Warning- LM317K duds from Digikey

[David Aurora]

Just a quick heads up in the hope nobody else loses hours of time like I just did- these "new" LM317K regulators they have in stock don't even come close to meeting spec https://www.digikey.com.au/en/products/detail/evvo/LM317K/25188516

I just watched 3 in a row eat shit in a bog standard regulator circuit as soon as current draw exceeded around 1A. Not a thermal thing either, in every case they worked unloaded or at half an amp or so but died instantly once I increased the load.

I don't think I've ever gotten fake parts from Digikey before so I spent ages double checking everything around them but couldn't find an issue. Then when I threw in some replacements from a different brand the power supply fired up like it's supposed to and has already been stress tested hard for about an hour without a hiccup.

Digikey were super fast with their response and have already credited me for the purchase, but who knows how long they'll still be up on the site for purchase so take my advice and don't touch them with a ten foot pole.

[wraper]

What was input voltage?

[squadchannel]

EVVOSEMI?
There have been a unknown and unheard of manufacturers making their way onto Digikey recently.

search on Digikey, I always narrow down the manufacturer.

to-3 maybe hard to find.

TI is $130  very expensive  maybe Hi-rel products.

[TimFox]

I had exactly the same problem with TO-3 317s from another vendor:  they would pop immediately when the load current exceeded a critical value.  They should have gone into thermal limiting and survived.  I’ll try to find the details.

[floobydust]

WTF the EVVOSEMI schematic pass transistor is not a Darlington, but has Sziklai pair which smells like the '1117 but it's a lower voltage IC.
LM317K STEEL I used back in the late 1970's in my first power supply, a faithful horse.

[coppercone2]

one of the biggest features the LDO has going for it is the thermal shutdown. Its like the only reason you would parallel up LDO instead of using a pass transistor. !!

I would not call it out of spec, I would call it defective.

I use those parallel circuits that are shown in some data sheets because... each chip has a integrated over temp protection. you lose the feature when you start going with external transistors. In this case, if it breaks, why the hell would you bother??

big fail

[David Aurora]

What was input voltage?

Nah it was not an input/output voltage issue. This is from a mixing console that had survived with the original regulators for 30 years (it failed after the owner shorted something on a DIY servicing attempt) and with a different brand of regulators appears fine too. Issue is specific to these ones Digikey are selling.

[David Aurora]

EVVOSEMI?
There have been a unknown and unheard of manufacturers making their way onto Digikey recently.

search on Digikey, I always narrow down the manufacturer.

to-3 maybe hard to find.

TI is $130  very expensive  maybe Hi-rel products.

Yeah these were literally the only stocked version of the LM317K I could find from a reputable source unfortunately, usually I'd never touch an unknown brand like that. Who'd have thought I'd have had better results buying off eBay 

[magic]

Who'd have thought I'd have had better results buying off eBay 

Maybe better or maybe same

WTF the whole thing, including DigiKey not noticing that this isn't even claimed to be a true 317 circuit. Maybe it went into suicidal oscillation with whatever capacitors are present in the circuit? (Assuming that the schematic found by floobydust actually represents what's inside - 1117 is not as stable as 317).

[AVGresponding]

WTF the EVVOSEMI schematic pass transistor is not a Darlington, but has Sziklai pair which smells like the '1117 but it's a lower voltage IC.
LM317K STEEL I used back in the late 1970's in my first power supply, a faithful horse.

Maybe the battery that supplies the non-inverting reference of the current limiter is dead...

[floobydust]

Another option is the LT317K, LT317AK. I dumpster dived some new LT338AK from work, if you're in a pickle (jar) let me know. But these have much greater output current.

Many smaller Taiwan and china semi manufacturers are using the classic part number but selling you a "cost improved" redesigned part.
Example I got comparators slow as molasses 10usec switching times and at first thought I got ripped off. Then, digging a bit:
Unisonic LMV331 is 12usec 
TI LMV331 is 0.6usec, ST is 0.375usec, On-Semi 1.5usec

I complained to the vendor and they replied "we are just broker". Lesson learned.

I find Digi-Key uses their fine reputation to pimp for these unknown semi manufacturers, and the marketplace vendors as well. Not sure what loser MBA came up with this strategy.
It does piss me off, side loading disreputable vendors - for what exactly? They think they are a fulfillment center or something.

[thm_w]

I find Digi-Key uses their fine reputation to pimp for these unknown semi manufacturers, and the marketplace vendors as well. Not sure what loser MBA came up with this strategy.
It does piss me off, side loading disreputable vendors - for what exactly? They think they are a fulfillment center or something.

Marketplace is fine, if it were off by default for searching, and only came up if nothing was found. Usually I go to digikey first to search for a replacement part even if they are obsolete. So if they can supply a marketplace product that is genuine, its of value.

The new semi manufacturers, some have really good niche products, like 3peak and Goford semi. Do they produce any garbage knockoffs? Don't know.

This EVVO looks like all they make is knockoffs: https://www.evvosemi.com/en/

[Geoff-AU]

The EVVO datasheet Digikey links to has only 3 pages, the first is a title page, last is a disclaimer, and page 2 has "actual specs" including a pinout (but no pins labeled anywhere, which one is which?).  No logic diagram or anything, and it's 1/3 the price of anything else.

Shame on you, Digikey.  Shame on you.

[wraper]

BTW isn't there some bipolar supply? I don't buy they are necessarily bad parts with no schematic or its description whatsoever. It may easily be they are less robust to something that goes wrong in your circuit, but part from another brand survives more of particular abuse than other. Circuit working does not necessarily mean it has no mistakes, oversights.

[magic]

If you still have a surviving sample you could check dropout voltage to confirm if it really is a 1117 rather than 317.
Then make sure that input/output capacitors follow 1117 stability guidelines and try again.
Or test it at low voltage (less heat if things go wrong) and see if it's oscillating, drifting out of spec, or doing some other suspicious stuff.

What was the behavior of the "dead" ones?

[Ice-Tea]

That datasheet 

High level of "trust me bro!"

[David Aurora]

BTW isn't there some bipolar supply? I don't buy they are necessarily bad parts with no schematic or its description whatsoever. It may easily be they are less robust to something that goes wrong in your circuit, but part from another brand survives more of particular abuse than other. Circuit working does not necessarily mean it has no mistakes, oversights.

You don't buy they're bad parts, when 3 out of 3 failed and alternate brands don't? On a dead simple straight-from-the-datasheet-design regulator? That sounds like a you problem.

[wraper]

BTW isn't there some bipolar supply? I don't buy they are necessarily bad parts with no schematic or its description whatsoever. It may easily be they are less robust to something that goes wrong in your circuit, but part from another brand survives more of particular abuse than other. Circuit working does not necessarily mean it has no mistakes, oversights.

You don't buy they're bad parts, when 3 out of 3 failed and alternate brands don't? On a dead simple straight-from-the-datasheet-design regulator? That sounds like a you problem.

You provided no description of the circuit whatsoever. I'd trust Digikey selling working parts rather than someone refusing to provide any context of how they fail. How could they fail without being bad parts? For example you may have bipolar PSU with no reverse polarity protection on the output, high output capacitance with no reverse direction diode across Vreg, high inductance on the output. As I already said, the fact of parts from some manufacturers surviving in your circuit does not necessarily mean that circuit in fine.

[Geoff-AU]

I agree that usually the burden of proof lies with the designer, to prove that they haven't abused the part.  Few details of the failing circuit were given.

But in this case, given the datasheet....  they're junk parts.  Not worth the time publishing and defending a circuit design. 

[David Aurora]

BTW isn't there some bipolar supply? I don't buy they are necessarily bad parts with no schematic or its description whatsoever. It may easily be they are less robust to something that goes wrong in your circuit, but part from another brand survives more of particular abuse than other. Circuit working does not necessarily mean it has no mistakes, oversights.

You don't buy they're bad parts, when 3 out of 3 failed and alternate brands don't? On a dead simple straight-from-the-datasheet-design regulator? That sounds like a you problem.

You provided no description of the circuit whatsoever. I'd trust Digikey selling working parts rather than someone refusing to provide any context of how they fail. How could they fail without being bad parts? For example you may have bipolar PSU with no reverse polarity protection on the output, high output capacitance with no reverse direction diode across Vreg, high inductance on the output. As I already said, the fact of parts from some manufacturers surviving in your circuit does not necessarily mean that circuit in fine.

Not sure if you're going out of your way to be a shit stirrer or you're actually this daft, but I'll play along one last time.

First of all, I've already told you- it's literally the bog standard LM317 circuit seen in every data sheet ever written and about 10 billion products out there. A few caps, a couple resistors and some protection diodes. If you don't know what that circuit looks like you probably shouldn't be in this thread shooting your mouth off.

Second, as I also already mentioned- save for the studio owner blowing up the old regulator (I don't know exactly what he did, but the supply died while he was messing round doing a DIY recap of the board, apparently while he was probing something), the circuit has worked for decades with an old LM317 and is again working fine with a new one from a different brand. When 3 out of 3 of the Digikey ones failed within the first 60 seconds in the same circuit only an absolute muppet would be standing here blaming the circuit.

Third- Digikey provided a refund immediately without requesting returns or more than a one paragraph rundown of what happened. I'd bet dollars to donuts that means it's not the first issue they've had with these.

[bdunham7]

When 3 out of 3 of the Digikey ones failed within the first 60 seconds in the same circuit only an absolute muppet would be standing here blaming the circuit.

While I agree that a defective part is the most likely explanation by far, one not too farfetched possibility is that the circuit exceeds the specifications of the LM317K and the original and "good" parts also exceed their specs by enough to let them work without issue.  In this case the "bad" parts simply meet the specs and promptly blow up when they're exceeded.  I doubt very much that that is the issue, but it's not a ridiculous question.  Similar things have happened in other contexts. 

I've seen a similar issue with a TO-220 LM317 where the voltage would drop when loaded.  This was powering a very old 8088 controller board and under certain conditions the load would go up for a few CPU cycles causing a brownout or reset.  The issue was the apparent internal impedance of the regulator was too high and after I replaced it and bench tested the regulator with power supply, it behaved just like yours apparently did--at about 40-50% of rated current it just popped without the overheat protection or current limiting kicking in.  The regulator was a recent Amazon or eBay purchase (not by me).

[David Aurora]

When 3 out of 3 of the Digikey ones failed within the first 60 seconds in the same circuit only an absolute muppet would be standing here blaming the circuit.

While I agree that a defective part is the most likely explanation by far, one not too farfetched possibility is that the circuit exceeds the specifications of the LM317K and the original and "good" parts also exceed their specs by enough to let them work without issue.  In this case the "bad" parts simply meet the specs and promptly blow up when they're exceeded.  I doubt very much that that is the issue, but it's not a ridiculous question.  Similar things have happened in other contexts. 

I've seen a similar issue with a TO-220 LM317 where the voltage would drop when loaded.  This was powering a very old 8088 controller board and under certain conditions the load would go up for a few CPU cycles causing a brownout or reset.  The issue was the apparent internal impedance of the regulator was too high and after I replaced it and bench tested the regulator with power supply, it behaved just like yours apparently did--at about 40-50% of rated current it just popped without the overheat protection or current limiting kicking in.  The regulator was a recent Amazon or eBay purchase (not by me).

1A current draw does not exceed the specs of an LM317K. And it has nothing to do with circuit quirks either because the Digikey ones failed in both the unit itself as well as on a DC load.

I think what a couple of people are also missing here is that a genuine LM317 has multiple degrees of protection. So while I was using these as replacements in a circuit that is literally the typical application circuit in the datasheet that we've all seen 10,000 times(!), even if the designer had done something stupid like trying to pull too much current, giving them too much input voltage or inadequate heat sinking a genuine part should go into protection mode. These things appear to go open circuit permanently immediately upon the current draw hitting 1A or so.

I also notice that there doesn't seem to be any trace of an LM317K on this manufacturers website, it could be that it's a brand new offering from them and we're the beta testers, or it could be that they've already been pulled because of shit like this

[bdunham7]

I think what a couple of people are also missing here is that a genuine LM317 has multiple degrees of protection...... trying to pull too much current, giving them too much input voltage or inadequate heat sinking a genuine part should go into protection mode. These things appear to go open circuit permanently immediately upon the current draw hitting 1A or so.

We're aware of the protection features, but the one thing you have wrong there with protection is including input voltage--that will kill them instantly.  It isn't actually input voltage, it's the difference between input and output that matters.  One way protection functions can fail is when you have an input voltage higher than the maximum but the output voltage is also high enough so that the difference is less than the max.  Assuming that circuit survives startup, it will work fine until one day it overheats or gets overloaded and then right as the regulator either limits current or goes into thermal shutdown the max voltage spec is suddenly exceeded and it pops.  I think that's why the question arose. 

However, I think you just got regulators with the same crappy die that mine had and they went poof in the same manner and for the same reasons.  The input voltage question is mildly interesting because that crappy chip might also go poof at a lower-than-spec voltage differential and still work OK in a lot of circuits.  If you still have your original and fake replacement parts, perhaps peeling the can open and having a look would be interesting as well.

[squadchannel]

install a pot in an ADJ, VIN is within rating, Iout is within rating, and you use it normally, and it breaks, then it is either defective, designed that way, or are being stingy to keep it cheap.

curious what kind of die design it is.
if it is the same as the 1117 series, it will have to be broken.
1117 has an image of being easily broken.

[schmitt trigger]

If you still have an undamaged device, you may want to drop a line to Noopy, to see if he would be interested in taking some die pictures and compare them against other images of the same device. But different vendor.

[wraper]

1A current draw does not exceed the specs of an LM317K. And it has nothing to do with circuit quirks either because the Digikey ones failed in both the unit itself as well as on a DC load.

I did not say anything about the current draw as by itself it should not be an issue for a good part.

I think what a couple of people are also missing here is that a genuine LM317 has multiple degrees of protection. So while I was using these as replacements in a circuit that is literally the typical application circuit in the datasheet that we've all seen 10,000 times(!), even if the designer had done something stupid like trying to pull too much current, giving them too much input voltage or inadequate heat sinking a genuine part should go into protection mode. These things appear to go open circuit permanently immediately upon the current draw hitting 1A or so.

It is supposed to have only overcurrent and overtemperature protection. Other than that, it's not hard to kill it.

Third- Digikey provided a refund immediately without requesting returns or more than a one paragraph rundown of what happened. I'd bet dollars to donuts that means it's not the first issue they've had with these.

It has nothing to do with quality of parts. It's just how DK responds to customer complaints.

it's literally the bog standard LM317 circuit seen in every data sheet ever written and about 10 billion products out there. A few caps, a couple resistors and some protection diodes. If you don't know what that circuit looks like you probably shouldn't be in this thread shooting your mouth off.

I know how standard circuit looks like. But I don't have a crystal ball to know what you actually did, what you consider non-inportant and not worth mentioning may be very important to others for finding the failure root cause. And you still did not say input voltage.

[David Aurora]

I think what a couple of people are also missing here is that a genuine LM317 has multiple degrees of protection...... trying to pull too much current, giving them too much input voltage or inadequate heat sinking a genuine part should go into protection mode. These things appear to go open circuit permanently immediately upon the current draw hitting 1A or so.

We're aware of the protection features, but the one thing you have wrong there with protection is including input voltage--that will kill them instantly.  It isn't actually input voltage, it's the difference between input and output that matters.  One way protection functions can fail is when you have an input voltage higher than the maximum but the output voltage is also high enough so that the difference is less than the max.  Assuming that circuit survives startup, it will work fine until one day it overheats or gets overloaded and then right as the regulator either limits current or goes into thermal shutdown the max voltage spec is suddenly exceeded and it pops.  I think that's why the question arose. 

However, I think you just got regulators with the same crappy die that mine had and they went poof in the same manner and for the same reasons.  The input voltage question is mildly interesting because that crappy chip might also go poof at a lower-than-spec voltage differential and still work OK in a lot of circuits.  If you still have your original and fake replacement parts, perhaps peeling the can open and having a look would be interesting as well.

I didn't think I'd need to say this guys, but this isn't my first day at an electronics bench. Jesus Christ. Yes, I understand how to use and how not to use a regulator.

This post wasn't a question about whether or not the designer made a mistake, it was a warning to others not to use what I've confirmed as dud parts in the hopes that it would save someone else some time/money.

The reason I didn't bother including every measurement is because it was of no relevance, everything was already checked before I went blaming parts and found to be well within safe operating areas. And again, the design is a tried and true, very basic LM317 circuit that works in more devices than you can poke a stick at. Nobody is reinventing the wheel here pushing this thing beyond its capabilities.

The finer details about what these parts can take is kind of an interesting side note I guess, but not the point here. The point is that these parts are marked as LM317K parts but break when used well within the safe operating area of an LM317K.

[David Aurora]

If you still have an undamaged device, you may want to drop a line to Noopy, to see if he would be interested in taking some die pictures and compare them against other images of the same device. But different vendor.

That would be cool but all 3 I had failed

[David Aurora]

1A current draw does not exceed the specs of an LM317K. And it has nothing to do with circuit quirks either because the Digikey ones failed in both the unit itself as well as on a DC load.

I did not say anything about the current draw as by itself it should not be an issue for a good part.

I think what a couple of people are also missing here is that a genuine LM317 has multiple degrees of protection. So while I was using these as replacements in a circuit that is literally the typical application circuit in the datasheet that we've all seen 10,000 times(!), even if the designer had done something stupid like trying to pull too much current, giving them too much input voltage or inadequate heat sinking a genuine part should go into protection mode. These things appear to go open circuit permanently immediately upon the current draw hitting 1A or so.

It is supposed to have only overcurrent and overtemperature protection. Other than that, it's not hard to kill it.

Third- Digikey provided a refund immediately without requesting returns or more than a one paragraph rundown of what happened. I'd bet dollars to donuts that means it's not the first issue they've had with these.

It has nothing to do with quality of parts. It's just how DK responds to customer complaints.

it's literally the bog standard LM317 circuit seen in every data sheet ever written and about 10 billion products out there. A few caps, a couple resistors and some protection diodes. If you don't know what that circuit looks like you probably shouldn't be in this thread shooting your mouth off.

I know how standard circuit looks like. But I don't have a crystal ball to know what you actually did, what you consider non-inportant and not worth mentioning may be very important to others for finding the failure root cause. And you still did not say input voltage.

I feel like I'm arguing about whether or not the buoyancy of the toilets played a part in the sinking of the Titanic here.

Again- if there was an issue with the circuitry around it this wouldn't be a thread. I'm not asking for advice, I'm warning others not to trust something I've confirmed to be a dud batch of parts.

[David Aurora]

install a pot in an ADJ, VIN is within rating, Iout is within rating, and you use it normally, and it breaks, then it is either defective, designed that way, or are being stingy to keep it cheap.

curious what kind of die design it is.
if it is the same as the 1117 series, it will have to be broken.
1117 has an image of being easily broken.

This is essentially what I've explained a thousand times now already. In a valid and working design, well within ratings these parts failed. That's the whole point of the post.

[coppercone2]

somehow this thread is starting to feel like

[magic]

These things appear to go open circuit permanently immediately upon the current draw hitting 1A or so.

Can you find any diodes between the input and other pins? Maybe it's simply a bond wire fusing because it was too wimpy.

[wraper]

I feel like I'm arguing about whether or not the buoyancy of the toilets played a part in the sinking of the Titanic here.

I feel like "it's garbage because trust me bro". Why in hell would you withhold even input voltage? I see no plausible reason to do so except it being very close to the max spec or exceeding it.

[thm_w]

If you still have an undamaged device, you may want to drop a line to Noopy, to see if he would be interested in taking some die pictures and compare them against other images of the same device. But different vendor.

Digikey still sells them so you are free to order more, or I can send the funds if someone wants to in depth analyze them.

[David Aurora]

I feel like I'm arguing about whether or not the buoyancy of the toilets played a part in the sinking of the Titanic here.

I feel like "it's garbage because trust me bro". Why in hell would you withhold even input voltage? I see no plausible reason to do so except it being very close to the max spec or exceeding it.

Nobody is "withholding" anything. Can you not read? As stated over and over again, everything was WELL within safe operating areas. If something had been either exceeding ratings or even close to it there would be no thread about this because the explanation would be that the components were being abused.

You seem to have this idea in your head that the designer or myself are shoving some ridiculous voltage into the things, no idea where this assumption comes from or why you're so hell bent on repeating it. Once again- I'm not asking for your input on why the components failed, the answer to that is obvious to anyone with basic reading skills. I'm warning others that these components are failing in completely standard circuits under completely safe conditions according to their specifications in the datasheet. If you're too thick to comprehend this that's your issue.

[wraper]

Is that Vin number under NDA? LMAO  . Apparently explaining actual conditions of failure is too much to ask. I guess plebs like me do not deserve to know anything more than amorphous "WELL within safe operating areas".

[David Aurora]

Is that Vin number under NDA? LMAO 

For real mate- this post was a courtesy warning to other techs. If you don't like it you're absolutely welcome to go fuck yourself.

[David Aurora]

Is that Vin number under NDA? LMAO  . Apparently explaining actual conditions of failure is too much to ask. I guess plebs like me do not deserve to know anything more than amorphous "WELL within safe operating areas".

Oh god, you're doing post edits now after I reply?

Here's the thing- your sarcastic pleb comment actually nails it. You don't seem to be understanding that you truly are an absolute nobody to me. You're just some dude with seventeen fucking thousand posts on an internet forum(!) who has obsessively latched onto a completely irrelevant detail thinking he's cracked a major mystery like Scooby Doo.

What exactly is the payoff you're hoping to get from carrying on barking up the wrong tree for 2 pages? Like do you genuinely think at some point I'm going to suddenly go "Oh shit, wraper was right, I hadn't checked the circuit! Vin was actually 100V/1000/10,000V! What a genius!"? You know, despite having already specified not only that circuit conditions were absolutely not the issue, that 3 out of 3 of the brand in question failed yet none of the alternatives had and that the replacements had been stress tested in the same circuit and were having absolutely no issues. And to top it off, in a classic circuit straight out of the datasheet.

Time for you to get a hobby outside of shitposting I reckon.

[5U4GB]

somehow this thread is starting to feel like

Or this :-).

[David Aurora]

somehow this thread is starting to feel like

Or this :-).

Pretty much.

The whole point of the original post was just to say "Hey, these things have popped up on Digikey replacing out of production parts but be wary because the ones I received were dodgy". I really wasn't expecting a pissing match over having some random guy on here demanding detailed measurements 

[Shock]

Just say they are a chinese reseller, people tend to give you the benefit of the doubt then. The ones that still hold out are anti west shills.

[5U4GB]

The whole point of the original post was just to say "Hey, these things have popped up on Digikey replacing out of production parts but be wary because the ones I received were dodgy". I really wasn't expecting a pissing match over having some random guy on here demanding detailed measurements 

It's his posting style.  Having become aware of the poster's name I've now noticed it in other threads as well, e.g. here (that's the response to the post).

[Psi]

If you still have an undamaged device, you may want to drop a line to Noopy, to see if he would be interested in taking some die pictures and compare them against other images of the same device. But different vendor.

That would be cool but all 3 I had failed

It's a TO3, take a hacksaw to the metal can and lets have a look what's inside

[David Aurora]

If you still have an undamaged device, you may want to drop a line to Noopy, to see if he would be interested in taking some die pictures and compare them against other images of the same device. But different vendor.

That would be cool but all 3 I had failed

It's a TO3, take a hacksaw to the metal can and lets have a look what's inside

Definitely chucking that on the to do list, might even cut it open tomorrow if I get a chance

[Siwastaja]

I really wasn't expecting a pissing match over having some random guy on here demanding detailed measurements 

No one even asked for "detailed measurements" (although those would be helpful too), just the most basic information like a schematic and input voltage. What is so difficult in disclosing them? This is what every sensible engineer would ask for as the very first thing, and that's what the manufacturer is also going to ask when you complain. Now you only made yourself look like a total nutjob and given the extreme arrogance in the way you communicate the issue I have a very strong feeling that it is pretty likely you just did something wrong and the parts are fine.

"Digikey is sending bad parts, trust me bro I did everything right" why should I care?
vs.
"I suspect bad parts, I used them per this schematic, do you see anything wrong in what I did and would you agree that it could be bad parts?"  - and now we are talking.


I mean, if you are an engineer and not some kind of con artist, the very first thing you do upon suspecting bad parts is you do take detailed measurements and ask for second opinion, then work with the manufacturer or distributor (and discuss with peers e.g. on a forum like this), but now it seems you are here just to vent and throw temper tantrums?

[nctnico]

Buying parts from a reputable vendor doesn't mean they are good. Personally I avoid the Multicomp and RS-pro branded parts from Farnell and RS as these are rather crappy. I can buy better quality from Aliexpress. And then there are shady semiconductor manufacturers like Sipex which make devices which fail within a year. Likely EVVO is a similar manufacturer. Given the data David has provided the LM317K he bought are obviously fake. At least the thermal protection should have kicked in. The low price is also a bit of a giveaway. Paying less than US$10 for a highly antiquated part is a red flag.

[David Aurora]

I really wasn't expecting a pissing match over having some random guy on here demanding detailed measurements 

No one even asked for "detailed measurements" (although those would be helpful too), just the most basic information like a schematic and input voltage. What is so difficult in disclosing them? This is what every sensible engineer would ask for as the very first thing, and that's what the manufacturer is also going to ask when you complain. Now you only made yourself look like a total nutjob and given the extreme arrogance in the way you communicate the issue I have a very strong feeling that it is pretty likely you just did something wrong and the parts are fine.

"Digikey is sending bad parts, trust me bro I did everything right" why should I care?
vs.
"I suspect bad parts, I used them per this schematic, do you see anything wrong in what I did and would you agree that it could be bad parts?"  - and now we are talking.


I mean, if you are an engineer and not some kind of con artist, the very first thing you do upon suspecting bad parts is you do take detailed measurements and ask for second opinion, then work with the manufacturer or distributor (and discuss with peers e.g. on a forum like this), but now it seems you are here just to vent and throw temper tantrums?

Who am I supposed to be asking for a second opinion here? You? Wraper? I don't know who either of you are or whether you know your ass from your elbow, why would I feel the need to run measurements past you before posting a warning that I received some bad parts? As I've said earlier, I took measurements as soon as I noticed an issue and that led me to suspect bad parts, I was able to reproduce the issue multiple times, and reported the issue to Digikey before posting here. Did you actually read any posts in this thread before deciding to chime in? The original post literally says all of this.

Yet again for the slow kids:
- The circuit is straight out of any LM317 datasheet out there, you're welcome to look it up if you're so inexperienced that you can't immediately visualise it. Diode for diode, resistor for resistor, it's the same damn circuit you'll find in about a million power supplies out there.
- This is a repair job, not a design job. The old regulator worked for decades and it didn't die on its own (the owner was attempting DIY work and killed it). The new regulator works absolutely fine (even when stress tested much harder than the actual circuit requires). 3 out of 3 of the regulators from Digikey failed, one in circuit and two while testing the supply with a DC load. All failed instantly once current draw reached 1A or so (and I do mean instantly, not a single one of these things lived more than about a minute including the time they were being tested at 100mA/500mA/etc).
- There is no big secret or grand conspiracy about the circuit details. The voltage differential is something like 8 volts for fucks sake (25V or so in for 17V out), hence why I've said over and over again that everything was running well within safe operating areas. Would you also like to know exact heatsink temperature even though we know it didn't overheat? Current down to the uA even though we know current limit wasn't exceeded? Should we go back and check moon phase? Do you want to know what I had for lunch that day?
- The post was a courtesy warning about parts failing when they shouldn't be, not a request for a multi page chat about it. If you don't believe me and think the parts are fine then go use them. Best of luck with that.

[David Aurora]

Alright, I got curious and peeled some regulators open. The ST one is the original dead one, there's one of the dead EVVO ones and then I sacrificed a working iSC one for more comparison options. I have limited zoom and even more limited knowledge of what I'm looking for here, but the things that stuck out to me are all three dies have "317" written on them (which I would assume rules out a simple silkscreening lie on the cases) and the ST/iSC ones have a significant raised section of the case under the dies whereas the EVVO one is dead flat. Not sure if that has much effect on the thermal side of things?

Uploads coming in next few posts...

[David Aurora]

ST

[David Aurora]

EVVO

[David Aurora]

iSC

[wraper]

Personally I avoid the Multicomp and RS-pro branded parts from Farnell and RS as these are rather crappy. I can buy better quality from Aliexpress.

They are YMMV, not crappy. Often you can figure out their origin from the datasheets. For example I bought cheap Multicomp 3D printer filament that was eSun in disguise, and I figured it out before the purchase. Sometimes you can find top tier parts in disguise. You can get better quality Aliexpress, although your chances of that are very slim.

[magic]

As expected, you blew the input fuse.

[wraper]

iSC

Construction looks fine. Presence of the heat spreader under the die indicates they did not save every penny. Bottom of the barrel parts usually do not have it.

[David Aurora]

As expected, you blew the input fuse.

To be fair, I was pretty rough opening them so it's possible any high impedance air gaps were generated in the vice

I'll crack open the other two EVVOs later on and see if there's much difference

[David Aurora]

iSC

Construction looks fine. Presence of the heat spreader under the die indicates they did not save every penny. Bottom of the barrel parts usually do not have it.

These are NOT the dud parts from Digikey, these are the ones I got later that work fine. But it looks like the parts you've been defending don't have it. Interesting...

[wraper]

iSC

Construction looks fine. Presence of the heat spreader under the die indicates they did not save every penny. Bottom of the barrel parts usually do not have it.

These are NOT the dud parts from Digikey, these are the ones I got later that work fine. But it looks like the parts you've been defending don't have it. Interesting...

I was not defending it. I was asking for more info to make an informed opinion. As of EVVO part I did not notice on previous page, die size looks ok. Bond wires are somewhat thin but probably enough for rated current. Bonding to the package looks to be some transparent glue which is not good. Lack of heat spreader, it may be ok if the package has thick bottom, otherwise not good. Certainly not worth $10, I'd use $ 0.30 TO-220 part instead even if I had to pay $10 for it.

[David Aurora]

iSC

Construction looks fine. Presence of the heat spreader under the die indicates they did not save every penny. Bottom of the barrel parts usually do not have it.

These are NOT the dud parts from Digikey, these are the ones I got later that work fine. But it looks like the parts you've been defending don't have it. Interesting...

I was not defending it. I was asking for more info to make an informed opinion. As of EVVO part I did not notice on previous page, die size looks ok. Bond wires are somewhat thin but probably enough for rated current. Bonding to the package looks to be some transparent glue which is not good. Lack of heat spreader, it may be ok if the package has thick bottom, otherwise not good. Certainly not worth $10, I'd use $ 0.30 TO-220 part instead even if I had to pay $10 for it.

So we spent 3 pages to get your "expert" conclusion that it's a shitty part. Which I had already deduced from testing before my first post about it. Fantastic. You're my hero.

[wraper]

Interesting, found exactly the same die in some unknown 'LM338K' https://zeptobars.com/en/read/LM338K-5A-LDO-TO-3-TO-220

[magic]

That's a Darlington output so it isn't a 1117, they just copy-pasted wrong schematic into their datashit.

And somebody even upgraded it to 5A, nice

[bdunham7]

Alright, I got curious and peeled some regulators open. The ST one is the original dead one, there's one of the dead EVVO ones and then I sacrificed a working iSC one for more comparison options. I have limited zoom and even more limited knowledge of what I'm looking for here, but the things that stuck out to me are all three dies have "317" written on them (which I would assume rules out a simple silkscreening lie on the cases) and the ST/iSC ones have a significant raised section of the case under the dies whereas the EVVO one is dead flat. Not sure if that has much effect on the thermal side of things?

Nice pictures!  Unless I'm blind, there's no die damage on the EVVO part and there doesn't seem to be any obvious chicanery with the die size or type.  So perhaps it is just a thrifted (but not "fake") part that happens to have a defect that happened somewhere in the manufacturing process.  IDK how much the heat spreader matters because an LM317K isn't intended to dissipate hundreds of watts like a 2N3055 or some of the even larger power transistors.  In any case, even if the part were thermally insufficient it shouldn't have completely failed.  b/t/w, did it fail open or short? 

[David Aurora]

Alright, I got curious and peeled some regulators open. The ST one is the original dead one, there's one of the dead EVVO ones and then I sacrificed a working iSC one for more comparison options. I have limited zoom and even more limited knowledge of what I'm looking for here, but the things that stuck out to me are all three dies have "317" written on them (which I would assume rules out a simple silkscreening lie on the cases) and the ST/iSC ones have a significant raised section of the case under the dies whereas the EVVO one is dead flat. Not sure if that has much effect on the thermal side of things?

Nice pictures!  Unless I'm blind, there's no die damage on the EVVO part and there doesn't seem to be any obvious chicanery with the die size or type.  So perhaps it is just a thrifted (but not "fake") part that happens to have a defect that happened somewhere in the manufacturing process.  IDK how much the heat spreader matters because an LM317K isn't intended to dissipate hundreds of watts like a 2N3055 or some of the even larger power transistors.  In any case, even if the part were thermally insufficient it shouldn't have completely failed.  b/t/w, did it fail open or short?

I'll upload pictures later, but I cut open the second one and that one had some residue that looked like something had popped violently. Haven't opened the third yet. Will double check connections between pins when I do that and update

[David Aurora]

Pics of the other 2 failed units incoming. On both of these it looks a lot more obvious that the bond wires vaporised

Also, the manufacturer has now gotten back to me about questions I had sent them. They say it's not a new product to their line and while it wasn't up on the English version of their site it is there on the Chinese version, so it hasn't been pulled from production or anything like that.

[David Aurora]

Third unit

[bdunham7]

Pics of the other 2 failed units incoming. On both of these it looks a lot more obvious that the bond wires vaporised

CCA bond wires??   

Is there any way you can measure the thickness of those bond wires?

[wraper]

Pics of the other 2 failed units incoming. On both of these it looks a lot more obvious that the bond wires vaporised

CCA bond wires??   

Is there any way you can measure the thickness of those bond wires?

Aluminum bond wire is normal and AFAIK is a primary choice if not using much more expensive gold. Other ICs seem to use aluminum too. However others seem to have thicker wires and a pair of them from in/out. Die in EVVO IC also has a place for a pair of bond wires but there is only one bonded. Also input bond wire is very long in comparison which makes it heat more.

[KE5FX]

Not trying to defend the shady parts, but I assume you've checked the capacitance and/or ESR of the bulk cap(s) upstream of the regulator?  If there's more (good) capacitance downstream than upstream, and no reverse-bias protection diode is present, that could explain both the original regulator failure and the new ones.  They could be taking some damage at power-off time that curtails their operating life drastically...

[Psi]

Not sure how practical this is for you, but if you have a microscope it would be interesting to touch a ultra pointed probe on one of the intact bond wires at the die and push 1A through it from CC psu to one of the output terminals while accuracy measuring uV/mV across it using 2nd meter so you can calc resistance.

[David Aurora]

Pics of the other 2 failed units incoming. On both of these it looks a lot more obvious that the bond wires vaporised

CCA bond wires??   

Is there any way you can measure the thickness of those bond wires?

Probably not anywhere near accurately unfortunately 

[Ice-Tea]

Is it just me or is the workmanship on these bond wire pretty bad?

[David Aurora]

Not trying to defend the shady parts, but I assume you've checked the capacitance and/or ESR of the bulk cap(s) upstream of the regulator?  If there's more (good) capacitance downstream than upstream, and no reverse-bias protection diode is present, that could explain both the original regulator failure and the new ones.  They could be taking some damage at power-off time that curtails their operating life drastically...

All of these questions have been covered many times in this thread now, but given that it got unnecessarily derailed for a couple pages I'll recap again to save you sifting through:
- The original regulator failed due to the owner of the unit poking around inside and apparently shorting something out. These new EVVO regulators failed but the newer ones of a different brand have worked just fine
- All power supply components had already been tested and confirmed working. Protection diodes ARE present
- These regulators not only failed while powering the unit, but also when the power supply was only driving a DC load which rules out all other circuitry in the unit
- They didn't even last long enough to be powered down once so definitely couldn't have copped that kind of damage. The first one failed at power up, and it was confirmed on the other two that any load exceeding about 1A would kill them instantly (the actual load is something like 1.2A from memory so it makes sense why the first one died while powering on the unit). On that note, I can't remember if I mentioned it originally but just to state the obvious, the supply WAS checked and working before connecting to the unit, just not heavily load tested as I had no reason to think these couldn't provide the rated current and I'd already confirmed the rest of the unit powered up OK with a bench supply

[David Aurora]

Not sure how practical this is for you, but if you have a microscope it would be interesting to touch a ultra pointed probe on one of the intact bond wires at the die and push 1A through it from CC psu to one of the output terminals while accuracy measuring uV/mV across it using 2nd meter so you can calc resistance.

That would definitely be interesting! Might be a bit more effort than I want to throw at it at this point though, plus I don't know how much faith we can realistically place on the remaining bond wires. I feel like it's anyones guess whether they were also compromised during failure or if they'd still be in the same shape they left the factory in

[floobydust]

Those fused bonding wires look too skinny. The ADJ teminal I could see a thinner gauge possible but whoever packaged the part seems to have saved a millipenny.
I wonder the die size if they shrunk it too much or it's also for 150mA versions etc . and they zap current-sense resistors to set that?
I don't see a heat spreader either.

[mikerj]

Might be worth sending the pics to Digikey so they can get an idea of the shoddy crap they are selling.

[mikerj]

looks different to me 

This was never in question.  All three dies are different.

[floobydust]

So these are surplus IC dies? Scrounged and then amateur packaged for great profit I guess. But gotta save micropennies lol.
Many tech papers on fusing limits for IC bonding wires. Aluminum ain't like copper or gold.

[wraper]

So these are surplus IC dies? Scrounged and then amateur packaged for great profit I guess. But gotta save micropennies lol.
Many tech papers on fusing limits for IC bonding wires. Aluminum ain't like copper or gold.

All 3 EVVO ICs have the same die. Other pictures are from ICs from different manufacturers.

[floobydust]

That makes more sense. I was trying to compare bonding wire thickness between them, then saw the heat spreader missing and gave up.
I thought the wire fusing is the problem with the EVVO parts.

What pisses me off is search engines are returning results from Digi-Key to EVVO semiconductor parts.
Searched for "2SC1623" and it's a top hit. UGH. WOULD NOT BUY based on OP's experience.

[David Aurora]

So these are surplus IC dies? Scrounged and then amateur packaged for great profit I guess. But gotta save micropennies lol.
Many tech papers on fusing limits for IC bonding wires. Aluminum ain't like copper or gold.

I don't know for sure, but I don't get the vibe they're surplus things like that. I've been chatting a bit with the manufacturer and they appear to be taking it seriously and wanting to gather info. I don't think a company just scavenging would bother with even replying to the first email I sent, but these guys are discussing it with me even though they're aware I only bought three of the things and therefore I'm clearly not an important customer

It's funny though, before this I've never given a shit about the guts of chips beyond the ideas of "Wow, it's cool that a chip can do this" or "The way they make dies is super interesting", but now I keep wanting a better microscope to look with more detail and I want to know everything about these bond wires 

[magic]

It's from China, of course it has to be as shoddy as they can get away with, and sometimes a little shoddier to test the waters. Somebody makes the dice, somebody packages them, everybody cuts all imaginable corners. Then somebody else sells them as LM338 on eBay. Business as usual.

For each customer who blows them up at 1.2A there is probably a bunch of customers who happily use them at 0.6A. It's all about optimizing the overall profit, specs be damned. They usually don't even bother creating their own datasheets and copy paste plots from other vendors.

BTW, do you still have some remaining parts? Could you see what happens with 3.3V input, ADJ to ground and 1Ω load on the output, to absolutely eliminate any possibility of transient overcurrents due to overvoltage, maybe rails going negative because of opposite polarity regulators, stuff like that? Would they blow up anyway?

[rsjsouza]

I don't know for sure, but I don't get the vibe they're surplus things like that. I've been chatting a bit with the manufacturer and they appear to be taking it seriously and wanting to gather info. I don't think a company just scavenging would bother with even replying to the first email I sent, but these guys are discussing it with me even though they're aware I only bought three of the things and therefore I'm clearly not an important customer

I agree with you. Also, if they ask for the parts back to do a Failure Analysis, then you know they are even more serious than the average supplier. FAs are expensive and time consuming processes and done mostly for manufacturers that want to stand behind their products.

It's funny though, before this I've never given a shit about the guts of chips beyond the ideas of "Wow, it's cool that a chip can do this" or "The way they make dies is super interesting", but now I keep wanting a better microscope to look with more detail and I want to know everything about these bond wires 

Due to the high import taxes in Brasil, which makes sourcing parts from reputable distributors an expensive venture, my friends there have been looking at dies for years to try to identify and spread the word about fake parts from the typical sources (aliexpress, banggood, mercadolivre, etc.). We are very fortunate to have options...

[floobydust]

It's a blessing the 317's didn't fail shorted - going full output to damage the console. A lot of op-amps downstream...
I've killed a few 317's over the decades. They have a natural weakness to backfeed or reverse-polarity spikes but this is well known and easily solved by adding the datasheet two protection diodes. LT317 only needs one.

I'll bet you could cut open a new EVVO part and watch the bonding wire glow at load.
It would show the packaging fail with too skinny a wire, and also a testing fail that should have caught that. Although it is a PITA to load test when there's no heatsink you can add.

[David Aurora]

It's a blessing the 317's didn't fail shorted - going full output to damage the console. A lot of op-amps downstream...
I've killed a few 317's over the decades. They have a natural weakness to backfeed or reverse-polarity spikes but this is well known and easily solved by adding the datasheet two protection diodes. LT317 only needs one.

I'll bet you could cut open a new EVVO part and watch the bonding wire glow at load.
It would show the packaging fail with too skinny a wire, and also a testing fail that should have caught that. Although it is a PITA to load test when there's no heatsink you can add.

Absolutely, I'd have been so pissed off if that had been the fault mode

If I had more spare time I'd have ordered a few more and played around with deeper testing/cutting open a working one and observing it/etc. I still might grab some next Digikey order and do it whenever I get to it

[Analog Kid]

I'll bet you could cut open a new EVVO part and watch the bonding wire glow at load.

Hey, why doesn't someone here actually do that?
Not only would it confirm substandard construction, it would also be kewl to watch!
Noopy?

[magic]

I don't think there is any doubt that the wires fused, particularly on samples 2 and 3.

The only interesting questions are:
1. Are they simply not rated for the 1.5A current of LM317, or was it some transient due to shoddy external circuit.
2. What would happen to the chip if the wires didn't fuse.

Other than that, it's just a classic case of "buy Chinese, buy twice".

[David Aurora]

I don't think there is any doubt that the wires fused, particularly on samples 2 and 3.

The only interesting questions are:
1. Are they simply not rated for the 1.5A current of LM317, or was it some transient due to shoddy external circuit.
2. What would happen to the chip if the wires didn't fuse.

Other than that, it's just a classic case of "buy Chinese, buy twice".

1. I really think we've covered the circuit conditions about 10 billion times at this point. The parts were in no way abused, and the failures weren't even all under the same conditions (one failed in the unit the power supply was from, two failed on a dummy load). 2 other brands of the same part have both worked fine there. Also, none of the devices ever reached 1.5A of current anyway (again, something I've covered thoroughly here), they died around 1A. Like I've said over and over again here, NO ratings were even close to being pushed here.
2. That is definitely the interesting question.

The "buy Chinese, buy twice" thing is dumb (frankly I find the constant Sinophobic echolalia around here absolutely moronic). The LM317K is out of production. The options were to buy these from Digikey or buy from random sellers on eBay/etc. I assumed Digikey would be the more reliable source for genuine parts than eBay etc, this didn't turn out to be the case but I think it was quite reasonable to make that assumption.

[magic]

1. I really think we've covered the circuit conditions about 10 billion times at this point.

The question has been asked 10 trillion times actually, but never answered.

There is more than one LM317 datasheet and typically more than one application schematic in each of them. And in dual rail systems additional protection is required, which IIRC isn't mentioned in those datasheets.

The parts were in no way abused, and the failures weren't even all under the same conditions (one failed in the unit the power supply was from, two failed on a dummy load).

The latter is closest to a reasonable argument you have made so far, albeit still no details of the test circuit.

2 other brands of the same part have both worked fine there.

Not an argument. They may have worked out of spec. This may have contributed to the failure of the original part.

Also, none of the devices ever reached 1.5A of current anyway (again, something I've covered thoroughly here), they died around 1A.

I gather that you have measured inrush current with a scope and aren't just making things up?

Like I've said over and over again here, NO ratings were even close to being pushed here.

Because "trust me bro".

Life would be much simpler if only people who are right were confident that they are right.

The "buy Chinese, buy twice" thing is dumb (frankly I find the constant Sinophobic echolalia around here absolutely moronic). The LM317K is out of production. The options were to buy these from Digikey or buy from random sellers on eBay/etc. I assumed Digikey would be the more reliable source for genuine parts than eBay etc, this didn't turn out to be the case but I think it was quite reasonable to make that assumption.

Ain't libel if it's true.

If DigiKey starts carrying transistor brands like ICS or Savantic, they will still be the garbage they are. And it looks like power ICs are similar story. And then of course lots of ICs with completely nonsense/misleading names, such as CMOS LDOs sold as 1117. Which kinda works most of the time, until it doesn't because you run into some important difference between these technologies.

[wraper]

The "buy Chinese, buy twice" thing is dumb (frankly I find the constant Sinophobic echolalia around here absolutely moronic). The LM317K is out of production. The options were to buy these from Digikey or buy from random sellers on eBay/etc. I assumed Digikey would be the more reliable source for genuine parts than eBay etc, this didn't turn out to be the case but I think it was quite reasonable to make that assumption.

It's not out of production. The third option is to pay though the nose https://www.ti.com/product/LM317-N-MIL/part-details/LM317K%20STEEL/NOPB?HQS=ocb-tistore-invf-invftransact-invf-store-octopart-wwe
IMHO the most sensible option is just to use TO-220 part. You can make a terminal adaptor PCB if you want to make it look nice.

[David Aurora]

1. I really think we've covered the circuit conditions about 10 billion times at this point.

The question has been asked 10 trillion times actually, but never answered.

There is more than one LM317 datasheet and typically more than one application schematic in each of them. And in dual rail systems additional protection is required, which IIRC isn't mentioned in those datasheets.

The parts were in no way abused, and the failures weren't even all under the same conditions (one failed in the unit the power supply was from, two failed on a dummy load).

The latter is closest to a reasonable argument you have made so far, albeit still no details of the test circuit.

2 other brands of the same part have both worked fine there.

Not an argument. They may have worked out of spec. This may have contributed to the failure of the original part.

Also, none of the devices ever reached 1.5A of current anyway (again, something I've covered thoroughly here), they died around 1A.

I gather that you have measured inrush current with a scope and aren't just making things up?

Like I've said over and over again here, NO ratings were even close to being pushed here.

Because "trust me bro".

Life would be much simpler if only people who are right were confident that they are right.

The "buy Chinese, buy twice" thing is dumb (frankly I find the constant Sinophobic echolalia around here absolutely moronic). The LM317K is out of production. The options were to buy these from Digikey or buy from random sellers on eBay/etc. I assumed Digikey would be the more reliable source for genuine parts than eBay etc, this didn't turn out to be the case but I think it was quite reasonable to make that assumption.

Ain't libel if it's true.

If DigiKey starts carrying transistor brands like ICS or Savantic, they will still be the garbage they are. And it looks like power ICs are similar story. And then of course lots of ICs with completely nonsense/misleading names, such as CMOS LDOs sold as 1117. Which kinda works most of the time, until it doesn't because you run into some important difference between these technologies.

Aaaand here we go again.

Get a life mate. You're the last man standing on this massive reach to look for anything but the obvious conclusion that the parts were duds. And you're still making the bold assumption that I am seeking or give the slightest shit about your opinion on it.

[David Aurora]

The "buy Chinese, buy twice" thing is dumb (frankly I find the constant Sinophobic echolalia around here absolutely moronic). The LM317K is out of production. The options were to buy these from Digikey or buy from random sellers on eBay/etc. I assumed Digikey would be the more reliable source for genuine parts than eBay etc, this didn't turn out to be the case but I think it was quite reasonable to make that assumption.

It's not out of production. The third option is to pay though the nose https://www.ti.com/product/LM317-N-MIL/part-details/LM317K%20STEEL/NOPB?HQS=ocb-tistore-invf-invftransact-invf-store-octopart-wwe
IMHO the most sensible option is just to use TO-220 part. You can make a terminal adaptor PCB if you want to make it look nice.

We're getting way into the weeds here if we're going to say it's not out of production because you can buy a version that costs more than 10 times the price of a standard one. Yes, the steel version exists, but over $120AUD for a single regulator IC is not the same as them being readily available. For all intents and purposes the LM317K is out of production.

It's easy to be an armchair expert with hindsight and someone else's experience, but if one of the most trusted major suppliers is selling new production parts for a reasonable price (i.e. these weren't $1 from eBay) why the fuck would my first instinct be to start hacking in a part with a totally different footprint? It's not at all unreasonable to have expected these parts to meet specs.

[thm_w]

Dropout is about 1.1V at no load. Quiescent current ~2mA. With 650R R1, adjusted R2 so output ~5.1V.

Maybe I'll measure current limiting at low input voltage tomorrow. Bond wires don't yet show up on thermal cam.

[Analog Kid]

Jesus F. Christ, those wires look tiny.

[floobydust]

Is that a visible kink in the middle of the output bonding wire? It's a long run and maybe they are damaged during assembly.

The LM317 has current-limiting and SOA protection for the pass transistor(s). It's not going to belt out a lot more current without its pass transistor failing. I won't fault the console's circuit here.

[David Aurora]

Dropout is about 1.1V at no load. Quiescent current ~2mA. With 650R R1, adjusted R2 so output ~5.1V.

Maybe I'll measure current limiting at low input voltage tomorrow. Bond wires don't yet show up on thermal cam.

(Attachment Link)

Certainly looks the same inside!

Same batch/date code (mine were all 2437)?

[bdunham7]

Is that a visible kink in the middle of the output bonding wire? It's a long run and maybe they are damaged during assembly.

That brings up the question of why they placed the die there instead of closer to or between the pins.  They could use thicker or double wires and not spend any more because they'd be less than half as long.  Sometimes cheap products aren't so much overthrifted as they are poorly implemented. 

[electrodacus]

Base on what I read and could see the fail was thermal fusing of the bond wire.
Voltage is fairly high 25V input and 17V output and there was a mention of DIY replacement of the capacitors.
If the new capacitors are higher capacity and or much lower internal impedance it cold be that fusing of the bond wire happens at power up due to transient  in charging the newly installed capacitor's on the output side.
If I understood correctly they failed few seconds after power up and in that case the bond wire is just incorrectly sized for the job.
The placement of the die seems fairly silly as having it closer to the input pin will have made that bond wire at least half the length reducing resistance and power drop on that to at least half.
So likely a product that does not meant spec and in that case Digikey should remove that component from sale.

[magic]

It's easy to be an armchair expert with hindsight and someone else's experience, but if one of the most trusted major suppliers is selling new production parts for a reasonable price (i.e. these weren't $1 from eBay) why the fuck would my first instinct be to start hacking in a part with a totally different footprint? It's not at all unreasonable to have expected these parts to meet specs.

It's unreasonable to expect any specs from a manufacturer who can't even be assed to produce a datasheet other than copy-pasting chunks of various text form other vendors which don't make sense together.

Or maybe they just mean something else than you think they do. Because actually, you have presented no evidence that the part fails to meet its specification:


 

Of course it might still be that the wire will fuse at 0.5A and that otherwise the die would overheat and short out at 1.0A.

[David Aurora]

Base on what I read and could see the fail was thermal fusing of the bond wire.
Voltage is fairly high 25V input and 17V output and there was a mention of DIY replacement of the capacitors.
If the new capacitors are higher capacity and or much lower internal impedance it cold be that fusing of the bond wire happens at power up due to transient  in charging the newly installed capacitor's on the output side.
If I understood correctly they failed few seconds after power up and in that case the bond wire is just incorrectly sized for the job.
The placement of the die seems fairly silly as having it closer to the input pin will have made that bond wire at least half the length reducing resistance and power drop on that to at least half.
So likely a product that does not meant spec and in that case Digikey should remove that component from sale.

As mentioned many times already, only the first failure happened in the unit that the owner had recapped, both other failures occurred on a DC load which rules out any capacitance issues downstream.

But yes, the units worked unloaded/lightly loaded but failed as soon as load was increased.

[David Aurora]

It's easy to be an armchair expert with hindsight and someone else's experience, but if one of the most trusted major suppliers is selling new production parts for a reasonable price (i.e. these weren't $1 from eBay) why the fuck would my first instinct be to start hacking in a part with a totally different footprint? It's not at all unreasonable to have expected these parts to meet specs.

It's unreasonable to expect any specs from a manufacturer who can't even be assed to produce a datasheet other than copy-pasting chunks of various text form other vendors which don't make sense together.

Or maybe they just mean something else than you think they do. Because actually, you have presented no evidence that the part fails to meet its specification:
(Attachment Link)
 

The placement of the die seems fairly silly as having it closer to the input pin will have made that bond wire at least half the length reducing resistance and power drop on that to at least half.

It seems to be placed right on the OUT pin, perhaps for thermal relief.

Sometimes cheap products aren't so much overthrifted as they are poorly implemented. 

The opposite is often true - carefully engineered to be maximally cheap and dodgy.

Are you new to electronics? Do you have any idea how many data sheets are copy/pastes of old data sheets from other manufacturers? Jesus Christ. Either way, if we assume a manufacturer is providing inaccurate data sheets and Digikey is still selling the part then why are you obsessed with trying to pin the blame on the end user for *checks notes* trusting a manufacturer and a supplier to make/sell a jelly bean part that has been around for nearly 50 years and sold by damn near every IC manufacturer that has ever existed? Get a grip.

Evidence wise, why do you think I owe you evidence? Why do you think your opinion matters?

It's Shrödingers Regulator apparently- you claim the part is cheap Chinese shit, but simultaneously keep saying the part only failed because of some mystery misuse conspiracy. Pick a shit theory and stick with it, you absolute peanut.

[magic]

Either way, if we assume a manufacturer is providing inaccurate data sheets and Digikey is still selling the part then why are you obsessed with trying to pin the blame on the end user for *checks notes* trusting a manufacturer and a supplier to make/sell a jelly bean part that has been around for nearly 50 years and sold by damn near every IC manufacturer that has ever existed? Get a grip.

I just suggest that you get a grip because this is the new normal now

It's Shrödingers Regulator apparently- you claim the part is cheap Chinese shit, but simultaneously keep saying the part only failed because of some mystery misuse conspiracy. Pick a shit theory and stick with it, you absolute peanut.

Shit circuit, shit part, smoke goes out. What's surprising about that?

Also, it's not me who blew this shit up so I have no vested interest in proving any theory about what went wrong. (Neither is it me who is on the hook for this thing continuing to work for another N years). I am equally prejudiced against audio gear and Chinese semiconductors, so whichever gets the blame I'm fine with it. But I'm curious what exactly went wrong and not accepting hand wavy non-explanations for this reason.

[David Aurora]

Either way, if we assume a manufacturer is providing inaccurate data sheets and Digikey is still selling the part then why are you obsessed with trying to pin the blame on the end user for *checks notes* trusting a manufacturer and a supplier to make/sell a jelly bean part that has been around for nearly 50 years and sold by damn near every IC manufacturer that has ever existed? Get a grip.

I just suggest that you get a grip because this is the new normal

It's Shrödingers Regulator apparently- you claim the part is cheap Chinese shit, but simultaneously keep saying the part only failed because of some mystery misuse conspiracy. Pick a shit theory and stick with it, you absolute peanut.

Shit circuit, shit part, smoke goes out. What's surprising about that?

Wait, it's a shit circuit? I thought you kept saying I hadn't provided enough info on the circuit? Interesting.

[temperance]

EEVBlog Forum, The Place to discuss your adventures with with real panda manure ad ∞

[magic]

Wait, it's a shit circuit? I thought you kept saying I hadn't provided enough info on the circuit? Interesting.

You did provide some information: the factory installed voltage regulator failed.
But it's not enough information to know for sure that the new one is crap.

You claim that something is junk, people ask questions. Surprising, isn't it?

[bdunham7]

But I'm curious what exactly went wrong and not accepting hand wavy non-explanations for this reason.

It seems pretty clear to me that the primary issue is the bond wires.  The regulator has current limiting so unless that is bogus as well, the only way the circuit could cause an overcurrent is by reverse biasing it.  And since he tested them separately with a DC load, even if the current limiting was bogus or he hooked them up backwards, the current was externally limited and shouldn't have blown the wires off.

Bogus (meaning they don't work to spec) LM317 series regulators are legion these days so the only suprise here is that you can get them from Digi-Key directly instead through their Marketplace portal where most of the fake stuff is. 

[magic]

One more thing I could imagine is the positive rail going negative if the negative rail came up first and then overvoltage, breakdown and overcurrent.

But realistically, the wimpy wires are the most likely culprit indeed. And of course, if their datashit is to be believed, the chip is capable of sourcing more current than it can survive

But again, realistically, the wires fuse at 0.5A, the chip smokes at 1A, the spec says 1.5A and somebody inevitably sells it as a 5A part on eBay.

[floobydust]

One more thing I could imagine is the positive rail going negative if the negative rail came up first and then overvoltage, breakdown and overcurrent.

But realistically, the wimpy wires are the most likely culprit indeed. And of course, if their datashit is to be believed, the chip is capable of sourcing more current than it can survive

But again, realistically, the wires fuse at 0.5A, the chip smokes at 1A, the spec says 1.5A and somebody inevitably sells it as a 5A part on eBay.

'317 does not have a ground pin, if the IC fails there is no overcurrent other than it what it can push into the load by overvoltaging it. The ADJ pin resistor limits that.
But... in a bench power supply the ADJ pin potentiometer smokes (add a polyfuse) when you dial it down wondering why the output voltage is too high. Then, when you short the output, the IC finally melts.


EVVO Semiconductor screwed up twice - defective packaging, and defective test. No biggie in china, this is par for the course.
Do they have an independently-audited quality system in place?
Shame on Digi-Key for letting anyone into their supply. For all that markup of theirs, they have no vetting process it appears.
LM317K TI is $1.00 on LCSC verses $6.61 from EVVO on DK.

[wraper]

LM317K TI is $1.00 on LCSC verses $6.61 from EVVO on DK.

Those are not TO-3 317K but plastic packages at LCSC.

[floobydust]

OK I must be too hasty, in my day the "K" was TO-3, and "T" suffix TO-220 but it means nothing now. I'd found it as a TI "TO-3-2" I think but now the listings price is up there past $40 ea.
Aliexpress N/S LM317K TO-3 $1.52 for two or $3.79 for five. Ask for lucky one. Well, it's not the steel costs I guess.
/s You could just buy some TO-3's and cut off the top to do your own QA. Then tack solder a real TO-220 or D2PAK in there, and have something better. It's that ridiculous.

[bdunham7]

You could just buy some TO-3's and cut off the top to do your own QA. Then tack solder a real TO-220 or D2PAK in there, and have something better. It's that ridiculous.

Perhaps an improved version of this? 

https://www.ebay.com/itm/266761106914

[floobydust]

Looks reasonable but a bit expensive. In the old days it was common to see a TO-3 voltage reg on a small tiny heatsink, so a few watts at best dissipation, which by today's standards was a huge waste of metal.
I'm just postulating where the markup truly is for these parts. Not the TO-3 can, not the die. You have to wonder.

Many "semiconductor" vendors are buying the dies and just do their own packaging of them. That is their core business.

[wraper]

You could just buy some TO-3's and cut off the top to do your own QA. Then tack solder a real TO-220 or D2PAK in there, and have something better. It's that ridiculous.

Perhaps an improved version of this? 

https://www.ebay.com/itm/266761106914

You'd be way better off with TO-220 package + 2 wires. Thermal conductivity is simply not there as it's just a usual 2 layer PCB with vias. So unless component is lightly loaded, this thing is pretty much worthless.

[floobydust]

I remember the TO-220 form factor sizing was to replace the TO-66. You can bolt one in place there but can't remember trying it.

[bdunham7]

You'd be way better off with TO-220 package + 2 wires. Thermal conductivity is simply not there as it's just a usual 2 layer PCB with vias. So unless component is lightly loaded, this thing is pretty much worthless.

That was why I suggested an improved version.  Perhaps a machined copper insert and low-temp solder for the center pad?  IDK how good you could get it with 4-layer 2oz copper and vias everywhere possible.  I've done TO-220 retrofits and I've never been happy with the way the look--often TO-3 devices are filling holes in a chassis. 

[floobydust]

With the TO-3 (TO-204) obsolete it will become more of an issue for replacement.
The automotive industry is using thermal adhesive under the PC board, that also flows into the vias, to transfer heat from semi's to the die-cast enclosure in ECU's. It's proven to work but I'm not a glue fan really.
I thought the LM317 is not high power, dissipation max. 20W compared to power transistors in the same package.

[wraper]

You'd be way better off with TO-220 package + 2 wires. Thermal conductivity is simply not there as it's just a usual 2 layer PCB with vias. So unless component is lightly loaded, this thing is pretty much worthless.

That was why I suggested an improved version.  Perhaps a machined copper insert and low-temp solder for the center pad?  IDK how good you could get it with 4-layer 2oz copper and vias everywhere possible.  I've done TO-220 retrofits and I've never been happy with the way the look--often TO-3 devices are filling holes in a chassis.

I already suggested PCB of half TO-3 footprint + TO-220. No temperature compromise. But it's just unneeded bells and whistles. You can nicely fit TO-220 screwed to one of the holes + 2 wires soldered directly to IC terminals as their order is the same as on TO3 if fit like on the sketch I posted.

[David Aurora]

But I'm curious what exactly went wrong and not accepting hand wavy non-explanations for this reason.

You're "not accepting" certain responses? You really are an authority of some sort in your own mind aren't you?


P.S. Enough with the post edits after replies, I keep having to go back to shitposts of yours I've already replied to after you change them

...
One more thing I could imagine is the positive rail going negative if the negative rail came up first and then overvoltage, breakdown and overcurrent.
...

You've shown that you can imagine a lot of things, regardless of their basis in reality.

...
But realistically, the wimpy wires are the most likely culprit indeed.
...

OK cool, see you again in your next post where we go back to the conspiracy theories

[thm_w]

Shorted the output, raised current to 2A, didn't even see 1.5A on the display and the bond wire had blown open. Vin was 6V, and happened so fast, so no issue with die temperature.

I don't have proper calipers here but wire is ~0.04mm.
Managed to get it to glow red at only 1A with careful probing. So aluminum could make sense.

Bonding looks rough too:

Same batch/date code (mine were all 2437)?

Yes same batch.
I'll ask digikey for refund.

[David Aurora]

Shorted the output, raised current to 2A, didn't even see 1.5A on the display and the bond wire had blown open. Vin was 6V, and happened so fast, so no issue with die temperature.

I don't have proper calipers here but wire is ~0.04mm.
Managed to get it to glow red at only 1A with careful probing. So aluminum could make sense.

Bonding looks rough too:
(Attachment Link)

Same batch/date code (mine were all 2437)?

Yes same batch.
I'll ask digikey for refund.

Thanks for putting in the effort to confirm!

[magic]

Shorted the output, raised current to 2A, didn't even see 1.5A on the display and the bond wire had blown open.

OK, but can you be certain that 1.5A was not actually reached, even transiently?
Remember, this is the absolute maximum rating of the part you bought, and internal current limit is suggested to be higher and certainly not guaranteed to be lower.

If I were DigiKey I would reject your refund

[thm_w]

OK, but can you be certain that 1.5A was not actually reached, even transiently?
Remember, this is the absolute maximum rating of the part you bought, and internal current limit is suggested to be higher and certainly not guaranteed to be lower.

Read the post again please, and count the number of bond wires seen in the photo.
1 for Vin, 1 for Adj, 2 for Vout

[magic]

Yes, Vout is Kelvin sensed.

OK, you got it to glow red at 1A, fair enough I guess. Not sure how I missed it first time.

[David Aurora]

...
Not sure how I missed it first time.
...

Maybe because you've spent 5 pages shit stirring and the signal to noise ratio in this thread at this point is completely cooked?

Yeah. Might be that I reckon.

[magic]

5 pages

I think you have confused me with 10 quadrillion other people who also doubted your competence, but appear to be convinced by thm_w's post at last, for some reason.

[bdunham7]

I don't have proper calipers here but wire is ~0.04mm.
Managed to get it to glow red at only 1A with careful probing. So aluminum could make sense.

Aluminum melts before it glows.

0.04mm is ~1.5mils, that would be good enough for about 0.5A in free air if it were gold.

https://www.renesas.com/en/document/atc/power-systems-design-estimating-bond-wire-current-carrying-capacity?srsltid=AfmBOoqrX5L5AJugpkSg_Ak_EBoo_f-i2Y7UH1AhLFXev6J48zieFkpR

[magic]

Aluminium melts above 600°C and in darkness you should be able to see a faint glow at this kind of temperature.

[David Aurora]

5 pages

I think you have confused me with 10 quadrillion other people who also doubted your competence, but appear to be convinced by thm_w's post at last, for some reason.

Same end result whether or not you and a couple of your buddies doubted my competence- I was right. Maybe you should spend less time mis-judging other peoples competence and more time reflecting on why you feel the need to do that.

[thm_w]

Its definitely silver color, so silver or aluminum. If you are saying aluminum doesn't glow then it would have to be silver.
https://tanaka-preciousmetals.com/en/products/detail/bonding-wires/

There isn't a ton of space on the die for larger wire. But maybe slightly thicker gold wires, and shorter length would have been OK.

[wraper]

Its definitely silver color, so silver or aluminum. If you are saying aluminum doesn't glow then it would have to be silver.
https://tanaka-preciousmetals.com/en/products/detail/bonding-wires/

There isn't a ton of space on the die for larger wire. But maybe slightly thicker gold wires, and shorter length would have been OK.

There is a space for thicker wire and unused in/out pads for a second wire in parallel. Silicon die looks like pretty normal LM317 die, maybe a little smaller than others shown but still of reasonable size.

[bdunham7]

Its definitely silver color, so silver or aluminum. If you are saying aluminum doesn't glow then it would have to be silver.
https://tanaka-preciousmetals.com/en/products/detail/bonding-wires/

There isn't a ton of space on the die for larger wire. But maybe slightly thicker gold wires, and shorter length would have been OK.

From that website apparently palladium-coated copper is a low-cost alternative to gold as well. 

There's room for more wires, right?  And it looks to me like you could double the diameter of the wires as well, although I'm no bonding expert.

[floobydust]

There is a space for thicker wire and unused in/out pads for a second wire in parallel. Silicon die looks like pretty normal LM317 die, maybe a little smaller than others shown but still of reasonable size.

It would interesting to measure and compare the die sizes.
To me it looks like the usual aggressive die shrink- resulting in the bonding-wire pads being tiny. Problem is you do that over many generations and hit a limit.
Compared to the pics here of other makes, there is no room for thicker wire. That means you'd have to use a second wire (input) and I see the die has provision for that but it would be hairy routing because the second (output) wire is in the way.
I'm guessing this is the 1.5A version die, not usable for the lesser current 317 versions if the sense resistor can be fuse modified.

[Siwastaja]

Same end result whether or not you and a couple of your buddies doubted my competence- I was right. Maybe you should spend less time mis-judging other peoples competence and more time reflecting on why you feel the need to do that.

You definitely were right.

I'm still enjoying my mental image of you contacting Digikey and their engineer as a very first thing asking for a schematic and parameters of your test circuit. Are you going to continue your one-man war and generate 6 pages of worthless email, or, do you suck it up, accept your loss, and use 5 minutes to draw that schematic, which you could have done here in the first place if your ego allowed you to operate like a normal sensible engineer?

Anyway, great job. I actually enjoy reading troll threads. And you definitely were right, just like a high-quality troll should be certain about before they start, so that they can do that nice ending reveal. Incompetent you are not - just an asshole who enjoys social games over being clear and concise.

[floobydust]

Same end result whether or not you and a couple of your buddies doubted my competence- I was right. Maybe you should spend less time mis-judging other peoples competence and more time reflecting on why you feel the need to do that.

You definitely were right.

I'm still enjoying my mental image of you contacting Digikey and their engineer as a very first thing asking for a schematic and parameters of your test circuit. Are you going to continue your one-man war and generate 6 pages of worthless email, or, do you suck it up, accept your loss, and use 5 minutes to draw that schematic, which you could have done here in the first place if your ego allowed you to operate like a normal sensible engineer?

Anyway, great job. I actually enjoy reading troll threads. And you definitely were right, just like a high-quality troll should be certain about before they start, so that they can do that nice ending reveal. Incompetent you are not - just an asshole who enjoys social games over being clear and concise.

To blame the application circuit over a melted bonding wire is totally hilarious.
I think you enjoy being the asshole here. Your trolling is fine if you are correct as the know it all.
But you blow up and personal attack too quickly to work out and hear any discussion. This makes the forum a hostile place and derails threads. PFO.

[wraper]

To blame the application circuit over a melted bonding wire is totally hilarious.

First 1/3 of this thread was basically: "this IC is garbage, trust me bro", and irritation when asked for more info. Nobody knew what was actual failure mode and conditions when it failed.

[floobydust]

Is this blaming the victim? New IC fails -> gimme full schematic -> no -> flames -> you are to blame -> flames

More worriesome is that we can't have a reasonable discussion on eevblog without the insults and personal attack.

[magic]

You got it wrong.

IC fails -> gimme basic details -> no -> how can we know that you are not to blame -> flames

Go back to page 1 and see for yourself.

[Analog Kid]

From my vantage point, the two bad guys here most responsible for muddying the waters are @magic and @wraper.

The OP has done everything that's required of someone making a report on defective parts. You nagging ankle-biters have no case.

More popcorn!

[wraper]

The OP has done everything that's required of someone making a report on defective parts. You nagging ankle-biters have no case.

More popcorn!

Read first 2 pages, zero info other than: "this IC is trash, other ICs are good, I've done everything correctly". Asking for more info: "how dare you question me!" You may notice that when OP provided some info about the part, conversation suddenly became constructive.

[magic]

Aluminum melts before it glows.

Aluminium definitely glows before it melts.


Sorry for the noise, this had to be investigated

[Analog Kid]

Aluminum melts before it glows.

Aluminium definitely glows before it melts.
(Attachment Link)

Sorry for the noise, this had to be investigated

Or maybe it glows as it melts?
How many angels can dance on the head of a pin?
Sheesh.

[magic]

Or maybe somebody is trolling due to having nothing meaningful to say.

[Analog Kid]

OK, so maybe it glows before it melts. (At least in some cases.)
Yay for you.

[bdunham7]

Aluminium definitely glows before it melts.

Sorry for the noise, this had to be investigated

Hmmm.  My statement was based on seeing (many times) molten aluminum being poured and not seeing it glowing.  I'd never tried heating foil with current.  I've no idea why there would be a difference, that looks almost white hot.

Edit:  I was unable to replicate your result with standard kitchen aluminum foil, it would repeatedly fuse before glowing.  Any specifics on how you did that?  Current, foil source, size, etc? 

[magic]

Narrow (maybe 1mm or so) strip of kitchen foil mounted to a transformer soldering iron. I guess the current has to be several amps because I tried shorting out a NiMH cell with a similar strip and it didn't glow.

Admittedly, my iron has built-in power adjustment (triac dimmer?) and I had to tweak it a little to get it to glow continuously without fusing. BTW, for this kind of stuff constant voltage is better than constant current, because metals typically have positive TCR so the former decreases power dissipation at high temperature while the latter increases it.

And it's not glowing white, that's just limited dynamic range of the camera. It really is orange to the naked eye.

Not sure what happens here, but if it melts, it stays in its original shape surprisingly well. At any rate, I think we can safely say that aluminium conductor glows before it fuses

[floobydust]

There are many research papers on aluminum bonding wires. Would be good to know how thick the wire is EVVO used. I also note the ends (toes) are huge and I wondered if they are touching any features?
neat failure analysis: https://www.semlab.com/examples/deviceexamples/
They call it "EOS electrical overstress" when fusing happens mid-span due to heatsinking at the ends. But we know... there was no overstress aside from forum members fusing 

Interesting is a tiny amount of silicon is added to the aluminum, then a double annealing process. IF this is done wrong, you get cracking in the wire - which might be something here, that the bonding wire itself failed due to cracking.
Personal experience with chinese metallurgy is they have no quality control, the country is quite used to things failing and their solution is to simply replace the product when you point out the annealing or heat treating is no good. No biggie 

Graph from Heraeus Electronics bonding wire

[Analog Kid]

Wouldn't it be cheaper (and better!) in the long run just to use copper for bond wires?
I mean, how expen$ive could that be? Even on a run of 100,000 pieces, the cost of the wire has got to be totally marginal.
Especially if you factor in the cost of all those other treatments, the silicon additive and the annealing.

[wraper]

Wouldn't it be cheaper (and better!) in the long run just to use copper for bond wires?
I mean, how expen$ive could that be? Even on a run of 100,000 pieces, the cost of the wire has got to be totally marginal.
Especially if you factor in the cost of all those other treatments, the silicon additive and the annealing.

You forget it's not all that simple. You cannot just ultrasonic bond anything to anything and expect it to be reliable. Not to say it seems copper bond wire is a recent thing, like 15 years of real mass production. Nobody will update old designs with that. Also material as such is not even an issue here. It's inadequately thin wire, and running only one when they could run two thicker wires at barely any additional cost. Another problem with using single wires is that BJT is likely not evenly loaded because current needs to run over the die from the opposite side.

[floobydust]

The boss needs to buy a Lamborghini. Think of all the profits EVVO reaps using cheapola bonding wire raking in the millipennies /s
Someone reviewed the bonding wire spec enough to use super thin for the ADJ pin, so they screwed up for that one high current wire. Oh and no testing done either.
Noopy die shot of 317H 0.5A TO-39 here's how it's done it's got the beef lol.

But seriously, I have never seen BoM costs for something like this IC.
I believe most of the costs are the die, but it seems to have been overshrunk really there is not enough room for bonding wires and where is that heat spreader anyway.

[wraper]

^Bond pad area is utilized by 50% at best. Also do not forget that 2x cross section would require just 1.4x increase in wire diameter. Not to say half of the bond pads were left unused. Regardless of the die quality, this problem has nothing to with the die but with inadequate packaging.

[floobydust]

The second failure is in testing, or lack thereof- that should have caught this packaging error.
They purport to have an ISO9001 quality system in place and clearly that failed as well, third failure.
I would have no confidence in the dies then, how are they for line/load regulation, noise, current-limit etc. ? LM337 would have the same problems.

[wraper]

IMHO they did nothing at all. Considering it's the same die/packaging as that fake LM338k I mentioned, just ordered the part from whoever sells this garbage with their brand name printed. I doubt it was an accidental error.

[exe]

I agree some numbers would support a more healthy discussion. I have no doubts that the parts are bad, but it took a lot of posts to get at least some understand how exactly they bad.

I guess we came here for an engineering discussion, but for OP it was a way to vent out. Which he couldn't do it because people are more interested in technical details.

[thm_w]

Sort of, OP was trying to warn others. The chance of anyone being on here buying LM317K is very low but its still good to see, as now I think we can consider the whole EVVO product line as suspect. And should tell us something about digikeys trust in manufacturers.

I sent an email to digikey to remind them to consider delisting the part from the site. I don't think they even look at the refund request reason unless the refund is >$50.

[exe]

BTW, what I found curious is that many folks expected digikey to provide only quality parts. Interesting. I didn't have that expectation, I only expect them to sell genuine parts.

I do buy parts from unknown to me Chinese brands from lcsc, but I test them. I also have no trust in parts that come with fake datasheets (copy-pasted from other datasheets, replaced logos, or too short datasheet omitting essential details). And there is always that dilemma: pay X times less for an unknown part, or pay a premium for brands and parts like AD/LT/TI/etc.

[David Aurora]

Same end result whether or not you and a couple of your buddies doubted my competence- I was right. Maybe you should spend less time mis-judging other peoples competence and more time reflecting on why you feel the need to do that.

You definitely were right.

I'm still enjoying my mental image of you contacting Digikey and their engineer as a very first thing asking for a schematic and parameters of your test circuit. Are you going to continue your one-man war and generate 6 pages of worthless email, or, do you suck it up, accept your loss, and use 5 minutes to draw that schematic, which you could have done here in the first place if your ego allowed you to operate like a normal sensible engineer?

Anyway, great job. I actually enjoy reading troll threads. And you definitely were right, just like a high-quality troll should be certain about before they start, so that they can do that nice ending reveal. Incompetent you are not - just an asshole who enjoys social games over being clear and concise.

How do you still not get this?

Digikey got all details right off the bat because they're the supplier and therefore relevant to the situation. After my initial email I actually phoned them with more information from my tests.

You didn't get the information you want because you're some random nobody on the internet whose opinion on the matter is of no interest to me. It was a take it or leave it warning, I didn't think anyone out there had so little going on in their day that they'd actually pick an argument over it. Do you people not have actual lives or jobs?

[David Aurora]

The OP has done everything that's required of someone making a report on defective parts. You nagging ankle-biters have no case.

More popcorn!

Read first 2 pages, zero info other than: "this IC is trash, other ICs are good, I've done everything correctly". Asking for more info: "how dare you question me!" You may notice that when OP provided some info about the part, conversation suddenly became constructive.

As I said to that other dick- Digikey and the manufacturer got all the info straight away (and follow ups too). Because their input on the matter is actually important. Yours is not.

I don't know why your ego is so massive that you think everyone has to run their tests and findings past you before coming to a decision, but that's your insecurity issue to deal with. At the end of the day, I put out a warning in the hope of saving others time and headaches. It was intended as a gesture of kindness towards the electronics community, but I suspect you don't have the basic social skills to comprehend that.

[David Aurora]

Sort of, OP was trying to warn others. The chance of anyone being on here buying LM317K is very low but its still good to see, as now I think we can consider the whole EVVO product line as suspect. And should tell us something about digikeys trust in manufacturers.

I sent an email to digikey to remind them to consider delisting the part from the site. I don't think they even look at the refund request reason unless the refund is >$50.

It's quite concerning to me that they're still up there at this point, I think I'll follow up with them again too

[magic]

There is nothing wrong with those chips, I'm sure there are plenty of happy customers using them at 0.1A or so.
You guys are weird

[gnif]

The OP has done everything that's required of someone making a report on defective parts. You nagging ankle-biters have no case.

More popcorn!

Read first 2 pages, zero info other than: "this IC is trash, other ICs are good, I've done everything correctly". Asking for more info: "how dare you question me!" You may notice that when OP provided some info about the part, conversation suddenly became constructive.

As I said to that other dick- Digikey and the manufacturer got all the info straight away (and follow ups too). Because their input on the matter is actually important. Yours is not.

I don't know why your ego is so massive that you think everyone has to run their tests and findings past you before coming to a decision, but that's your insecurity issue to deal with. At the end of the day, I put out a warning in the hope of saving others time and headaches. It was intended as a gesture of kindness towards the electronics community, but I suspect you don't have the basic social skills to comprehend that.

https://www.eevblog.com/forum/chat/forum-rules-please-read/
Take this as a warning, there is no need to name call or insult other members here.

[exe]

btw, I do think that it would be beneficial for the community if digikey removed it from the listing. But, I dealt with their support once when they sent me someone else's parts along with my order (some logic ICs). I was expected them to at least say thanks to me and ask for details, such us the order number on the package so they could notify the original owner of the parts. But all they said was that I can return parts if I want. Since those parts where cheap, and sending them back would cost more than the price of the parts, I didn't bother.

[squadchannel]

btw, I do think that it would be beneficial for the community if digikey removed it from the listing. But, I dealt with their support once when they sent me someone else's parts along with my order (some logic ICs). I was expected them to at least say thanks to me and ask for details, such us the order number on the package so they could notify the original owner of the parts. But all they said was that I can return parts if I want. Since those parts where cheap, and sending them back would cost more than the price of the parts, I didn't bother.

I've been there too. A couple of times in the past I have received the wrong item.
Each time, I contacted digikey and told me, "No need to return it, I'll send you a new one." I feel lucky.

The most expensive wrong receieved item was a TDK-Lambda SMPS, 50W, which is expensive to buy new.

[thm_w]

BTW, what I found curious is that many folks expected digikey to provide only quality parts. Interesting. I didn't have that expectation, I only expect them to sell genuine parts.

Its expected, and should be.
If I go to digikey and order a transistor that is rated for 1A, it sure as hell better handle 1A of current.
Same if I go to mcmaster and order a bolt rated for 1,000kg.

[David Aurora]

BTW, what I found curious is that many folks expected digikey to provide only quality parts. Interesting. I didn't have that expectation, I only expect them to sell genuine parts.

Its expected, and should be.
If I go to digikey and order a transistor that is rated for 1A, it sure as hell better handle 1A of current.
Same if I go to mcmaster and order a bolt rated for 1,000kg.

Exactly. The "quality" vs "genuine" thing is absolute hair splitting. It shouldn't be remotely controversial to expect an LM317 from Digikey to work like an LM317.

[David Aurora]

The OP has done everything that's required of someone making a report on defective parts. You nagging ankle-biters have no case.

More popcorn!

Read first 2 pages, zero info other than: "this IC is trash, other ICs are good, I've done everything correctly". Asking for more info: "how dare you question me!" You may notice that when OP provided some info about the part, conversation suddenly became constructive.

As I said to that other dick- Digikey and the manufacturer got all the info straight away (and follow ups too). Because their input on the matter is actually important. Yours is not.

I don't know why your ego is so massive that you think everyone has to run their tests and findings past you before coming to a decision, but that's your insecurity issue to deal with. At the end of the day, I put out a warning in the hope of saving others time and headaches. It was intended as a gesture of kindness towards the electronics community, but I suspect you don't have the basic social skills to comprehend that.

https://www.eevblog.com/forum/chat/forum-rules-please-read/
Take this as a warning, there is no need to name call or insult other members here.

And it was not insulting to be called incompetent or a liar for 6 pages?

[gnif]

The OP has done everything that's required of someone making a report on defective parts. You nagging ankle-biters have no case.

More popcorn!

Read first 2 pages, zero info other than: "this IC is trash, other ICs are good, I've done everything correctly". Asking for more info: "how dare you question me!" You may notice that when OP provided some info about the part, conversation suddenly became constructive.

As I said to that other dick- Digikey and the manufacturer got all the info straight away (and follow ups too). Because their input on the matter is actually important. Yours is not.

I don't know why your ego is so massive that you think everyone has to run their tests and findings past you before coming to a decision, but that's your insecurity issue to deal with. At the end of the day, I put out a warning in the hope of saving others time and headaches. It was intended as a gesture of kindness towards the electronics community, but I suspect you don't have the basic social skills to comprehend that.

https://www.eevblog.com/forum/chat/forum-rules-please-read/
Take this as a warning, there is no need to name call or insult other members here.

And it was not insulting to be called incompetent or a liar for 6 pages?

You resorted to name calling, thus I replied to your message. This warning/reminder is for all here, not just you.

[wraper]

And it was not insulting to be called incompetent or a liar for 6 pages?

Nobody called you incompetent or a liar, if you extrapolate what was actually said to that, it's a non existing problem you invented in your brain. Since when raising doubts when conversation partner refuses to answer a few most basic technical questions, and not immediately accepting everything on face value (on engineering forum at that), became an insult? Actually I don't recall anyone except you giving any insults or personal attacks.

[AVGresponding]

And it was not insulting to be called incompetent or a liar for 6 pages?

Flamebaiting is flamebaiting; you fell into the troll's trap. Mods/admins here are well known for not bothering with context, as they mostly don't have the time and just respond directly to an individually reported post and not the backstory.

Kind of ridiculous that this thread is still going, all the relevent information was imparted several pages ago...

[gnif]

And it was not insulting to be called incompetent or a liar for 6 pages?

Flamebaiting is flamebaiting; you fell into the troll's trap. Mods/admins here are well known for not bothering with context, as they mostly don't have the time and just respond directly to an individually reported post and not the backstory.

Kind of ridiculous that this thread is still going, all the relevent information was imparted several pages ago...

Sorry, but I had time, I read through and saw that David Aurora was the only one that actually crossed the line and started to name call, clearly out of frustration. Normally personal insults result in a ban, but in this instance it was clear that things here got a bit heated which resulted in this instead. The post was a warning/reminder to all, to settle down. While I can't speak for the other mods here, I always try to get the context before I act.

I do agree though that this thread has run it's course, as such, locking.