Golden Age Comp54 compression "sticking"

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Sisu MT

Well-known member
Joined
Jun 25, 2015
Messages
81
Location
Austin, TX
When I turn power on, full signal is passing through with the compressor engaged and it acts as expected. When I feed signal into the unit (a clap)... The signal is attenuated and the gain reduction meter moves to reflect this as expected. The problem begins at this point, as the volume gets stuck in the low position that it was attenuated to, and the gain reduction meter stays stuck at the highest point of the transient. So basically the compressor is stuck in compression and doesn't reset back to a null point. Does that make sense? Then if I clap again, the compression will kick in and it'll max out the gain reduction and the meter will be pinned all the way to the right. I've done a fair bit of troubleshooting... I've got another one of these so I tested voltages on many of the components on the back to see where there were major differences... I did find a few places. I'm not sure if that means those components have issues or if those are merely symptoms of a different issue. I've swapped the input transformers from one unit to the other but that didn't make any difference. I've also tested all of the diodes and they are all working fine. I know this is a complicated circuit and troubleshooting from a distance is next to impossible, so I'm simply asking if any of you have seen a similar response from a compressor before And if so were you able to diagnose the issue? Thanks!
 
That looks to me like the discharge path of the timing capacitor is not connected. Check the connections.
A schemo (or a link to one) would help. I don't know what a gap54 is, and I guess I'm not alone...
 
abbey road d enfer said:
That looks to me like the discharge path of the timing capacitor is not connected. Check the connections.
A schemo (or a link to one) would help. I don't know what a gap54 is, and I guess I'm not alone...

The GAP54 is a chinese made clone of the 2254.  I believe that the company is Swedish & they out source the manufacture of their range.  A lot of their issues are down to low quality construction & components ..........
 
abbey road d enfer said:
That looks to me like the discharge path of the timing capacitor is not connected. Check the connections.
A schemo (or a link to one) would help. I don't know what a gap54 is, and I guess I'm not alone...

Thanks for the thoughts. I used "GAP". This is short for Golden Age Project. Updated this in the title. Been looking for weeks for a schemo online, but no luck. Have emailed them a couple times with no reply. Seems they are only interested in selling products and not supporting them once they are in the field. As Rob mentions, this unit is a clone of the 2254, so using a diode bridge to reduce gain. Pic attached (note that one of the input xformers has been removed in this pic). I'm actually pretty impressed with the sound (the other one, which is working). Imparts a very nice saturation when pushed. ive triple checked that all the wiring connections are plugged in (molex connectors) but its possible one of the wires broke loose from the wiring harnesses, so I'll test them for continuity. Capacitor wise, there are electrolytics, tants, and polypropylene. I'll study the traces and pull and test some of the caps. Would you expect the discharge capacitor to come right after the diode bridge?
 

Attachments

  • PXL_20210422_025533504.jpg
    PXL_20210422_025533504.jpg
    895.9 KB · Views: 48
Sisu MT said:
Would you expect the discharge capacitor to come right after the diode bridge?
No, the diode bridge is the gain cell.
The timing capacitors are in the BA185 or BA191 board.
In each of these module, there is a 4-transistor circuit that drives the timing cap.
Since the GAP uses a very different construction you have to find clues, for example the trimmer labels, if any.
Does this phenomenon appears in the compressor or the limiter? That would help trace the problem.
 
abbey road d enfer said:
Does this phenomenon appears in the compressor or the limiter? That would help trace the problem.
Thanks. This comp left off the limiter (presumably to keep cost down), so it only compresses. I see there are some good DIY 2254 threads, so will search those for some learnings. I toyed with the trimmers. They are well labelled (GR, Threshold, meter etc..) and was able to change the circuit's response, but none of them improved the core issue. I'll figure out which of the transistors you're referring to and test them as well. Thanks again for the ideas!
 
Sisu MT said:
Thanks. This comp left off the limiter (presumably to keep cost down), so it only compresses. I see there are some good DIY 2254 threads, so will search those for some learnings. I toyed with the trimmers. They are well labelled (GR, Threshold, meter etc..) and was able to change the circuit's response, but none of them improved the core issue. I'll figure out which of the transistors you're referring to and test them as well. Thanks again for the ideas!
I think the problem is something disconnected, so check where the release switch is connected; the capacitor is supposed to be connected directly to it. I believe an ohmmeter is your best friend there.
 
Thanks again. to be clear, when you say "release switch" are you referring to the potentiometer that sets the release speed or something else?
 
C33 is the timing cap. Release switch should be connected between + on the timing cap and the side of R55 that is connected to one of the diodes. At the lowest release time you should read around 6k ohm.
 
Sisu MT said:
Thanks again. to be clear, when you say "release switch" are you referring to the potentiometer that sets the release speed or something else?
On the original 2254 schemo I have, Release is a rotary switch. probably reolaced with a pot on teh GAP, cause much less$$
 
abbey road d enfer said:
On the original 2254 schemo I have, Release is a rotary switch. probably reolaced with a pot on teh GAP, cause much less$$
Thx understood. Actually I think these are rotary switches upon closer inspection. They have detents, a solid click when turned. Feels sturdy like switches on higher end gear.
 

Attachments

  • PXL_20210428_012954752.PORTRAIT.jpg
    PXL_20210428_012954752.PORTRAIT.jpg
    2 MB · Views: 23
Sisu MT said:
Thx understood. Actually I think these are rotary switches upon closer inspection. They have detents, a solid click when turned. Feels sturdy like switches on higher end gear.
Well, it's labelled SW4, so it must be a switch!  ;)
Now, I would suggest you check it with an ohmmeter.
However, it actually looks like an encoder, so there may be a decoding logic somewhere; that maybe the simple logic included in a CD4051, that requires 3 adresses for 8 positions.
It's not an absolute coder, doesn't have enough pins.
Apparently it has 5 pins, which is exactly what's needed for a hexadecimal coder.
 
abbey road d enfer said:
Well, it's labelled SW4, so it must be a switch!  ;)
Now, I would suggest you check it with an ohmmeter.
However, it actually looks like an encoder, so there may be a decoding logic somewhere; that maybe the simple logic included in a CD4051, that requires 3 adresses for 8 positions.
It's not an absolute coder, doesn't have enough pins.
Apparently it has 5 pins, which is exactly what's needed for a hexadecimal coder.
Ha, good catch! There is another row of pins on the bottom side..more studying with magnifying glasses and found this to be made by Alpha...perused their site and found one that looks very similar here:http://www.taiwanalpha.com/downloads?target=products&id=222. Pins 4, 5, 9 connect to R55 on the main PCB (via flex cable) and pin 10 on the switch connects to the timing cap (c33).What pins would you suggest taking readings on? Am I looking for a resistance reading or capacitance? Thank-you very much!
 
Sisu MT said:
Ha, good catch! There is another row of pins on the bottom side..more studying with magnifying glasses and found this to be made by Alpha...perused their site and found one that looks very similar here:http://www.taiwanalpha.com/downloads?target=products&id=222. Pins 4, 5, 9 connect to R55 on the main PCB (via flex cable) and pin 10 on the switch connects to the timing cap (c33).
So there are only 4 pins in total, yes? Since this switch has 8 positions, if I believe the pictures on the website, that means these 8 positions are binary-coded-decimal, which needs a decoding logic, that is probably in a CD4051 or 74HC4051. Inspect the circuit for a chip with 4051 in its name. Check the voltages on pins 9, 10 & 11; these are the address lines. If these are correct, the chip is probably duff.
 
abbey road d enfer said:
So there are only 4 pins in total, yes? Since this switch has 8 positions, if I believe the pictures on the website, that means these 8 positions are binary-coded-decimal, which needs a decoding logic, that is probably in a CD4051 or 74HC4051. Inspect the circuit for a chip with 4051 in its name. Check the voltages on pins 9, 10 & 11; these are the address lines. If these are correct, the chip is probably duff.
Actually 10 pins (2 rows of 5). No IC's in this circuit. Only silicon are a couple BJT's and some PNP Transistors. Thx!
 
Tekay said:
C33 is the timing cap. Release switch should be connected between + on the timing cap and the side of R55 that is connected to one of the diodes. At the lowest release time you should read around 6k ohm.
Ok, so as an experiment I bypassed C33 and that immediately reset the timing..meter reset...volume rebounded to normal and compression even started to be working properly while C33 was bypassed. I'll try swapping it with another 10uf tant tomorrow...do you think this simply means it's a bad capacitor? Also, as far as the resistance test you mentioned above...what points should i measure from? +side of C33 and switch side of R55? Thanks!
 
Sisu MT said:
Ok, so as an experiment I bypassed C33 and that immediately reset the timing..meter reset...volume rebounded to normal and compression even started to be working properly while C33 was bypassed.
"Bypassed", you mean shorted?

what points should i measure from?
the discharge path from the transistor's emitter, is a diode, a 6.8k resistor, the switch and one of the selected resistors that go to ground. If you measure between the wiper of the switch and ground, you should see various continuities, depending on the switch position.
 
abbey road d enfer said:
"Bypassed", you mean shorted?
the discharge path from the transistor's emitter, is a diode, a 6.8k resistor, the switch and one of the selected resistors that go to ground. If you measure between the wiper of the switch and ground, you should see various continuities, depending on the switch position.
I'm not sure which is the right term. i ran a bare wire across the leads of C33. Would that considered shorting the circuit? Thanks!
 

Latest posts

Back
Top