AMB Laboratories DIY Audio

[RudeWolf]

The unit in question is a rev. C DGardner board Dynalo powered from a Sigma 22. Input stage has AMB matched FET's and output stage has regular unmatched bipolars. Amplifier board currently has all four biasing trimpots populated and an option to use a DC servo if needed (opamp unplugged). The R64;65 trimpots are left unpopulated because I couldn't figure out if they are needed.

When I first powered up the board there were no problems with DC on output and via trimming I could basically zero in to 1-5mV. I even decided not to use the servo. Then disaster struck- I listened to the amp and while bedazzled by the sound rushed to fight minor troubles with wiring and for a few seconds accidentally powered the amp in reversed polarity. Thorough checks with DMM showed that while input stage and current sources survived the output stage was toast. So I swapped out all BJT's and powered up the board to do some biasing.

It seems that it is impossible to get right biasing for both input and output stages. If I bias the output stage to 40mV (R23 readings) then input stage gets tortured by 12V (R2 readings), when I start with input stage to get 5V then output stage starts to cook with 400mV.

Is there a chance that something more may be damaged by miswiring judging from this?

[amb]

Do a BJT sanity check on all the non-FET transistors.

[RudeWolf]

Even if I just swapped out ALL of the BJT's?

I was about check all of the FET's in spite that R1;2 and R8;9 (same on the other channel) voltages are really symmetrical.

I did another try on biasing yesterday evening and with trimpots on 10K DC was lingering at 5V. Also the DC voltages on output fluctuate like crazy in range of 100mV. I was able to calm them down by disabling the feedback but that isn't a real solution.

I have a bad feeling that it's either the FET's or I got fake BJT's (from a brick&mortar store btw).

[amb]

Sorry, I missed the bit about you having replaced all the BJTs. If the new parts are suspicious then you should get genuine ones. The only other active parts in the circuit are the JFETs and the LEDs. Other than electrolytic capacitors, I don't think any of the passive parts would be damaged by this mishap. But the only electrolytic caps in this circuit are on the supply rails, not likely to cause the type of symptoms you describe.

[RudeWolf]

Well, I did buy them from a physical store, but I have no idea where are they getting them from. Markings looked genuine but there are some bad rumors about their sources.

In any case I plan to do a full transistor check tonight.

Btw, is there a chance that JFET's could not be completely dead from the accident but only damaged?

[amb]

Btw, is there a chance that JFET's could not be completely dead from the accident but only damaged?

Yes, I've seen some cases of degraded JFETs, although not due to polarity reversal.

[RudeWolf]

Oh well, time will tell. Most of the local guys have hard time believing that these FET's in this circuit can actually be damaged by a 15V current.

Prior to this I have talked with Spritzer and Pars and they just told me to swap everything. Maybe that would have saved me some nerves, but hey- now I know my amp inside out!

[RudeWolf]

I have bad feeling about these FET's. All of the bipolars measured ok and FET's on low bias voltages showed symmetrical values. However when higher bias currents are applied they go crazy.

Today I'll build a full fledged FET tester-


There will be trimpots on R1 and R2. That should allow me measure the full range of voltages.

[RudeWolf]

After a FET swap all seems ok. Honestly I am both happy and clueless. Undead FET's is a first.