Adjustable Buck Voltage Regulator - AliExpress - Exiciting finds

[ez24]

Hi 

I just have to share my exciting finds today.  I found AliExpress today on this forum and found this
buck regulator:

http://www.aliexpress.com/item/Wholesale-1pcs-DC-Step-Down-Converter-DC-4-0-38V-to-1-25V-36V-5A-75W/32285636716.html

Scroll down and you will find a treasure :  specs and documentation (readable)

Boy did I have to get a couple of these,  and heat sink, standoffs, documentation   

wow  2 finds today       

Sure glad to see ebay competition

Wasn't that hard to sign up and order - one password for Ali and one for payment (no paypal)  - same as fleabay

Edit:  Another exciting find - I can print out the complete ad with the specs.  Ebay cuts the ad and details off.  With ebay I have to copy and paste into Notepad to print the details out

Edit:  Another exciting find - AliEpress allows product reviews like Amazon 

[Canobi]

I've used Aliexpress myself a couple times and was happy with what I got. Another one I use is Deal Xtreme. I got some relatively good quality scope leads just recently for very little, shipping is also free on most items, though it's less component oriented than Aliexpress.

[wraper]

Just don't expect it running at full rated power.

[splin]

Hi 

I just have to share my exciting finds today.  I found AliExpress today on this forum and found this
buck regulator:

http://www.aliexpress.com/item/Wholesale-1pcs-DC-Step-Down-Converter-DC-4-0-38V-to-1-25V-36V-5A-75W/32285636716.html

Scroll down and you will find a treasure :  specs and documentation (readable)

Boy did I have to get a couple of these,  and heat sink, standoffs, documentation   

“Danger, Will Robinson!”   

Those 50V 220uF filter capacitors (in the heavily annotated picture) have a 650mA ripple current spec:

http://www.mouser.in/ProductDetail/Lelon/VZH-221M1HTR-1010/?qs=sXzuR3ywNuc25mceIPLw6w%3d%3d

(That says 553mA but the datasheet shows 650mA). With a 5A output current there's a *very* good chance that the input capactitor is going to explode and possibly take somebodies eye out - ask me how I know (it wasn't this particular convertor and it missed my eye btw)!

This shows the equations for calculating the input and output capacitor ripple currents:

http://rohmfs.rohm.com/en/products/databook/applinote/ic/power/switching_regulator/capacitor_calculation_appli-e.pdf

Look at the equation on page 3. You would probably find it a bit tricky to determine the inductor ripple current but its not a very significant term anyway. Assuming the inductor ripple current is 1/3 the 5A output current; with an output of 5V and input voltages between 6 and  36V  the peak input ripple current is 2.52A at Vin = 2 * Vout and a minimum of 1.39A at Vin = 36V.

If you have a low impedance power source and *very* short, low inductance connecting leads then you may be OK as most of that ripple current will come from the source rather than the filter capacitor, but how are you going to measure it to be sure?
Bear in mind that the filter capacitor has a specified impedance of 120m ohms at 100kHz so the total source impedance of your power source and the connections and the PCB traces from the input connector to the switching mosfet needs to be less than approx 120 milli-ohms * .65A/2.52A milli-ohms , or  < 31 milli ohms to keep the capacitor's ripple current below 650mA.

At a switching frequency of 300kHz the reactance of an inductor consisting of a single 8mm diameter loop of wire will exceed that, never mind the wire resistance. The convertor's switching freqency could even be 500kHz or higher making the problem worse.

I would replace the input capacitor, at least, with something with a much higher ripple current rating such as:

http://www.digikey.com/product-detail/en/EKZN500ELL681MK25S/565-4071-ND/4843881

An aluminium-polymer capacitor is much smaller for the same rating but more expensive:

http://www.mouser.in/ProductDetail/Nichicon/RNU1H470MDN1PH/?qs=sGAEpiMZZMs9%2fSpGH%2fyc7JYN24Rot9XpwJjNTEwdkxU%3d

Now you know why those switching convertors are so cheap - they cost reduce them to the bare minimum and then carry on removing essential parts!

Note that ripple current ratings vary enormously with capacitor type and size so you need to check the exact specs.

Alternatively, you could leave the existing capacitor and connect a suitable capacitor as close as possible to the input connections, but beware the inductance again from the leads and PCB traces.

The ripple current is much less in the output capacitor and probably won't be a problem, but worth trying to estimate from the Rohm link.

In the mean time, either limit the current output to not much more than 1A, or a tin-hat and safety vizor might be a good idea!

[Yansi]

lol,  220uF basic aluminium cap for 5 amps, what a joke! 

Might be interesting to consider polymer caps, but I've never used those nor I know much about theirs typical specs. But I smell that ripple current vs. capacitnce should be much better than standard aluminium do have.

[ez24]

Some good info - thanks splin

Will have to study all that information.

I would never expect to get full rating from Chinese modules.  I will test them and put up the test results, starting at 70% and work down.  I hoped to get a good 2a from this.

[paulie]

LOL. I was wondering how long it was going to take for the china haters and fear mongers to chime in.

Out of millions (several thousand personally) of similar type out there I doubt even one has taken any eye out or exploded due to ripple spec. It's true that these do not perform as advertised, even with the heatsink warned about in most listings,  but no problem running an amp or two. Now let's bring up the "huge ripple" which might not actually interfere with most applications but might just be detectable with a scope with enough sensitivity. Or what about "fake" chips? That subject's always good for some healthy hand wringing.

Listening to problems brought up by units "properly" designed here from both hobbyists and professionals, IMO the Chinese seem to have a much better handle on these than crybaby western experts.

I'm positive running with scissors has taken a bigger toll on eyesight and don't forget those wolves under the bed.

[splin]

LOL. I was wondering how long it was going to take for the china haters and fear mongers to chime in.

Out of millions (several thousand personally) of similar type out there I doubt even one has taken any eye out or exploded due to ripple spec.

*Mine* did 

Explode - that is - fortunately I wasn't looking at it at the time. Hardly surprising when the input capacitor's ripple current was nearly 5 times the specified limit - that's 25 x the heating power - how is it not going to explode???

but no problem running an amp or two.

No, as I said in my last post, 'don't run it much beyond an amp' - because at approximately 1.2A the ripple current is still in spec! At 5A its way beyond the limit which is what I was warning about. The vast majority won't explode because they won't be running at maximum 'rated' current and at the most stressful input/output ratio, as it turns out mine was when testing it.

I'm not a China hater - I love the fact that we can get these things dirt cheap - but you have to be aware of the dangers and their limits. I bought a Chinese hot air station recently; it came with a 3 pin mains adapter which, through terrible design, didn't make contact with the earth pin. I might well have become a 'China hater', for a little while at least, if I'd got 240 of our finest National Grid's volts across my chest! And before you say it, that was a definite possibility because the live, very thinly, single insulated wire feeding power to the heater was trapped hard up against an earthed connection in the air wand. I bought it cheaply not expecting it to be safe, and it wasn't but I could remedy that.

IMO the Chinese seem to have a much better handle on these than crybaby western experts.

I've no doubt that Western safety standards are set very high - un-necesessarily and expensively so in many cases - but I think you've got a very odd perspective. My friend got severe chemical burns from a Chinese made leather sofa. Does that make us cry-babies because the products were banned?

Now let's bring up the "huge ripple" which might not actually interfere with most applications but might just be detectable with a scope with enough sensitivity

But that's 2.5A RMS of ripple current in the input capacitor - you wouldn't need a super-sensitive scope to detect it, an unloved finger and a burns specialist should do just fine.  (Because it looks like a bandaged thumb!).

It won't have any significant impact on the output voltage ripple so you're application will work just fine until the big bang and the amps go on strike or go elsewhere.

[paulie]

tin-hat and safety vizor might be a good idea!

Well that kinda says it all!. And don't forget the black helis.

Seriously though, as part of the day job handling large numbers of these, both "proper" from "official" channels and cheapies from The Evil Empire, there has been no big difference in failure rates. I've encouraged clients to return defects for analysis (AKA fix and drop in my parts box). Out of several dozen very few faults involving those caps and those related to solder joints or mechanical damage. No puffies or explosions. Dead silicon far more common. So from my perspective these are a great deal at 1/5th to 1/10th the cost.

Anyway your repeated mention of 5a does not sound like direct experience with these particular units. They simply will not supply that and in fact long before 3a simply shut down. Then you "assume" 2.5a ripple which makes comments even less credible. So it don't look like we are even on the same wavelength here.

Feel free to throw money out there for the Real Thing. Personally I hate playing the sucker.

[ez24]

thanks all

Lots of info on ripple - I sure have to study this

I will test these to 2 amp and hope I get that

[kjs]

If you use Chinese parts you have to be aware how the fairy tales (= Chinese for specifications) are made. If the switching transistor has a maximum current limit of 5A it's a 5A power supply. If the capacitor is marked with 40V maximum input voltage is 40V. There's no such thing as design margin or any concern about reliability or lifetime.
Don't get me wrong, we use LM2596 modules as we can't even buy the IC for the $0.54 we pay for the entire module but we have completely characterized the stuff and use them at limits which are far away from what they specify. All modules go through regular final test and whatever isn't as it should be is returned and replaced for free. Much cheaper and as it isn't a part which adds value it doesn't even matter whether the LM2596 are genuine or XL……. marked with TI/National/ON logo.

[Fungus]

Don't get me wrong, we use LM2596 modules as we can't even buy the IC for the $0.54 we pay for the entire module but we have completely characterized the stuff and use them at limits which are far away from what they specify.

Yep. I've had some of those voltage converters running for years. Just run them well below claimed spec, keep an eye on the temperatures, be happy that you paid a quarter of the price!

[paulie]

If you use Chinese parts you have to be aware how the fairy tales (= Chinese for specifications) are made.

I agree 100% on that. I prefer the term "bare faced liars" as more descriptive.

Fortunately Ebay buyer protection saves the day. Out of literally thousands of transactions last few years there are couple hundred cases opened. Actually it rarely gets to that because a quick message to seller almost always results in instant refund. They know what they did and just don't care because few bother to complain. They simply write off guys like me as cost of doing business.

If the switching transistor has a maximum current limit of 5A it's a 5A power supply.

We'll have to agree to disagree on this one. Transistor is actually a minor part of the system. In fact I'd put it a few down on the list below things like inductor and, yes that's right, caps. Splins warning about finger fatality is close to true, not the caps though. For a few rare applications I do have to tack on another cap for noise suppression but it's the coil that gets real hot real quick near limits. That toroid in OPs module is one reason it actually puts out close to 3a instead of 2a for the half buck version that has no pots or display and cheaper inductor.

So in my world a max current limit of 5a is a 2-3a supply.

ps. Generally for personal use I prefer the buck/boost (SEPIC) versions since there are no concerns at all regarding input or output voltage. A few cents more but worth it to make life easy. The new XL6009 buck/boost are even better with another amp over the old LM2596 and cheaper still.

[Fungus]

All this is true of any electronic component. Nothing should be run at the "absolute maximum rating" listed in the datasheet.

Even if it's something like a MOSFET bought from the most reliable supplier in the world, the "absolute maximum rating" isn't a working value, it's a theoretical limit for the part either under ideal conditions (which don't exist) or for very short bursts of time.

If you want to switch a 5A current using a MOSFET then buy a MOSFET rated for 30A. If you want a 5A power supply that will run 24/7 then buy a 12-15A power supply. That's the reality no matter where it's manufactured (unless you work at NASA/military and pay a fortune for mil-spec power supplies).

[macboy]

Don't get me wrong, we use LM2596 modules as we can't even buy the IC for the $0.54 we pay for the entire module but we have completely characterized the stuff and use them at limits which are far away from what they specify.

Yep. I've had some of those voltage converters running for years. Just run them well below claimed spec, keep an eye on the temperatures, be happy that you paid a quarter of the price!

If you are running a "5 Amp" module at 1 Amp to avoid self-destruction, did you really get a bargain on that 5 A module, or did you overpay for a poorly made 1 A module?

[Fungus]

If you are running a "5 Amp" module at 1 Amp to avoid self-destruction, did you really get a bargain on that 5 A module, or did you overpay for a poorly made 1 A module?

I got a decent price on a reasonably well made 2A module.

(If it's putting out enough amps for the purpose and it doesn't burn my fingers then it's probably going to last as long as anything you're overpaying for).

[ez24]

ps. Generally for personal use I prefer the buck/boost (SEPIC) versions since there are no concerns at all regarding input or output voltage. A few cents more but worth it to make life easy. The new XL6009 buck/boost are even better with another amp over the old LM2596 and cheaper still.

what is this ??  (SEPIC)   Do you have any links to the devices you use?
 I think some of the modules I ordered have the XL chip because I could see "X", I will not know until I get them.

Searching ebay for SEPIC I came up with this interesting one:

- high sales
- current set with jumpers (I like this)
- BUT I do not know how to set the voltage (shows my ignorance)

[url=http://www.ebay.com/itm/150-1500mA-Buck-Regulator-LED-Driver-for-1-50W-High-Power-LED-/161026417634?pt=LH_DefaultDomain_0&hash=item257dec23e2]http://www.ebay.com/itm/150-1500mA-Buck-Regulator-LED-Driver-for-1-50W-High-Power-LED-/161026417634?pt=LH_DefaultDomain_0&hash=item257dec23e2[/url]

here is another one:

[url=http://www.ebay.com/itm/LM2577-Auto-Step-Up-Step-Down-Power-Supply-Module-Solar-Power-Panel-DC-DC-SEPIC-/330809583178?pt=LH_DefaultDomain_0&hash=item4d05c9a64a]http://www.ebay.com/itm/LM2577-Auto-Step-Up-Step-Down-Power-Supply-Module-Solar-Power-Panel-DC-DC-SEPIC-/330809583178?pt=LH_DefaultDomain_0&hash=item4d05c9a64a[/url]

- this does not have current limit
AND basically how is this different (other than specs) than the one I found (mine has current limit)

Now I wish I could find one with

- current limit with jumpers
- voltage limit with pot
- LED
- large coil
- new chip
- heat sink

(ha ha probably no such animal)

FYI  I once lived in Taiwan and went into a company that made audio amps.  I asked how much power it had and they said anything I wanted.  In other words I could have had a 5000 watt amp that put out 1 watt.

[kjs]

SEPIC stands for "Single-ended primary-inductor converter". It can convert up and down. This means your output voltage can be higher or lower than the input voltage.

Example (LM2577 based, just a random link):
http://www.ebay.com/itm/DC-DC-adjustable-auto-step-down-up-power-supply-module-solar-power-panel-LM2577-/111185749981?pt=LH_DefaultDomain_0&hash=item19e32fe7dd

Generally i will avoid using such overrated modules, but you can be sure that it won't explode at full power, but the efficiency will drop, and temperature and ripple will increase.

I am still searching for one example which isn't over-rated....... And yes, if you operate them at the "specified" limits the inductor can blow up as well as the capacitors. Even when they don't blow up their lifetime is very limited. We measured the temperatures of the LM2596 modules (C and L) at room. 145C when operated to the fairy tale limits. Just use them to what the components are designed for and ignore the fairy tale and they are fine.

[Fungus]

One of these arrived yesterday, it seems to work well and it's not too badly made (has nicely rounded corners and 'feet'):

http://www.ebay.com/itm/391036308181

It can display output volts and amps and has a settable current limit (short the output terminals, dial the limit current). Very handy for seeing how much power a circuit is using. Yes it's only 10mA resolution but that's good enough for a lot of jobs.



The main chip is an MP2307 buck converter rated for 3A continuous output.

$8.60 for a "poor man's bench power supply"? Bargain!

[Rick Law]

One of these arrived yesterday, it seems to work well and it's not too badly made (has nicely rounded corners and 'feet'):
...

It can display output volts and amps and has a settable current limit (short the output terminals, dial the limit current). Very handy for seeing how much power a circuit is using. Yes it's only 10mA resolution but that's good enough for a lot of jobs.
...
The main chip is an MP2307 buck converter rated for 3A continuous output.

$8.60 for a "poor man's bench power supply"? Bargain!

I would not call that a bargain.  As said, you can't run that 3A sustained for long.  You need to give yourself some headroom.

I have been using a few similar ones as my "bench supply" for around a year.  They are fairly high noise and generally below spec (ie: give yourself head room).  If noise is not an issue, they are not too bad but this one is not a bargain.  The only thing that distinguish this one from the $3-$5 (5Amp) alternatives is the build in LED display.

This is an analog control unit using 3296 Trimpot.  Trimpot is not designed for frequent adjustments.  Looking up specs from Bourns, their relatively high quality 3296 trimpots are designed for 100 turns.  Once you use it a while (50+ times) the wiper begin to wear out.  It will start to get difficult to adjust with the jitter.  So, you better get good at trimpot replacement.  On one of mine, I replaced the trimpot so many times the copper trace was beginning to lift (got impatient, pull too hard before solder fully melt).

I switched over to using the B3603 digitally controlled.  It can do close to 2A without external cooling.  I did a "mini review" on it here:
https://www.eevblog.com/forum/reviews/b3603-dcdc-buck-converter-mini-review-and-how-the-set-key-could-be-fatal/

There are some inaccuracy there, but it give you a general feel of that machine.  I've learned a lot more since I wrote that.  The learning since then pointed me to at least one embarrassing fact: The unit I reviewed actually had a problem and I didn't even know it at the time!  The noise I screen printed is about 1/3 more than it should.  I have since repaired my unit.  My own inexperience for sure, but that is the process of learning.   I plan to update that review but I haven't got around to do it yet.

Those corrections are documented in that thread, so if you found that a potential buy, read the whole thread.  You get a feel of how the unit works and the unit's strength and weakness far better than if you just read the review.

[ez24]

[url=http://www.ebay.com/itm/391036308181]http://www.ebay.com/itm/391036308181[/url]

I had not seen this one - it has an inductor that looks half way between the small one and large one
so I ordered a couple of these to test

thanks

[kjs]

They replaced a half way decent pot-core with a lousy axial one to save on cost. Will radiate nicely on the switching frequency.

[ez24]

https://www.eevblog.com/forum/reviews/b3603-dcdc-buck-converter-mini-review-and-how-the-set-key-could-be-fatal/

thanks I have been looking for a review like this

the comment on the pot (100 turns) was a shocker - I did not see this coming
back to the drawing board


thanks

[ez24]

They replaced a half way decent pot-core with a lousy axial one to save on cost. Will radiate nicely on the switching frequency.

pot-core  ??

axial one  ??

are you talking about the inductor

"Will radiate nicely"  is this good or bad ?

thanks

[ez24]

Radiate == bad, because you don't want to have an EMI generator with you.

Pot core means magnetically closed circuit, or shielded.

Acial one means the magnetic circuit is not closed, so it will radiate to free space.

thanks I understand now BUT one more stupid question

are we talking about the inductor, it is the only thing I can think of because I think "pot core" means pot metal and the inductor is the only thing I can think of unless it is the caps.  I do not think you are talking about the ICs.

Will I be able to test EMI using a scope by just holding the probes over the thing (once I find out what the thing is ha ha)

thanks