Adcom GFA-565 with very high DC offset

[Mayank]

I had picked up a pair of original, unmolested, unopened units a couple of months ago. *The former owner had kept them in storage for a few years but listed them as "in perfect condition".

The first test I did on reaching home was to check for DC offset. One had 63V and the other 40mV. I am working on the very high offset unit first.

Initial finding: There was capacitor leakage on the underside of the driver board.The 10R resistor R157 was bad. The servo IC 2A had a short on pins 4&6. Input stage transistors Q107 & Q108 (C3478/A1376) were bad. In the Voltage Gain section, transistor Q109 (A1015) had the B-E short.

I washed the board in Simple Green solution (4 times with careful scrubbing). Replaced the five electrolytics, servo IC, 10R resistor, Q107, Q108, Q109 transistors. Checked all the rest of transistors, diodes and resistors. Everything looked good and there was no bad smell from the leaked electrolyte.

On powering up, the 1.2K resistor (R154) on the negative rail on the board smoked instantaneously. I know, I should have powered up with the variac

What am I missing? Could this be due to a short on one of the output transistors?*

Any suggestions/guidance appreciated.

Thanks
Mayank

 

[AMV8]

Check the input ac voltage to the bridge rectifier and the dc output voltage of the bridge rectifier. It could be a rectifier fault or a connection to the rectifier that is faulty.
Then start up with a variac if possible.
Don

 

[Mayank]

Thanks for responding Don.

Measurements with everything disconnected except the soft start circuit:

AC input to bridge rectifier: 62.1V and 62.1V
DC output from bridge rectifier: +57.1V and -57.0V

I had disconnected the two large main psu capacitors.

Shouldn't the DC voltage be 62.1 x 1.414=87.8V?

Rgds
Mayank

 

[lanchile]

Thanks for responding Don.

Measurements with everything disconnected except the soft start circuit:

AC input to bridge rectifier: 62.1V and 62.1V
DC output from bridge rectifier: +57.1V and -57.0V

I had disconnected the two large main psu capacitors.

Shouldn't the DC voltage be 62.1 x 1.414=87.8V?

Rgds
Mayank

Yes it should be!!! But you "have to" put back those PSP caps and measure voltage again! Check for shorted output transsistors, disconnect them and measure them.

 

[Mayank]

With the two PSU caps connected, the voltages look OK at +85.2V and -85.6V. So I am ruling out bad power supply or rectifier bridge.
I am removing the two huge heat sinks to check the 10 pairs of output transistors for a short.

Mayank

 

[Mayank]

I think I found the issue (was home all day due to Hurricane Sandy).

4 of the 10 output transistors D424 had a dead short on the B-C and also had their 10R base resistors open. On the other side, all 10 of the B554 were found to be OK.

Before I go ahead and replace the bad outputs and base resistors, I would like to know why this happened and what was the reason just 4 of these were affected and not all. I think I should replace all 10 just in case.

Mayank

 

[lanchile]

I think I found the issue (was home all day due to Hurricane Sandy).

4 of the 10 output transistors D424 had a dead short on the B-C and also had their 10R base resistors open. On the other side, all 10 of the B554 were found to be OK.

Before I go ahead and replace the bad outputs and base resistors, I would like to know why this happened and what was the reason just 4 of these were affected and not all. I think I should replace all 10 just in case.

Mayank

it is not easy to say, but it can be many reasons why! if it was not pushed to the limits, shorted it out by accident,it could be just stress. even if amp has fuses as protection, that will not guarantee 100% protection for output transistors. But I suggest you to change "all" 10 output transistors. make sure you test them before and after soldering them. also after taking out all output transistors "measure" all components near by like: resistors, caps,diodes. and then put the output transistors. use the "lamp" method to plug it...just in case if there is a short somewhere.also do not connect speakers until you are sure you do not have high voltages to the speakers. Good luck!

 

[Mayank]

Update:

I replaced the D424 output transistors and all the corresponding 10R base resistors. Also, 4 of the 10 0.33R/5W ceramic emitter resistors had also opened which were replaced. I checked all the other resistors, diodes and capacitors on the output boards which were found to be OK.

On powering on with a 100W bulb in my newly made Light Bulb Limiter - I got an initial bright glow which quickly "simmered" to a faint red. Hmm, I thought this looks nice - no shorts, the output stage is good and the amp should now be OK.

However, after about 10 seconds with the lamp continuing to be faint red: FZZZ - the same 1.2K 1/4W resistor (R154) on the negative rail on the input board, smoked happily.

Before I start the test with a variac, is there any obvious component/fault that I am overlooking? Also, if I were to measure the current on the negative rail, shouldn't it be less than square root [1/4W / 1.2K] = 14.4 mA?

Thanks
Mayank

 

[weinstro]

Update:

I replaced the D424 output transistors and all the corresponding 10R base resistors. Also, 4 of the 10 0.33R/5W ceramic emitter resistors had also opened which were replaced. I checked all the other resistors, diodes and capacitors on the output boards which were found to be OK.

On powering on with a 100W bulb in my newly made Light Bulb Limiter - I got an initial bright glow which quickly "simmered" to a faint red. Hmm, I thought this looks nice - no shorts, the output stage is good and the amp should now be OK.

However, after about 10 seconds with the lamp continuing to be faint red: FZZZ - the same 1.2K 1/4W resistor (R154) on the negative rail on the input board, smoked happily.

Before I start the test with a variac, is there any obvious component/fault that I am overlooking? Also, if I were to measure the current on the negative rail, shouldn't it be less than square root [1/4W / 1.2K] = 14.4 mA?

Thanks
Mayank

Maybe take a look at the long-tailed pair that is ultimately sinking current through that resistor.

Regards,

Rob

 

[Mayank]

Thanks for the suggestion Rob.

The resistor in question R154 (1.2K) is on the negative rail for the Bias section. The Input & Voltage Gain sections have a seperate (-) and (+) supply.

Studying the schematic, I thought there could only be two reasons for the high current across R154 (a) film capacitor C117 (1uF/100V) has a short to ground or (b) diode D115 and/or R150 are short and Q119 (2SA970) has a B-C short to ground.

Unfortunately all these 4 components test OK.

What have I overlooked ?

Rgds
Mayank

 

[weinstro]

Thanks for the suggestion Rob.

The resistor in question R154 (1.2K) is on the negative rail for the Bias section. The Input & Voltage Gain sections have a seperate (-) and (+) supply.

Studying the schematic, I thought there could only be two reasons for the high current across R154 (a) film capacitor C117 (1uF/100V) has a short to ground or (b) diode D115 and/or R150 are short and Q119 (2SA970) has a B-C short to ground.

Unfortunately all these 4 components test OK.

What have I overlooked ?

Rgds
Mayank

Well, something, but I don't know for sure what.

What you say is true, but the current through the long-tailed pair formed by Q101 and Q105 are sinked though R154 via Q116.

First, check R145 (499 ohms).
Next, follow the collector of Q116. It connects to the emitters of the differential input pair Q101 and Q105. The collectors of these connect to the emitters of Q103 and Q107. In addition, there is a DC servo IC, IC101, which connects to the bases of Q103 and Q107 and also to the collectors of Q101 and Q102. Theoretically, a DC servo should be maintaining the output to have negligible DC offset.

I suspect that one or more of these devices or the resistors feeding these transistors has a problem. A short in this area could result in a larger voltage drop across R154, hence the burn. There could be a problem with the opamp and/or C108, too. I think C107 and C108 are electrolytics. Replace with a plastic film or high quality ceramic cap.

The good news is that the circuit is symmetrical push-pull, so you should be able to compare voltages and currents of Q101, Q103, Q105, and Q107 with those of Q102, Q104, Q106, and Q108.

Good luck with the diagnostics.

Regards,

Rob

 

[weinstro]

Initial finding: There was capacitor leakage on the underside of the driver board.The 10R resistor R157 was bad. The servo IC 2A had a short on pins 4&6. Input stage transistors Q107 & Q108 (C3478/A1376) were bad. In the Voltage Gain section, transistor Q109 (A1015) had the B-E short.

Some more thoughts:

I see you already took a look at least some of the input stage transistors. Pin 6 of the IC on the schematic looks like the output pin to me. If you had a short on this pin, that chip is likely dead. Unfortunately, Adcom only shows a house part number.

Given that R157 was dead, did you check the zeners D116 - D119? Also check R155 and R156.

Also, there are 3 other fusible resistors: R121 (negative feedback to ground, 10 ohm), R135 and R136 in series with base of pre-driver transistors.

Regards,

Rob

 

[Mayank]

Calling GFA-565 owners

I am still working on this

Quick question: Can anyone confirm if the red wire from the 12A (+) fuse goes to pin 10 on the input board and the corresponding red wire from pin 5 to the +tive terminal on the main capacitor?

This is how it is on the schematic, but they are reversed in my unit. I am confident that no one has worked on this unit - wonder if these wires were accidently switched on the factory assembly line (no biggie as both are +86V).

Thanks
Mayank