4 hours ago, dingding123 said:

17k smd resistors are quite hard to find. Would you recommend 20k or 15k ohm resistors?

I didn`t check the schematic. However, the input resistors(Signal input) shouldn`t be a probably if you go with 15k or 20k.

Has anyone tried building this? Not sure if it's just me - after power on, the board overheats within a minute and then comes the magic smoke. PSU voltages are correct but output balance (i.e. L+ vs L-) goes from -2 to -60v quite quickly then suddenly normalises to around 4v).

Pretty sure there are no shorts or issues with soldering on my board...

i am aware of the overheating issues.  i have an assembled board coming to me to look into this.  something clearly has changed from the original parts i used and there is some positive thermal runaway that did not happen with the prototype. there is a later version board now, 4 layers, still does not fix the problem.

putting 12v zeners in series with the 1.1meg resistors may introduce enough negative thermal compensation. otherwise the bias needs to be trimmed a lot.

hate the parts situation i do.

you are aware that you need a heatsink touching all the sot transistor.  still with heatsink it typically can take about 5 minutes to reach high temps.

On 8/7/2024 at 5:16 PM, kevin gilmore said:

all three power supplies a total of $80.62 with shipping from china.

fully regulated 12v to 18v input.

i could not build these things for that price, no way.

I was thinking, Kevin, are those bricks rechargeable? Does it need a battery management board?

These are DC-DC converters 

2 hours ago, judo said:

These are DC-DC converters 

All right, thought they were batteries... So perhaps it's not really "portable", to the office yes, but around mmmm

Perhaps a little early, compiled a BOM:

https://www.mouser.de/api/CrossDomain/GetContext?syncDomains=www&returnUrl=https%3a%2f%2fwww.mouser.com%2fTools%2fProject%2fShare%3fAccessID%3dd8cfd9e908&async=False&setPrefSub=False&clearPrefSub=False

Tried a test build, with silicone thermal sheet, alu plate on some alu piece.

Yes, there's thermal runaway. Power consumption begins at 24V/0.8A and slowly takes up to almost 2A, when the board reached 80°C I shut it down.

The 400V B+ rail (so the B-) started around 9mA, up to 20mA before shutdown.

Offset was around 2V on one channel, 5V on the other one.

Amp was in idle condition.

@kevin gilmoreDr., you were saying putting a 12V sender in series to 1.1Meg resistor, newbie question, with what polarity?

What about trimming the bias ? Do you foresee adding a trimpot in the next release?

Thanks, hope it helps other fellow builders.

Unfortunately the resistors drift with temperature as I've tried building it before. I'd recommend adding a fan to cool the top side.

7 hours ago, psyduck said:

Unfortunately the resistors drift with temperature as I've tried building it before. I'd recommend adding a fan to cool the top side.

Yes, but it introduces noise 😁

8 minutes ago, Michelag said:

Yes, but it introduces noise 😁

There's an updated board on google drive that has a 10k trimmer with dn2540 that you can adjust the current till it's stable.

https://drive.google.com/file/d/16cuirLgLsadI1xa9c6OoaIqW6KevyZPa/view?usp=drive_link

2 hours ago, psyduck said:

There's an updated board on google drive that has a 10k trimmer with dn2540 that you can adjust the current till it's stable.

https://drive.google.com/file/d/16cuirLgLsadI1xa9c6OoaIqW6KevyZPa/view?usp=drive_link

You are one of the greatest, here on my side it's easier to open the mouth instead of go checking updates on the two repositories 😁😔

Hope it addresses also the thermal runaway... 

Edited October 19, 2025Oct 19 by Michelag

When adjusting the trimmers, make sure they're balanced to minimize DC offset between outputs, recommend having another DMM for that.

To measure the current, you need to measure the voltage between the 100R resistors and divide the voltage by 100. The total current for one side is the sum of all 4 of them.

never expected thermal runaway because birgir has been selling the much higher power version and its completely stable.

those particular resistors due to the heat really drift.  like 10%.  and in the wrong direction. with the resistors replaced by current sources its stable.

so much easier to make really higher power things.

What resistors are people using on this?  I did it all through hole so 50ppm but I've never had real drift issues with 1206 and 2010 parts before.  

34 minutes ago, spritzer said:

What resistors are people using on this?  I did it all through hole so 50ppm but I've never had real drift issues with 1206 and 2010 parts before.  

Same here, 1206 and 50ppm, but drifting

PBHV2160ZX seem to be out of stock at Mouser?

On 10/20/2025 at 11:27 AM, spritzer said:

What resistors are people using on this?  I did it all through hole so 50ppm but I've never had real drift issues with 1206 and 2010 parts before.  

I've used Vishay's CRCW 2010 and 2512 resistors rated at 400V. They are thick film resistors.

On 10/21/2025 at 2:46 AM, Michelag said:

PBHV2160ZX seem to be out of stock at Mouser?

Unfortunately due to political issues.

CRCW's should be just fine so something else might be drifting there.  

Sorry if I posted this in the wrong thread.

But it seems like this is also a clone of this amplifier.

You're talking about the amplifier board itself overheating, right?

That is, the resistors on it are drifting and therefore i need to be replaced with ones with greater thermal stability...

This is where the power supply failed, and I've already ordered the DC-DC converter that Kevin recommended.

I hope everything works reliably after installing them.

I initially thought that the problem with this thing was in the power supply, but now I realized that it’s not that simple...

 

 

Edited November 17, 2025Nov 17 by Omega_ELS

15 minutes ago, Omega_ELS said:

Sorry if I posted this in the wrong thread.

But it seems like this is also a clone of this amplifier.

You're talking about the amplifier board itself overheating, right?

That is, the resistors on it are drifting and therefore i need to be replaced with ones with greater thermal stability...

This is where the power supply failed, and I've already ordered the DC-DC converter that Kevin recommended.

I hope everything works reliably after installing them.

I initially thought that the problem with this thing was in the power supply, but now I realized that it’s not that simple...

 

And so, yes it is, quite unstable due to drifting.

Problem is not the PS, it's that the current draw grows slowly with time, then it begins pulsating, then flames or short.

I'm now assembling the stabilized version with dn2540 and pot, so that you can dial in the desired current, yet I'm having some issues with that, but as said, I started this morning (and now it's early afternoon).

Who sold this unit made the error of "marketing" it too early, perhaps not testing nor listening at it.

Here it's all about patience, testing, doing it better and listening.

So probably a new PS will not solve the issue.

 

4 minutes ago, Michelag said:

And so, yes it is, quite unstable due to drifting.

Problem is not the PS, it's that the current draw grows slowly with time, then it begins pulsating, then flames or short.

I'm now assembling the stabilized version with dn2540 and pot, so that you can dial in the desired current, yet I'm having some issues with that, but as said, I started this morning (and now it's early afternoon).

Who sold this unit made the error of "marketing" it too early, perhaps not testing nor listening at it.

Here it's all about patience, testing, doing it better and listening.

So probably a new PS will not solve the issue.

 

Thanks for your message.
Then I'll try building a cheap power supply with 4 DC-DC converters (i have some for 200V) and test its stability when running 24/7.
And only if it runs 24/7 will I order those power supplies. They're quite expensive at $25 each, and i need two of them, which means $50 plus bias.

Edited November 16, 2025Nov 16 by Omega_ELS

On 10/19/2025 at 11:30 AM, psyduck said:

When adjusting the trimmers, make sure they're balanced to minimize DC offset between outputs, recommend having another DMM for that.

To measure the current, you need to measure the voltage between the 100R resistors and divide the voltage by 100. The total current for one side is the sum of all 4 of them.

I guess it should be max 10mA. These bricks output max 20mA.

If we take only the right or left half, the upper amp (+) draws 10mA, that are "vacuumed" by the lower PS (-400V).

Then add other 10mA for the lower part, and you're at max acceptable draw.

I'm testing with the 10k pot, until now it drew really more than 10mA, so the PS started "pulsing" under my dim bulb tester. 

I'll keep you updated.

the power supply shown is definitely a flyback single ended switcher. which means the 2 rectifier diodes see twice the power than a push pull switcher would generate. which means its going to get hot. diodes need to be increased in power handling.  would like to see a real picture of the amp board in as high resolution as possible.