Atten APS3003S/3005S Variable Power Supply – Review and Teardown

[Mint.]

Hello!
Recently I have bought an Atten APS3005S variable power supply off eBay and after using it for a bit I have created the following review and teardown. I have put it up onto my personal blog and wish to share it with you as this is my first proper review and teardown of such an instrument. Also I am surprised there is not much talk about this instrument here on the forum.
http://mintelectronics.wordpress.com/2012/07/09/atten-aps3003s3005s-variable-power-supply-review-and-teardown/
Any constructive critique would be welcome!
-Cheers

[David_AVD]

I have one of the APC3003S here at home for fiddling with Christmas lights, etc.  It's fine for the $95 delivered I paid a while back, but I did make a couple of mods.

First I replaced the fan with a better quality (and quieter) one.  I also made a small board (with PIC) so the fan only turns on at a preset heatsink temperature.  Most of the time I'm drawing very little current so the fan rarely kicks in.  This is nice as it's in my home office.

Interestingly, mine has ventilation down the whole of each side, not just half like the one reviewed.  I thought the build quality was nothing to write home about, but perfectly acceptable considering my use and the price point.

[saturation]

Nice review, mint, but if that voltage spike is true during turn on/off, this device can damage circuits.  This is fully unacceptable behavior in any PSU, even for hobbyists, considering you can get a cheap Chinese PSU, like the copies of this supply by Mastech, that provdie better performance for the same money.  I'd rate it 0/10, its really very bad, one of the worst I've seen.

[Mint.]

Nice review, mint, but if that voltage spike is true during turn on/off, this device can damage circuits.  This is fully unacceptable behavior in any PSU, even for hobbyists, considering you can get a cheap Chinese PSU, like the copies of this supply by Mastech, that provdie better performance for the same money.  I'd rate it 0/10, its really very bad, one of the worst I've seen.

I kind of agree with you, I updated the post with another picture of the waveform of the spike itself, its really short and it does not seem to be that damaging in my opinion. However I am not sure if this is true or not since I don't know much about such circumstances, I assume its like static electricity where you can get a high voltage and current hit you, but its over so quickly you don't get hurt?

[Mint.]

I have one of the APC3003S here at home for fiddling with Christmas lights, etc.  It's fine for the $95 delivered I paid a while back, but I did make a couple of mods.

First I replaced the fan with a better quality (and quieter) one.  I also made a small board (with PIC) so the fan only turns on at a preset heatsink temperature.  Most of the time I'm drawing very little current so the fan rarely kicks in.  This is nice as it's in my home office.

Interestingly, mine has ventilation down the whole of each side, not just half like the one reviewed.  I thought the build quality was nothing to write home about, but perfectly acceptable considering my use and the price point.

There is a thermostat that is supposed to go onto the heatsink from the two power transistors, but its not present in the 3003S version I think. Somebody on YouTube posted a video about the upgrade.

 

[David_AVD]

Yeah, I saw that video a while back, but had already done my fan PCB.  I was going to implement proportional fan speed but the PWM made the fan whine unacceptably.  I might revisit the fan code / drive someday when I'm bored.

[Devian]

Nice review, mint, but if that voltage spike is true during turn on/off, this device can damage circuits.  This is fully unacceptable behavior in any PSU, even for hobbyists, considering you can get a cheap Chinese PSU, like the copies of this supply by Mastech, that provdie better performance for the same money.  I'd rate it 0/10, its really very bad, one of the worst I've seen.

Hi, saturation,thank you for your sharing, you think PSU supplied by Mastech is good, which exactly of the PSU model?

[nctnico]

A spike during turn-on/turn-off will damage circuits for sure. So kick it in the dumpster before you pull your hairs out.

[David_AVD]

It would also be wise to check that there really is an output spike during power-up.

Remember, what you see on a 'scope is not always the real story.   

[Mint.]

It would also be wise to check that there really is an output spike during power-up.

Remember, what you see on a 'scope is not always the real story.   

Any idea how I can test this?

[Mint.]

A spike during turn-on/turn-off will damage circuits for sure. So kick it in the dumpster before you pull your hairs out.

Thats not the way to go! Even if it has such a flaw, I can still disconnect the circuit and then turn the supply off.

[nctnico]

A spike during turn-on/turn-off will damage circuits for sure. So kick it in the dumpster before you pull your hairs out.

Thats not the way to go! Even if it has such a flaw, I can still disconnect the circuit and then turn the supply off.

One day you'll forget that or someone else who is unaware of the problem uses the PSU to test some delicate equipment. I think I posted a story before about such a PSU and the amount of problems it caused.

[David_AVD]

All I meant was make sure the spike you're seeing is really present on the output, not induced another way.

I've not had any issues with my PSU, but it would be interesting to capture the peak voltage output at switch on.

[Mint.]

All I meant was make sure the spike you're seeing is really present on the output, not induced another way.

I've not had any issues with my PSU, but it would be interesting to capture the peak voltage output at switch on.

I've ran a few tests this morning and I am still puzzled by this phenomenon on the scope it shows the waveform, but any load or meter that I connect to it does not show anything. Even though I remember I hooked up an LED last time I tested this and it lit up a tiny amount. Hmm. It flashes a bit, but then if I disconnect the scope and try again its a little bit less brighter of a flash and doesn't always work. Maybe its the scope.

[David_AVD]

You could build a little comparator circuit to detect voltages over a certain threshold.  Set the comparator to trip (and latch) at a little over the desired test voltage.  Might be a simple and fun project for you?

[saturation]

You won't get hurt, naturally mint, static electricity can be kV of very low current, but your circuits will, that's why MOS devices are protected from static; this Atten is like having a static discharge every time you turn it off.  damage its just a question of time.  You can't remember to remove the DUT under power before turning the Atten off all the time, and if the designers of Atten made such a simple albeit heinous error, what else will fail later on, that you haven't tested for.

ATTENtion to detail is something so much Atten gear lack.  If the designers were perfectionists it means, most anything else has been addressed.  If they let little things slide, there can be more problems later that you haven't tested for.

Consider the Atten 858D rework station, the error there are so heinous, it will injure the user, but the good news its easy to fix.  However, there is no other device like it [ The Yihua or WEP brand are all built similarly, with problems] so buyers have no choice.

But in PSU, you have many choices.  Why choose one with such errors?

..it does not seem to be that damaging in my opinion. However I am not sure if this is true or not since I don't know much about such circumstances, I assume its like static electricity where you can get a high voltage and current hit you, but its over so quickly you don't get hurt?

[saturation]

Anything in their catalog is worth considering, its a question of what you want.  Reviews for some units are in the eevblog archives:

http://www.mastechpowersupply.com/variable-power-supply/linear-power-supply/cat_6.html

Nice review, mint, but if that voltage spike is true during turn on/off, this device can damage circuits.  This is fully unacceptable behavior in any PSU, even for hobbyists, considering you can get a cheap Chinese PSU, like the copies of this supply by Mastech, that provdie better performance for the same money.  I'd rate it 0/10, its really very bad, one of the worst I've seen.

Hi, saturation,thank you for your sharing, you think PSU supplied by Mastech is good, which exactly of the PSU model?

[leppie]

All I meant was make sure the spike you're seeing is really present on the output, not induced another way.

I've not had any issues with my PSU, but it would be interesting to capture the peak voltage output at switch on.

I only have an analog scope, so I could not check it, but from my UNI-T 61E in peak mode, I have seen spikes up to 85V when powering off. The unit was set to 5V output, CV mode.

Now I wonder if I should return it...

[David_AVD]

All I meant was make sure the spike you're seeing is really present on the output, not induced another way.

I've not had any issues with my PSU, but it would be interesting to capture the peak voltage output at switch on.

I only have an analog scope, so I could not check it, but from my UNI-T 61E in peak mode, I have seen spikes up to 85V when powering off. The unit was set to 5V output, CV mode.

Now I wonder if I should return it...

Maybe try with a 10K resistor across the meter leads so that the high input impedance of the DMM isn't picking up stray spikes.  85V is more than the raw supply rail in that PSU.

[leppie]

Maybe try with a 10K resistor across the meter leads so that the high input impedance of the DMM isn't picking up stray spikes.  85V is more than the raw supply rail in that PSU.

I will try that.

Some more interesting tests (all on no load):

30V: -45V / 110V
5V: -35V / 85V
1V: -32V / 60V  eeeekkk!

I just put the meter on peak mode, and turned the PSU on and off about 10 times at about 5-10 second intervals.

Another issue besides the crap fan is the humming from the transformer causes the case to vibrate very audibly. Thinking of adding some rubber padding as dampening it with my finger stops the metal rattling.

[leppie]

Tested with 10K and a 10R 10W resistors.

It was much less reproducible, and the worst overshoot was about 30%. No undershoot could be detected.

Also noted the stupid banana sockets are only 3.5mm in diameter and my nice ones wont fit   Will have to replace them too.

[McMonster]

Does anyone happen to have a schematic for 3005S version of this PSU? Current regulation is gone in my unit and I want to repair it.