TGM8 - my best amplifier, incredible bass, clear highs, no fatigue (inspired by Rod Elliot P3a)

[Bigun]

Strange, not clear to me the issue yet. I will wait for your further test results. I guess you already checked for bad solder joints or other connections.

 

[pinnocchio]

Voltage on emitter of Q2 is -8.8Vdc, voltage on D1 is -12.1Vdc

Bad solder join checker through magnifier glass done prior to first power up.

 

[Bigun]

Are these voltages stable ? we're looking for the source of the drift. If these voltages are solid then the problem is elsewhere.

 

[pinnocchio]

ok, voltage drift after 15 minutes for D1 is -0.1Vdc and for Q2 -0.2Vdc and has not moved from 5 minutes and 10 minutes read out.

Could it be that VR2 25K pot has become damaged and is causing the adjustment issues? Could there be excessive current going through VR2 on hard clipping the amp?

Thanks
Do

 

[Bigun]

I can't think of a mechanism for the potentiometers to be damaged, an over-current at the output stage shouldn't results in excess current through the potentiometers. It could have harmed the drivers and perhaps the VAS. But I can't see how any of the dc-offset circuitry should have been damaged.

The voltage drift on Q2 was over 15 minutes, much longer timescale than the dc-offset drift you saw on the output (you posted it was 150mV in a minute or so) which indicates the problem is not with Q2 perhaps. Did you replace D1 ?

Could C3 be damaged I wonder ?

It is difficult to debug this! the simplest is to replace more and more parts until it's solved but this is inefficient. I feel like I could be more helpful if I had the board on my bench and could do more measurements and try things myself. I'd be happy to do it if you were to send me the board but I am on travel the last week of Oct and 1st week of Nov which would mean no progress at all for that time. But I offer my services if you like ?

 

[pinnocchio]

Hi Bigun,

I would gladly accept your offer. I would send you the module and the removed components. I removed the SSR Mosfets and put in a bypass and I removed Q17 to shutdown the protection circuit. Do you want me to re-install them before sending? I replaced Q1 and Q2. Lost a pad on Q1 but it was not connected underneath the PCB so no consequences on that side. I won't be sending the output Mosfets to you since I decided to not take my time when I removed them and just cut the legs... I don't have spares right now but that won't stop you from testing the amp anyways.

You can PM me your mailing address and also how much for checking the module.

Thanks
Do

 

[Bigun]

Hi Bigun,

I would gladly accept your offer.

OK, I'll send a PM. There is no cost (and no guarantee!). I just wish to be helpful. Don't re-install any parts.

Will try to ensure you have it for xmas 🙂

 

[pinnocchio]

OK, I'll send a PM. There is no cost (and no guarantee!). I just wish to be helpful. Don't re-install any parts.

Will try to ensure you have it for xmas 🙂

Can I ship it without the heatsink? It is quite heavy with it.

Thanks
Do

 

[Bigun]

Yes, sure you can, I have a suitable heatsink all ready and waiting for it.... although I won't have time to do much until I'm back from some business trips (early next month). Please also clarify in a list what you have replaced so I know which parts should be guaranteed undamaged and that you got them from a trusted source.

Also, are there parts which you have no spares, e.g. the speaker protection solid state relay FETs were damaged - probably we have to obtain some new ones, please confirm what is missing. I will be ordering some stuff from Digikey for my pre-amp project and could easily add a couple of FETs to the list.

 

[pinnocchio]

Hi Bigun,

You can order a set of IRFP240/9240 and the mosfets for the SSR protection. Q1 and Q2 have been replaced, Q17 been removed.

Let me know how much to send you for the parts when you get them.

I will pack things hopefully this week but with my traveling it could be next week.

Also, all parts are genuine. I only order from mouser and Digi-Key and the likes, never use eBay or AliExpress for electronic parts.

Thanks a million!
Do

 

[Bigun]

OK, and just to be clear on how the damage was caused: the amplifier was working fine, playing music into a fairly normal speaker and the bias and dc-offset were both stable. Then at some high volume setting the amplifier failed - music stopped playing and the speaker protection warning light came on. You discovered that the drains / collectors of the power devices had a low resistance to the heatsink. You replaced the isolating thermal pads which cleared this 'short'. Afterwards the dc-offset adjustment was very-sensitive and you could not adjust it to zero, the offset also drifted, the speaker disconnect relay was blown-shot but the bias current was stable. This is the summary as I remember it. Does it seem correct ?

Did you have an output inductor in use at the time of failure ?
Remind me/confirm the voltage of the zener diode in the speaker protection circuit D5 ?

 

[pinnocchio]

OK, and just to be clear on how the damage was caused: the amplifier was working fine, playing music into a fairly normal speaker and the bias and dc-offset were both stable. Then at some high volume setting the amplifier failed - music stopped playing and the speaker protection warning light came on. You discovered that the drains / collectors of the power devices had a low resistance to the heatsink. You replaced the isolating thermal pads which cleared this 'short'. Afterwards the dc-offset adjustment was very-sensitive and you could not adjust it to zero, the offset also drifted, the speaker disconnect relay was blown-shot but the bias current was stable. This is the summary as I remember it. Does it seem correct ?

Did you have an output inductor in use at the time of failure ?
Remind me/confirm the voltage of the zener diode in the speaker protection circuit D5 ?

Correct!

Yes, had an inductor in place. D5 can't be measured in the protection circuit since Q17 has been removed and thus no more power to the circuit. I could always put it back and measure it.

Do

 

[Bigun]

for D5 no need to measure.

 

[pinnocchio]

All SMD parts on second module are soldered. Will finish it early next week 😀

Do

 

[Bigun]

 

[pinnocchio]

Module is on bench and very well behaved. DC offset holds steady at 0.4mV and bias is also stable. I'll let it sit for 1 hour then run some square waves and sine waves through it, then music.

Do

 

[Bigun]

I'm always careful with square waves - a high frequency square wave into a load can over-stress the amplifier. I use 1kHz square wave from the probe-calibration test point on my oscilloscope and avoid any thought of driving 20kHz full swing into a capacitive load or it will blow up !

 

[Mark Johnson]

I'm always careful with square waves - a high frequency square wave into a load can over-stress the amplifier.

Many of Nelson Pass's DIY project writeups include high frequency square wave testing. Here's a figure from his First Watt F5 manual.

_

 

[thimios]

During the testing period always i push hard every amplifier.

 

[thimios]

And even harder!