Son of Ampzilla - Rebuild & Restoration

Mine was true to the original schematic and I replaced parts based on what was in the amp.

I'm not 100% sure the changes to the schematic were factory.
 
Just received a very clean and legible set of scans of the original schematic.

Thank you nj pheonix :thumbsup:
 
Hi All:

An update on the rebuild/restoration progress.

1. Power Supply - COMPLETE. Transformer was found to be OK; replaced the diode bridge KBH2502 which had a dead short between two of its prongs and the four 1N4003 diodes in the 51V supply (1 was short). The original 10,000uF/50V main PSU capacitors were replaced with 12,000uF/50V which were almost half their size! The rebuilt power supply has the required outputs +/- 52V and +/- 40V.

2. Output Transistors daughter board - COMPLETE. Both channels had 2 of 4 output transistors short. It was an assorted cocktail - GAS, SK & ECG confirming prior repair attempts. I replaced all 8 with (shiny) new matching Motorolla MJ15003/MJ15004 devices. The two 30 ohm/2W and the 2.7 ohm/1W resistors on the output line appeared cooked. These were replaced.

3. Driver & VAS daughter board - IN PROCESS. While both the drivers pairs (2N6316/2N6318) and the VAS pairs (GAS116/GAS117) tested OK, based on the recommendations posted on a DIYaudio thread, I plan on replacing with them with Onsemi MJE15031/MJE15030 and MJE15029/MJE15028 respectively. All 15 electrolytic capacitors are being replaced with audio grade Nichicon capacitors.

One channel is complete.

Rgds
Mayank

The driver/VAS board "before"

IMG_0430.JPG

the driver/VAS board "after" the rebuild with the output board

IMG_0429.JPG
 
Are you replacing any of those potentiometers?

That is a bit further down my to do list - probably along with replacing the ceramic capacitors. Most of the multi-turn trimmers I found on Mouser had 5 mm pin spacing. The existing ones in the GAS SOA PCB are 8.5 mm. No biggie as I can drill new holes.

Current focus is to complete the Driver and VAS replacements and recap the board for the second channel. Then to set bias and calibrate DC offset.

Rgds
Mayank
 
Nice,
I forgot you were doing this build. I would of spent some time talking to about it at the M A meet. My apologies. Next time
 
It is ALIVE!

Both channel boards have been completed. All output transistors replaced along with corresponding modern drivers and VAS devices with new mica insulators and fresh heatsink compound. All electrolytics re-capped.

While the unit is still on the table with both channel PCB heatsinks unattached to the sides, I measured DC offset and Bias. Interestingly DC offset was 270mV/690mV and bias (measured across one of the 0.2 ohm emitter resistor) was 6mV/11mV.

Adjusting the DC offset trimmer I could get it down to a respectful 10mV/14mV and adjusted the bias trimmers to set the bias at 30mV.

I am going to listen to some music in this "open state" before planning next steps :)

Rgds
Mayank
 
Last edited:
This is getting good. Glad another one saved.

Thank you Blue - but ... Houston, we have a PROBLEM!

With the amplifier playing so well (even with my test bench $20 speakers), before assembling the heatsinks and PCBs back inside the chassis, I needed to test the VU meters and the lamps. Obviously none of the lamps were original and had been incorrectly substituted by truck indicator lamps of 24V - but that will be dealt with later.

The problem is only one of the VU meter works :oops: The other one shows open coil winding. There was a 22 ohm resistor in series connected to the speaker output which was burnt, probably as a result of the output transistors shorting leading to the open VU meter coil. The options I see now are:

1. Try and see if the meter coil can be repaired (highly doubtful - especially since I am no longer able to do fine soldering work with my progressive lenses)
2. Pray very hard to find a replacement meter (BT/eBay/DIYaudio swap meet)
3. Fit a different pair of meters till I eventually source an original.

I am leaning towards option #3 especially since I didn't understand why GAS made the meters with dark black acrylic faces which needed bright lamps to make them visible (bright=hot enough to melt the lamp covers).

Any advise/suggestions?

Rgds
Mayank
 
send the meter to the guy that I think EW mentioned that fixes meters. At least find that source and contact for more info to make the decision.
 
thank you for the 'blow by blow'.. don't get too many who do that for a good read. "Collin Clive" frankenstein. any thunder or fly a kite? :D

heck with the meters button it up.

I'm a bit surprised that without heatsinks not attached you didn't get thermals on the outputs. Did you run on a Variac or full power with low gain?
:thumbsup: let 'er settle in before major exercise and check over all 'balance' etc.
 
Thank you all for the suggestions and feedback!

send the meter to the guy that I think EW mentioned that fixes meters. At least find that source and contact for more info to make the decision.

I found EW's post. I'll call Instrument Meter Specialties, Lancaster CA tomorrow to find out what they recommend for the meter.

I'm a bit surprised that without heatsinks not attached you didn't get thermals on the outputs. Did you run on a Variac or full power with low gain?

Glad you found this an interesting read. The PCBs had been fixed on the heatsinks, but the heatsinks themselves not screwed on the chassis - had used several long quick connect cables to operate in the "open state". Initial power on was with different settings on the variac prior to full power.

These are original lamps. They're in series so if one goes out , they're all out.
http://www.mouser.com/ProductDetail/VCC/1891/?qs=sGAEpiMZZMsEIsGLxVc9kOQsnPKHG/1%2b
Oh, PS they do make an LED equivalent though i never tried.

The meters are a stock Modutect , don't know if there around if there not any VU. Meter that will fit the chassis will work too ,has to be a VU type for proper ballistics

Thanks for the heads up. I'm leaning towards replacing the lamps with LEDs. Almost all of the SOA pictures on the web have the lamp shields burnt/melted with the heat.

I believe this is the original incandescent lamp (type 1891 14VAC)

1891-2.jpg

I want to replace with these

s-l1600.jpg

Rgds
Mayank
 
Replacing the incandescent lamps with LEDs is a bit more involved than I originally thought. As nj pheonix stated above, the original 4 incandescent lamps were in series on 64VAC supply directly from the secondary winding of the main transformer. Also, the PCB designers did not foresee DC LED lamps being fitted in the future - the copper tracks on the PCB for lamp holders need to be carefully reconfigured for DC.

Here is the original circuit and my plan for modifying it for DC LED lamps. The 1N4003 diode in the original circuit was for reverse voltage protection for the LED indicator:

GAS Son 1001.jpg

I'll modify the PCB as under:

Lamp board revised.JPG

Comments welcome!

Rgds
Mayank
 
If this lighting circuit works, I'll be doing the same on my SOA.

Why reduce the original 10 watt resistor to 2 watts?
 
I never really played around with the idea much (just gave it some thought). First is there any reason LEDs wouldn't light on AC? I was just thinking the half cycle 60 times a second probably would light (i don't know whether you'd get a strobe effect). I like you idea BTW. My only other concern would be the dispersion pattern of the the LEDs, whether you'd get a similar lighting effect.:idea:
Oh, point of note . Should you opt for the incandescents, i've been led to believe by Mike B the the chinese lamps don't last long and recommended the PN i gave you.
 
they'll light, but it might flicker. If you want it to not flicker, stick a diode and a cap in there to turn it into DC. They don't pull enough current to justify more than a simple half wave and single cap filter rig.
 
Hello friends,
i got this amp from "barn" which means that is really in bad shape. end transistors burned, power resistors burned, drivers burned, one power cap exploded and so on. Im looking for unmodified electronic scheme. I readed here that nj phoenix send to Mayank.
@ nj phoenix
Will You be so kind and send also to me this scheme?
 
Hello,
Another question...what is working temperature of this amp? Let say, that its powered up for 4h and have speaker load of 8 ohms and without any input signal? Im asking because both channels warming up same...aprox. 50-60 degree celsious (120-140F). Output DC is 12mV on both channels.
 
I'd wait for someone else to chime in but that seems high to me.
Where are you taking temp reading?
 
Back
Top Bottom