EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: Macbeth on August 26, 2015, 11:40:52 am
-
I recently purchased some LTZ1000's from AliExpress. I was fully expecting used/remanufactured, even faulty devices for the low price, but thought fakes are probably a myth. I mean, it's a pretty niche low market device. Who would bother faking them really?
Now the first thing I did was do some ohms / diode tests and things don't appear to be in order. e.g. the heater is listed in the datasheet as ~300 ohms. Mine are both open circuit H+ to H-, and reading 1.4v (2 x diodes) the other way.
Doing some EBC diode tests on the NPNs, Q1 behaves as expected but Q2 measures 1.4v b-e and 2.1v b-c (with 0.9v in reverse instead of o/c).
Or am I reading too much into the simple internal diagram on the datasheet? I've also noticed the HP3458A 7V reference diagram is more detailed on the internals, but that doesn't match up with my measurements either.
Has anyone with a genuine LTZ1000 got ohms measurements of the heater (pins 1,2) and diode tests of Q1 EBC (pins 7,4,5) and Q2 EBC (pins 7,6,8) to share?
I will get some pics up later. The devices have a date code of T0612 and do not have gold plated pins.
-
(http://i.imgur.com/d5de5jm.png)
(http://i.imgur.com/di8uy2b.png)
(http://i.imgur.com/EXKF4XD.png)
-
Looks 100% fake to me. Why?
Marking too big and looking cheap low-resolution paint silkscreen,
No lot code.
Pins are too big, LTZ's I had/saw have more thin pins.
Key facing wrong direction, usually it's facing up, if looking on LTZ labeling straight.
Bottom side is gold-plated on original LTZ's
Crack one open, and let's see what's inside :)
Why to fake LTZ's - well, it's niche market, but at 50$/a part still some voltnuts buying them in ten's, so even with half-price deal - do the math :)
-
I also think, that these are fake ones.
Same reasons, as TiN already mentioned, especially the printing is really bad, and the signet on the edge of the case.
The heater should read these 200..400 Ohm in one direction, although in the wrong direction, you really might measure the bulk diode. But that should read 0.55V, then.
The case normally is also gold coloured on the bottom, and I've never seen original parts with silver legs..
But might be reworked.
Last point, I found some pictures, origin is a facsimile archive from the volt-nuts list:
http://volt-nuts.febo.narkive.com/FIfiYHU5/ltz1000ach-and-ltz1000ch-in-china (http://volt-nuts.febo.narkive.com/FIfiYHU5/ltz1000ach-and-ltz1000ch-in-china)
These fake ones there have exactly the same DC as yours.
Sorry - Frank
PS: original volt-nuts thread: https://www.febo.com/pipermail/volt-nuts/2011-August/000961.html (https://www.febo.com/pipermail/volt-nuts/2011-August/000961.html)
(not faked O0 )
-
Thanks guys. Looking at the block diagrams the heater should measure a few hundred ohms both ways around, but I can see from the last diagram if it went o/c then there is a diode in parallel which could be measured instead (but 1.4v?). Anyway I tabulated diode tests on all leads and both devices are consistent with each other and bear no relation to the block diagrams. I tested my meter didn't put out 7V as I don't want to be accused of "zenering" them by the seller (like they would know what that is!) :-DD BTW. These only cost me £7 each so I knew they would be on the shitty side, maybe one a failure, but not an outright relabelled part.
(http://i.imgur.com/htUitD8.png)
I'll have a play and see if the supposed zener actually works. If this was the real deal would it be safe enough to just apply a PSU with a limiting resistor across 3 and 4 and measure the voltage? The datasheet mentions a certain order of pins being higher voltages than each other.
I will go and find my dremel and cut the lid off one of them and try and take some pics after I've done that ;)
-
Looks 100% fake to me. Why?
Marking too big and looking cheap low-resolution paint silkscreen,
No lot code.
Pins are too big, LTZ's I had/saw have more thin pins.
Key facing wrong direction, usually it's facing up, if looking on LTZ labeling straight.
Bottom side is gold-plated on original LTZ's
Crack one open, and let's see what's inside :)
Why to fake LTZ's - well, it's niche market, but at 50$/a part still some voltnuts buying them in ten's, so even with half-price deal - do the math :)
I think just the color of the pins should be enough to call these fake.
So refund I guess.
-
It's supposed to be a voltage reference, not a resistance reference. Why not measure the reference voltage output and critique that ? Sure, you do say you went looking for a dodgy part, and may have well found one, but don't rule out the possibility it may well work. I'm not saying just because they found a newer cheaper way to do the same thing that they should label it the old way instead of advertise it as the substitute, but what about the elephant in the room, are we going to get 50 pages discussing the resistance of a voltage reference and 50 pages in another thread discussing voltage drop across 1% resistors ? give us the beef ! I'm not asking for a graph with voltage across temperature range ( I'm just subliminally suggesting it :o ) but tell us the meter reading for voltage please !
If it works fine AS A VOLTAGE REFERENCE which is like, the idea, then hey, it's a way to put fifty bucks into your cosplay fetish rather than EE fetish. That would have to be a good thing right ?
....right ?
anyone ?
-
If it works fine AS A VOLTAGE REFERENCE which is like, the idea, then hey, it's a way to put fifty bucks into your cosplay fetish rather than EE fetish. That would have to be a good thing right ?
....right ?
anyone ?
You have no idea what you are talking about, do you?
I'll make a car analogy for you, because I was told those are easier to understand than electronics ( I disagree but whatever)
It's like ordering the best Ferrari ever made and find that it is something else with some spoilers painted red. It doesnt matter how fast that fake car goes, what brand it is, it is not a Ferrari, end of story.
-
....troll talk...
It's a temperature controlled voltage reference. The heater should work at a minimum.
-
....right ?
:palm:
-
MacBeth,
I have some brand new LTZ1000ACH's A Grade parts straight from Linear, They are in my inventory as project spares. Would it be of any help if I take pictures, and make measurements of these? They are still in the sealed anti-stat bags as we speak.
Jon
-
I know what it does and have made ovens before to do the whole ntc ptc thing. Grow up people. It's like measuring the resistance of your room heater. It's irrelevant to whether it does the job or not. Measure if it keeps you warm or not. In this case, you're trying to get voltage out. Unless your circuit specifically looks for a set resistance to regulate the temperature, then who cares ?
-
You have no idea what you are talking about, do you?
I'll make a car analogy for you, because I was told those are easier to understand than electronics ( I disagree but whatever)
It's like ordering the best Ferrari ever made and find that it is something else with some spoilers painted red. It doesnt matter how fast that fake car goes, what brand it is, it is not a Ferrari, end of story.
worst analogy ever.
It is blindly clear that people should read up about what these things are and what these things do. The heater resistance is IRRELEVANT. You need a thermostat circuit to control the heater. It's governed by the temperature as sensed by the temperature sensor in the package, if you design the circuit so that the temperature is governed by the resistance that's and EPIC FAIL. The point is the temperature and controlling it dynamically. NOT STATICALLY. It would defeat the purpose.
(https://www.eevblog.com/forum/projects/ltz1000-heater-resistance-possible-fakes/?action=dlattach;attach=167867;image)
-
MacBeth,
I have some brand new LTZ1000ACH's A Grade parts straight from Linear, They are in my inventory as project spares. Would it be of any help if I take pictures, and make measurements of these? They are still in the sealed anti-stat bags as we speak.
Jon
If it's not much effort, that would be great :-+ Though don't break the seal if you want to resell them or something like that on my account!
-
The only possible reason for measuring the resistance of the heater or calling for it to be a certain value is if you are going to MAKE A CIRCUIT THAT WILL NOT WORK.
If the circuit has a thermostat to control the temperature of the reference, which is the POINT of buying the reference then the resistance doesn't matter, it only matters that it has a heater. You CAN, and many people DO make their own by putting a resistor on the side of a reference or a diode they want to use as a reference and then wrapping it in insulation. You put the thermostat, the heater and the reference together and make a DYNAMIC temperature control.
A set resistance is only good if you're putting the heater across a set voltage for a set output, which means it cannot compensate for ambient temperature changes. That in turn makes the whole exercise pointless because the voltage will drift.
-
Here is a temperature controlled crystal oscillator. This method of doing it is older than I am.
heater, sensor, crystal.
(https://hackadaycom.files.wordpress.com/2014/07/oven.jpg?w=800)
from
https://hackaday.com/tag/oven-controlled-crystal-oscillator/ (https://hackaday.com/tag/oven-controlled-crystal-oscillator/)
-
The heater resistance is IRRELEVANT. You need a thermostat circuit to control the heater. It's governed by the temperature as sensed by the temperature sensor in the package, if you design the circuit so that the temperature is governed by the resistance that's and EPIC FAIL. The point is the temperature and controlling it dynamically. NOT STATICALLY. It would defeat the purpose.
And what this have to do with determining is it genuine or not? You really need to write useless shit in every tread :palm:. Does the increasing post count make you happier or what?
-
I don't know what the HeadlessChickens beef is. Read the LTZ datasheet and it states the heater is 200 min and 420 max ohms. I measure open circuit.
-
Missing the point people. You're trying to critique an item which you knew would probably be substandard. Then totally failing to show it is substandard by not measuring it's one and only function. REVIEW EPIC FAIL.
A decent review would graph it's voltage to temperature ratio and compare that to the graph in the datasheet.
A crappy review would at least measure the voltage at the ambient temperature at the time of writing.
This review didn't even measure the voltage of a voltage reference. Epic fail.
Do you know how to measure voltage ?
-
oh, but I'll give the review a C- for name-calling. Unprofessional.
-
Missing the point people. You're trying to critique an item which you knew would probably be substandard. Then totally failing to show it is substandard by not measuring it's one and only function. REVIEW EPIC FAIL.
No, rather you are missing your brain and this is not a review :palm:.
-
I recently purchased some LTZ1000's from AliExpress. I was fully expecting used/remanufactured, even faulty devices for the low price, but thought fakes are probably a myth.
He is not determining if these knock-off versions will work, he is trying to determine if they are genuine or not. They don't appear to be, but by measuring the resistance of the heater and it not matching the datasheet it suggests that they are damaged or fakes.
Missing the point people.
No you are!
You're trying to critique an item which you knew would probably be substandard. Then totally failing to show it is substandard by not measuring it's one and only function. REVIEW EPIC FAIL.
He's not trying to critique the item. I believe the "EPIC FAIL" is yours and your inability to read the original post.
Anywho, back to the topic - just from the print on the part I'd say they are fakes - all linear products tend to have really good quality printing on them.
Were you trying to find a cheap source for them, or were you after a couple to test with? If it is the latter, linear have sent me out a few of the lm399 as a sample - I'd assume they'd do the same for the ltz1000.
-
I recently purchased some LTZ1000's from AliExpress. I was fully expecting used/remanufactured, even faulty devices for the low price, but thought fakes are probably a myth.
He is not determining if these knock-off versions will work, he is trying to determine if they are genuine or not. They don't appear to be, but by measuring the resistance of the heater and it not matching the datasheet it suggests that they are damaged or fakes.
agreed. Speculating about whether they're fake or not based on what, the datasheet ? The genuine part may have had several different production runs with several different silkscreens. Where is the information on determining if it is real or not ? Heck, genuine parts are often out of spec, what does that prove, and at the end of the day, with no information to judge if it genuine or not, what is the point ? I'm thinking if the voltage output is not measured then nobody knows whether it is even worth keeping an eye out for these supposed fakes, which may be genuine without guidance from the original manufacturer it's hard to know. I want the half hour I've spent on this back.
Missing the point people.
No you are!
No you are, times two ! ( seriously, are we doing this ????? )
He's not trying to critique the item. I believe the "EPIC FAIL" is yours and your inability to read the original post.
Agreed, he's not trying to critique the item and yes, the EPIC FAIL to know what the h*** he is doing is mine. I don't know what the h** he is doing, and wonder is there anyone who does ? does he know ? what are you doing, and will we ever get a voltage reading ? No, no I don't want to know or have anything more to do with this, I need to preserve my (in) sanity !!!
Anywho, back to the topic - just from the print on the part I'd say they are fakes - all linear products tend to have really good quality printing on them.
:palm: according to what ? oh no I don't want to hope for a manufacturers web page link outlining the different silk screens that were used or how to spot a fake. I've decided not to care. Or at least to try not to care. I'm leaving this thread I hope. It's crazier than I am. and that's bad !!!!
-
Actually I am determined to see if they will "work". But I am not going to get, say, one of TiN's lovely boards all populated with expensive op-amps and resistors then put this turd in it and watch it all go up in smoke :-DD
Some simple out of circuit tests first. I mean if I fish some old transistors out the junk box I am going to at least check them before putting in a real circuit.
Checking the heater resistor was my first simple check (fail) then the transistors Q1 (appears ok) and Q2 (fail). To check the Zener I have to make a circuit. A quick read of the datasheet this seems like it's not so straightforward (any voltage differences >7V on some pins will destroy it). So here I am...
-
agreed. Speculating about whether they're fake or not based on what, the datasheet ?
BINGO!
with no information to judge if it genuine or not, what is the point ?
There is plenty evidence suggesting this is a fake.
Heck, genuine parts are often out of spec
Out of spec part = faulty
I want the half hour I've spent on this back.
I'd be much happier too if you haven't spent it writing the nonsense
PS
If you would have a jar of turd with a label saying "honey", will you taste it?
-
If you would have a jar of turd with a label saying "honey", will you taste it?
Mmmm, it does sound appetizing, is it genuine ? It's hard to tell with honey, not like software with 'microsoft' written on it, then you can be guaranteed it's a turd and you should snap it right up.
-
Were you trying to find a cheap source for them, or were you after a couple to test with? If it is the latter, linear have sent me out a few of the lm399 as a sample - I'd assume they'd do the same for the ltz1000.
Just after a couple to play with. I am only a white belt voltnut, so I don't mind playing with crusty old salvaged LTZ's on my path to zen harmony and voltnut wisdom. To be replaced with the genuine article after I have destroyed my first couple.
Linear don't do the LTZ1000 as samples, but thanks for reminding me about the LM399's :-+
-
Speculating about whether they're fake or not based on what, the datasheet ? The genuine part may have had several different production runs with several different silkscreens. Where is the information on determining if it is real or not ? Heck, genuine parts are often out of spec, what does that prove, and at the end of the day, with no information to judge if it genuine or not, what is the point ?
Yes the silkscreen can change, however, all the linear parts I've ever seen have silkscreens that are pretty much identical. So these might be from a different run, but I find it highly unlikely that they are. If you look at bunnie's issues with micro sd cards (http://www.bunniestudios.com/blog/?page_id=1022 (http://www.bunniestudios.com/blog/?page_id=1022)) some of the firsts hints that something might be wrong are dodgy silkscreens. Then add the fact that these are very high quality components (not cheap jelly bean parts) and the difference in the silkscreen makes it very likely that these are fakes.
If you are getting genuine parts that are out of spec., you need to change the manufacturer you buy from. If something doesn't meet its specification, either it is a faulty product or it has been damaged (perhaps during transit or when unpacking). No part should ever be out of spec..
I'm thinking if the voltage output is not measured then nobody knows whether it is even worth keeping an eye out for these supposed fakes, which may be genuine without guidance from the original manufacturer it's hard to know.
Again if these were jelly bean parts I might have agreed with you - but they are not. The ltz1000 is a very specialised reference, which no other reference can really rival. So is a knock off version going to perform equally well? I doubt it. Is it going to perform better considering the price? I doubt that too. It might, but chances are it won't - I can't see a random guy being able to better the lm399 (which is around the price these were bought for).
So it might be fun seeing if these work, but there is very little chance they will be good enough (and consistent enough) to keep an eye out for.
:palm: according to what ? oh no I don't want to hope for a manufacturers web page link outlining the different silk screens that were used or how to spot a fake. I've decided not to care. Or at least to try not to care. I'm leaving this thread I hope. It's crazier than I am. and that's bad !!!!
As I said above, the silkscreen being different can be a sign that it is a fake. Linear normally have a very high quality silkscreen - so what is more likely a random off day at the factory that produces high precision devices, or it being a fake?
-
Were you trying to find a cheap source for them, or were you after a couple to test with? If it is the latter, linear have sent me out a few of the lm399 as a sample - I'd assume they'd do the same for the ltz1000.
Just after a couple to play with. I am only a white belt voltnut, so I don't mind playing with crusty old salvaged LTZ's on my path to zen harmony and voltnut wisdom. To be replaced with the genuine article after I have destroyed my first couple.
Linear don't do the LTZ1000 as samples, but thanks for reminding me about the LM399's :-+
Fair enough, but if you are only wetting your feet, you might find the lm399 much easier to deal with. There are a lot of things you have to consider to get a stable output from it (i.e. not drown it out with noise and drifts from other components). With the ltz1000, there are a hell of a lot more - with a much higher need to use high precision resistors.
As a bit of an example, I've got an lm399 on a breadboard for testing - it currently is stable to around 30 uV, with the largest issue caused by the voltage supply drifting (due to temperature), which throws the constant current circuit out. I haven't even considered getting the output to the more ideal 10V yet. Whilst I haven't been doing this constantly (only odd weekends and evenings), I've been doing this since February.
I've just check the linear website and you are right - they don't offer the ltz1000 as a sample on their website. But it might be worth contacting them - I've got samples from different companies before by just dropping them a nice email.
-
Unlike the LM399, the "heater" in an LTZ1000(A) *IS* an actual resistor. It is made up of silicon, and because it is N-type material and the substrate is P-type material, it also forms a diode with the substrate. The value will vary depending on the process variables, but the data sheet has a range for it's resistance at room temperature [25C]. In all fake LTZ's that I know about, this heater resistor is not in the device.
Yes. Measures o/c in the proper direction. I tried powering it with 15V and even a 1N4148 going from pin 2 to ground. No current. Obviously no heat.
Not well known is that in China, if you [knowingly] use your merchant bank account to defraud someone, you can get 20 years in prison. It's a good bet that you can get a refund. They will claim ignorance, and offer your money back [you may not even have to send them back]. Contact the seller before you cut them open.
I have already asked for a refund. Bloody hell with the amount of fakes China is known for I didn't expect them to give 20 years for it. Ok, I know some of the real scandals have led to the death sentence which is pretty swift - bullet to the head and an invoice to the relatives for the cost of the bullet is what I have heard.
If they let you keep them, then you can cut them open. Find someone that has a microscope with a camera [the biology department in your local university would have one of these], and get a close-up of the die under different light colors. It would be interesting to see what's inside.
Yes I plan on this. I've got a microscope with camera (just a toy - Bresser Biolux thing) so that might do. I was going to tell the seller that Trading Standards and EU law prohibit me returning fakes if he tries asking for them back. But maybe going all legal is a bit much I don't want them to get 20 years or a bullet to the head! :-DD
-
People can report their broken parts. While I'm sure there are those who desire to live in a land of make-believe rather than checking sources, and most of the make-believers work in the western press, in China they do have the death penalty for people who sell deadly products or harmful medicines that kill people. Naturally some news outlets want to abbreviate that to just 'products' rather than include the words deadly or dangerous, and some people love to perpetuate that exaggeration as fact.
http://www.telegraph.co.uk/news/worldnews/asia/china/4315627/Two-sentenced-to-death-over-China-melamine-milk-scandal.html (http://www.telegraph.co.uk/news/worldnews/asia/china/4315627/Two-sentenced-to-death-over-China-melamine-milk-scandal.html)
Whereas if you want to make a deadly mess with impunity, try Texas. Where a fertilizer plant has no oversight even though it has more than 1,000 times more hazardous material than it needs to mandate it. The plant had no sprinklers, no fire walls, and no deluge systems, when it exploded and killed lots of people, nobody was held accountable. In fact the government praised everyone involved.
Governor Rick Perry says that no more regulation would have stopped this tragedy. Republican Texas State Senator Donna Campbell echoed these thoughts in an interview with the Texas Tribune. She said that there is "a point at which you can overregulate" companies. "I think we're doing a good job," she said. Fourteen people died, most of them first responders, and 200 more were injured.
So if you want to report the voltage reference as faulty product, please do, the manufacturer won't get imprisoned in Gitmo and tortured forevermore.
-
People can report their broken parts. While I'm sure there are those who desire to live in a land of make-believe rather than checking sources, and most of the make-believers work in the western press, in China they do have the death penalty for people who sell deadly products or harmful medicines that kill people. Naturally some news outlets want to abbreviate that to just 'products' rather than include the words deadly or dangerous, and some people love to perpetuate that exaggeration as fact.
http://www.telegraph.co.uk/news/worldnews/asia/china/4315627/Two-sentenced-to-death-over-China-melamine-milk-scandal.html (http://www.telegraph.co.uk/news/worldnews/asia/china/4315627/Two-sentenced-to-death-over-China-melamine-milk-scandal.html)
Whereas if you want to make a deadly mess with impunity, try Texas. Where a fertilizer plant has no oversight even though it has more than 1,000 times more hazardous material than it needs to mandate it. The plant had no sprinklers, no fire walls, and no deluge systems, when it exploded and killed lots of people, nobody was held accountable. In fact the government praised everyone involved.
Governor Rick Perry says that no more regulation would have stopped this tragedy. Republican Texas State Senator Donna Campbell echoed these thoughts in an interview with the Texas Tribune. She said that there is "a point at which you can overregulate" companies. "I think we're doing a good job," she said. Fourteen people died, most of them first responders, and 200 more were injured.
So if you want to report the voltage reference as faulty product, please do, the manufacturer won't get imprisoned in Gitmo and tortured forevermore.
:palm:
Is everything to you lead to anti US sentiments and anti Microsoft and anti Apple or any other US corporation?
You keep disrupting threads on this forum left and right and you are just asking for your feathers to be plucked.
Edit: derailing threads is probably a more accurate phrase.
-
So if you want to report the voltage reference as faulty product, please do, the manufacturer won't get imprisoned in Gitmo and tortured forevermore.
Something is really wrong with you. quadruple posting in every single thread that you post no relevant information...
All the hate towards the USA. I mean, do you realize this is not a political forum?
And attacking anyone who disagrees with you. Even if multiple engineers keep posting the same, no, you dont agree, because whatever makes you dont believe that.
I can only hope some moderator will review you account. :--
-
Sorry guys, I should have known I might "trigger" TheHeadlessChicken. For the first time ever I am considering using the filter option. SNR is way too much here which is a shame. Oh, and "Nurse! Nurse! he's out of his straitjacket again! Nurse! please help us!"
-
Ah, I found it - click on Profile > Modify Profile > Buddies/Ignore List :box:
-
Yip... That works.
-
That doesn't help because threads will still be de-railed.
-
Ah, I found it - click on Profile > Modify Profile > Buddies/Ignore List :box:
Thank you for poinitng that out. ^-^
-
Done... no more "TheElectricChicken".
Ditto, one less troll.
-
Hi,
If you open the LTZ1000ACH you will be able to see the die. It should look like this:
http://cache.ourdev.cn/bbs_upload782111/files_17/ourdev_464495.JPG (http://cache.ourdev.cn/bbs_upload782111/files_17/ourdev_464495.JPG)
It is very unique.
Also:
https://dev.xdevs.com/projects/kx/repository/entry/img/ltz/ltz1000a_1.jpg (https://dev.xdevs.com/projects/kx/repository/entry/img/ltz/ltz1000a_1.jpg)
Jay_Diddy_B
-
You can see die structures with naked eye as well, features are pretty big.
-
More interesting labeled photo of a LTZ1000 die:
https://doc.xdevs.com/doc/xDevs.com/KX/Die_LTZ1000_small.jpg
It seems there are several heater resistors available but not all of them are bonded.
-
Excellent pics. I hope I get something worthwhile to show.
I've let the guy know they are tested as not working and appear to be fake and asked for a refund but got the strange reply "You will not be able to test this is no problem with me is to test good delivery" :-// Even running it back and forth English/Chinese through google translate a few times (aka "chinese whispers") doesn't help :-DD
-
My chinglish decoder says "I tested them all good, I think your test procedure is incorrect, that's why your result - fail".
-
I'm trying to chat and he insists on chinese so I am using google translate. He is telling me "This is a test of good hair" and "Quality is no problem" :-DD I told him I will raise a dispute but I'll go ahead and peel one open anyway. Not only does the heater not work but the 7V zener doesn't clamp!
He's now asking me to send them back... |O (Sorry, thats illegal in my country - dispute time) :box:
-
Ah, I found it - click on Profile > Modify Profile > Buddies/Ignore List :box:
Thank you! Getting rid of that drivel is quite therapeutic.
-
!!! TRIGGER WARNING !!! ARACHNOPHOBES LOOK AWAY NOW! (oops, too late...)
The genuine AliExpress bargain basement LTZ1000ACH in all its naked glory:
(http://i.imgur.com/qPdSSX7.jpg)
Sorry about the pic. Camera phone and poor light. I'll see if my old microscope works in Windows 10 and try and get a better look at the die.
-
Poor thing, and it had such a nice head of hair too. Quality part indeed. :-DD
-
Die looks like it is an opamp, with those large compensation capacitors on it. Interesting that you can also see the welded extensions on the legs as well. Kind of like they did hair braiding.
-
Excellent pics. I hope I get something worthwhile to show.
Send me PM if you would like to know exact IC they are selling as LTZ.
-
I got out my microscope. Only 640x480, but not bad... I can make out an initials FB
(http://i.imgur.com/ysMap0R.png)
So what might it be?
-
Excellent pics. I hope I get something worthwhile to show.
Send me PM if you would like to know exact IC they are selling as LTZ.
Why not just tell us all?
-
I will need the die to use better magnification.
So if someone will send me the IC I will try to identify and of course post what it is or at least hires picture like I did for LTZ1000 from TiN.
-
Here are photos of one that I got from eBay. It came from a US seller and I paid around $45 for it. The pins are about .45mm in diameter. I measured the resistance of the heater and it's about 240 Ohms.
-
So what might it be?
The question is: what is cheap and comes in a TO-99 package?
The structures on the left side of the die look like 3 "power stage transistors". right?
The only chip I know with 3 output transistors is a NE555. (best guess).
No its not a genuine NE555, but it might be a CMOS or fake derrivate. >:D
With best regards
Andreas
-
There is mark FB on Macbeth chip. You ca guess what it is :)
-
I didn't know Facebook is fabricating its own silicon circuits. :-DD
-
Thanks plesa. So there's the FB in the middle and BF on the top left which my microscope didn't pick up. I must upgrade the firmware to that next model up that you use ;)
I'm guessing they are for lining up the back and front images correctly in manufacture rather than a makers mark.
Although it could be a top secret state of the art quantum processor component from Facebooks data warehouse that Zuckerberg had labelled LTZ1000 to hide it away from industrial espionage. Somehow it ended up on Aliexpress :-DD ... or maybe its an op-amp
-
Interesting tear-downs of various counterfeit parts:
http://sound.westhost.com/fake/counterfeit-p1.htm (http://sound.westhost.com/fake/counterfeit-p1.htm)
-
I recently purchased some LTZ1000's from AliExpress.
Did you get an answer back yet from Aliexpress? What happened?
I swear I wrote a small account of what happened but it seems to have just vanished. Suffice to say I got my refund in full.
-
Just buy them from LT direct shop ;)
-
Just buy them from LT direct shop ;)
That would be by far the best. I've given up on LTZ for now and got a pair of LM399A's and associated opamps direct from LT to play with (free samples :-+ )
-
Just buy them from LT direct shop ;)
To add my 2cents: simply buy the LTZ1000 CH, NOT the A version.
That's 35$ from LT, and it's better for building a REAL precision reference.
I'm still wondering, why so many people still copy the faults of HP!
I prefer copying superior Pickerings (7001), Datron 4901, Datron 1281 references, all using non-A @ 45°C or slightly above!
Frank
-
Can you not also run the 'A' at 45 degrees? Isn't that what HP3458A "Option 002" is?
-
Option 002 is just selected reference with lower drift, still @high temperature.
Some people use battery to run LTZ, so that would be better for A-version. Also after all, both versions would be not that far off after aging/conditioning completed.
-
Can you not also run the 'A' at 45 degrees?
Nope, the thermal isolation automatically increases necessary oven temperature by 10°C.
With 35°C max room temperature, plus 5°C regulation margin plus 10°C adder for A version, that's not feasible.
The "A" version better runs at 55°C, non-A can run at 45°C.
Environmental temperature has to be limited, and monitored, in either case.
As TiN already wrote, 45°C or 55°C, and even 65°C give nearly the similar stability, after some time.
45°C oven temperature is a non-brainer concerning drift, anyhow.
Frank
-
Can you not also run the 'A' at 45 degrees?
Nope, the thermal isolation automatically increases necessary oven temperature by 10°C.
With 35°C max room temperature, plus 5°C regulation margin plus 10°C adder for A version, that's not feasible.
The "A" version better runs at 55°C, non-A can run at 45°C.
Environmental temperature has to be limited, and monitored, in either case.
As TiN already wrote, 45°C or 55°C, and even 65°C give nearly the similar stability, after some time.
45°C oven temperature is a non-brainer concerning drift, anyhow.
Frank
I've read the above statement several times now in this forum, but from the data sheet of the the LTZ I don't see any evidence that the A version cannot be run at 45°C or 55°C. From what I understood the chip is just better isolated from the (bottom of the) case so that the same temperature can be achieved with lower heater power.
On page 3 of the data sheet "temperature rise vs heater power" it is shown that the A version just needs 0.1 W for 45°C while the non-A needs around 0.25W.
Do I miss something here?
J.C.
-
Hi all,
For reference........photo of an LTZ1000 I got from Digikey.
Also pictured some LM399's I got from Ebay..........hmmmmm!
Ian.
-
Those LM399's look dodgy, my samples direct from LT have PHILIPPINES E14237 stamped on the sides. Yours don't have anything!
-
Some photo's of quite old LTZ1000 's which are not fakes and have all silver pins (never been soldered, only plugged into a socket for burn in), for your further pictorial reference (Sorry for the pun , I will whip myself later!)
Work had purchased a 'Decent' quantity for the purpose of selecting after the aging was completed.
The measured heater resistance on 2 units from the pic was 276 ohms and the other 279 ohms, pins 1 and 2.
-
Those LM399's look dodgy, my samples direct from LT have PHILIPPINES E14237 stamped on the sides. Yours don't have anything!
Interestingly I also have a number of old LM399 refs and those have no text on the side either, and look the same as the pic from IanJ. ie National semi branded.
Perhaps his are old stock too !.
-
Some photo's of quite old LTZ1000 's which are not fakes and have all silver pins (never been soldered, only plugged into a socket for burn in), for your further pictorial reference (Sorry for the pun , I will whip myself later!)
Work had purchased a 'Decent' quantity for the purpose of selecting after the aging was completed.
The measured heater resistance on 2 units from the pic was 276 ohms and the other 279 ohms, pins 1 and 2.
Maybe it's only the 'A's that have the gold pins? I mean they need something to justify the extra cost!
-
Some photo's of quite old LTZ1000 's which are not fakes and have all silver pins (never been soldered, only plugged into a socket for burn in), for your further pictorial reference (Sorry for the pun , I will whip myself later!)
Work had purchased a 'Decent' quantity for the purpose of selecting after the aging was completed.
The measured heater resistance on 2 units from the pic was 276 ohms and the other 279 ohms, pins 1 and 2.
Maybe it's only the 'A's that have the gold pins? I mean they need something to justify the extra cost!
Both have gold plated pins now ( I have 4 LTZ1000 manufactured 1443).
-
Link to the CERN ultra precision 10mA current reference (based on LTZ1000A):
http://cds.cern.ch/record/643294/files/cer-002399331.pdf (http://cds.cern.ch/record/643294/files/cer-002399331.pdf)
-
Wow, 6 centuries ago.
That's some serious aging! Designed by a druid, manufactured by blacksmiths hammer and anvil, leads turned to gold by alchemy, then ovenised over the centuries in the blacksmiths forge and passed from father to son... :-DD
Link to the CERN ultra precision 10mA current reference (based on LTZ1000A):
http://cds.cern.ch/record/643294/files/cer-002399331.pdf (http://cds.cern.ch/record/643294/files/cer-002399331.pdf)
Thanks for the link! Good to know CERN has a couple of projects as well as messing around with the LHC and KiCAD development :)
-
Those LM399's look dodgy, my samples direct from LT have PHILIPPINES E14237 stamped on the sides. Yours don't have anything!
Interestingly I also have a number of old LM399 refs and those have no text on the side either, and look the same as the pic from IanJ. ie National semi branded.
Perhaps his are old stock too !.
Some of them are dated 1988/89.......
Anyways, i've just created a 10-way pcb that allows 10 LM399's to be plugged in and aged (partially). Each channel has it's own self-biasing op-amp and low tempco resistors........so this weekend I'll drop in a bunch of my LM399's, including some of the Ebay ones and see what occurs.
Ian.
-
Link to the CERN ultra precision 10mA current reference (based on LTZ1000A):
http://cds.cern.ch/record/643294/files/cer-002399331.pdf (http://cds.cern.ch/record/643294/files/cer-002399331.pdf)
Remarkably about this paper is one of the authors and main designer of this CC circuit.
Mr. Pickering is the wizard of the LTZ1000, i.e. he was working for Datron and developed the 7000 reference.
The major design elements, like the TaN resistors, went into this CC source.
Frank
-
Both have gold plated pins now ( I have 4 LTZ1000 manufactured 1443).
Hello,
Depends wether you have lead free or non RoHS devices:
http://www.linear.com/designtools/leadfree/ (http://www.linear.com/designtools/leadfree/)
With best regards
Andreas
-
Link to the CERN ultra precision 10mA current reference (based on LTZ1000A):
http://cds.cern.ch/record/643294/files/cer-002399331.pdf (http://cds.cern.ch/record/643294/files/cer-002399331.pdf)
From that paper:
The gain step to 10 V from the well-known and predictable zener reference LTZ1000A, at 7.2 V, is defined by “statistical” TaN film resistor arrays [2]. These have been shown to maintain stable ratios to less than one part in 10^6 over time and temperature. The 10 V is then converted to 1 mA in a transconductance stage, utilizing a 10-k ohm resistor constructed from four of Vishay’s new “Z-Foil” zero TC resistors.
Any ideas why they chose TaN statistical arrays for the 7.2 to 10V conversion and Z-foil for the 10V to 1mA conversion rather than using the same, either TaN array or Z-foil, for both? Could it be that 10mW (1mA x 10K ohm) would create more self heating errors in a TaN array than the Z-foil? They use another TaN array for the subsequent 1 to 10mA current mirror but presumably can use much lower value resistors to limit self heating to acceptable levels?
Does anyone have any good references to the long term stability of TaN resistors and even better, TaN statistical arrays?
-
After reading the data-sheets on TaN-film resistors [especially those from IRC/TT-Electronics] I see that the long term stability and stability with the various torture tests can be almost as good as Evanohm-based bulk metal foil resistors [i.e. Z-foil and Z1-foil]. Tantalum Nitride resistors laugh at humidity, and after an initial accelerated burn-in, their time stability is quite good.
I assume you mean the IRC Ultride resistors - they have shelf life spec of 100ppm absolute or 20ppm tracking/1 year which is nowhere near as good as the spec for the Vishay hermetic metal foils of < 2ppm over 6 years absolute. Yes the stability should improve after the first 1000 hours but I'd be surprised if it would get much better than 10 to 20ppm/year absolute.
Vishay thin film networks have similar specs. I couldn't find anything which suggests that TaN are any more stable than other types of precision thin film resistors - one Vishay document suggests that their nichrome and 'Tamelox' (a proprietary nichrome and TaN alloy) are more stable than TaN except in high humidity conditions.
I did find this ATP document http://www.thinfilm.com/equation.html (http://www.thinfilm.com/equation.html) which provides a TaN aging formula but it calculates improbably low aging drifts at 25C - eg. < .003ppm over 25 years!
Any ideas why they chose TaN statistical arrays for the 7.2 to 10V conversion and Z-foil for the 10V to 1mA conversion rather than using the same, either TaN array or Z-foil, for both? Could it be that 10mW (1mA x 10K ohm) would create more self heating errors in a TaN array than the Z-foil? They use another TaN
Having thought about it I guess the reason is obvious - the accuracy of the 10V to 1mA conversion depends on the absolute stability of the resistor and the TaN resistors are nowhere near good enough hence the hermetic Z-foil. The accuracy of the 7.2 to 10V conversion and the current mirrors depends on the resistors' ratio tracking stability where the much better tracking stability of TaN resistor networks, along with further improvement using statistical arrays, makes their use appropriate.
The best temperature coefficients spec I could find for TaN resistors is 5ppm absolute which wouldn't be good enough for the 10V to 1mA conversion but the best tracking TCRs are 1ppm which would also improve using statistical arrays. Z-foils could also be used in statistical arrays but the cost would be prohibitive.
-
I did find this ATP document http://www.thinfilm.com/equation.html (http://www.thinfilm.com/equation.html) which provides a TaN aging formula but it calculates improbably low aging drifts at 25C - eg. < .003ppm over 25 years!
It is all right with this formula. Aging drifts at 25C is 0.04 ppm/year.
-
I did find this ATP document http://www.thinfilm.com/equation.html (http://www.thinfilm.com/equation.html) which provides a TaN aging formula but it calculates improbably low aging drifts at 25C - eg. < .003ppm over 25 years!
It is all right with this formula. Aging drifts at 25C is 0.04 ppm/year.
Sorry I don't understand where you got .04ppm/year from. When I plug t=8760 hours, T=25C into their equation:
drift = [1.51 x 10^12 x t^.61 x exp(15070/(T+273.1))]%
I get .0004ppm drift per year. Using their example of t =10^8 hours and T=92.2C the equation above gives 0.133% which matches their result.
I simply don't believe any of these numbers are remotely credible. 0.13% drift from a resistor operating at 92.2C for 11,415 years??? Or .04ppm or .0004ppm/year at 25C?
-
Bad calculator? :)
-
Bad calculator? :)
No - that equation (claims) to give results in % hence our 100x disagreement!
-
This equation is full of it, useless, the only alloys which have proven long term stability over years that are below 0.1 PPM/year are Evanohm and Manganin (under very controlled conditions), these figures were confirmed over decades of time. I do not believe these aging calculations, not even close to real world, particularly not under normal operating conditions. I have not seen any published actual long term figures for any TaN version resistors that verify this silly equation, projections based on short term aging tend to be inaccurate. Maybe, under very favorable conditions, TaN might be closer to 1 PPM/year but under average operating conditions it is likely to be at least several PPM/year at best.
Remember the CERN document is 25+ years old and frankly I see absolutely nothing special about Pickering's circuit, for that matter I see nothing particularly 'special' about Pickering's LTZ1000 reference either. The Datron reference was about as good as Fluke's at the time, with perhaps some minor improvement in specs, that circuit was superceded in just a few years years.
-
Evanohm
Sorry off topic, but what was the standard/recommended/optimal case size of your Evanohm resistors under 100 kohm?
6.35mm x 12.7mm (1/4" x 1/2") maybe?
Probably mentioned somewhere, but I couldn't find.
-
Hi LTZ2000,
That is the size (805) that most customers purchase although I had one who insisted on my 708 which is 0.150" D x 0.300" L. In some cases where the customer wants a ratio match with close TCR tracking, depending on the values, I will recommend two different resistor sizes and method of thermal coupling for best performance. I have achieved TCR tracking of <0.2 PPM/°C when properly executed. I usually discuss the application with the customer to determine the best solution.