GSSL Output Stage Blowing Capacitor Problem

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

alisomay

Member
Joined
Mar 14, 2018
Messages
19
Hello GSSL builders. I built a GSSL for my friend a month ago, it worked at least around 12 -15 Hours before i gave him and maybe double after that.
Unfortunately i will be out of my country for months and he has a problem with the machine and i want to help him but i can not debug it from far.

The problem is about the four electrolytic caps in the output balancing stage.
He informed me that they are blown and sent me these pictures, i included them in the attachment. (How do we display images ? )

I made a guess that maybe a metal part was in the case and shorted the capacitor and he gave it to a technician and he changed the electrolytics.

The day after it happened again with the new caps.


Where or what problem in the circuit you would look for and start deebuging in this situation?
I really appriciate a brainstorm.
Thank you very much.

(Polarities are right  : ) )
 

Attachments

  • debug.jpg
    debug.jpg
    602.5 KB · Views: 43
The voltages need to be checked at the caps and double check at the IC pins. Somehow there is excessive power being routed there, and with this being a self etch board the possibilities are fairly numerous. Unless your friend knows how to troubleshoot this will more than likely something you will have to deal with in person when you return home.

Thanks!

Paul
 
I will assist my friend to check voltages at the caps but, i can not imagine in a circuit like this that that voltages could exceed cap's maximum ratings. If you are talking about the voltage becoming more positive in the - side of the capacitor than the + side, as you say it is a good idea to check this also.

What would be the expected voltages in capacitor legs?

About the phantom power (it is a good idea) because this device is connected as a send effect from Yamaha O1V96i to do parallel compression. There might be a little possibility like that. I checked with my friend he would definetly answer tomorrow morning.
I will get back to you.

Thank you.
 
Apparently somehow it's getting more VDC than what it is rated for, and I'm not sure how that is possible unless if both of the rails (15V and 12V) were somehow combined that would exceed the 25V rating for the electrolytics in the output section per the schematic, although I would think that only 2V over the caps rating would take longer to make it fail and wouldn't cause them to blow up in such a fantastic manner.

At any rate, I would make sure your friend didn't do something silly like apply +48V as Jakob suggested and then measure voltages at the IC sockets and capacitor solder pads after you remove the bad capacitors.

For the record, while I was testing preamp build years ago, I managed to send phantom power to the input which was connected to the output of a summing mixer and fried those output capacitors.

Thanks!

Paul
 
Hey people thank you for the brainstorm! It was indeed the phantom power! I always thought of a mistake of the circuit but apperantly i jumped the phantom in my mind :). Also thank you gyraf for the correct advice and the design itself. I self etched and built the circuit very carefuly and measured so i trust the circuit in a way but never over trust a circuit ofcourse :).

There is a local technician that i trust he will change the caps for him.
I also told him to buy spare IC’s to change them if in case something happened to them also.

Maybe it would be a good idea to change 8 electrolytics at the input and output stage, in case if something happens like this again to hold for a while. I would tell him to change them with 100v ones if he finds a small size around.

Do you think it is a good idea?

I would also open a thread to share the picture of the device we have built.

Thanks again and have a good day.
 
I think 50V is enough to be on the safe side for the 100uF output capatitors. But you will need to orient these four electrolytics so that their +side is towards the XLR's - that is, opposite from what the schematic/pcb shows.

You can probably still manage to blow the outputs by giving it phantom power if you are unlucky enough - but then it would only be a matter of replacing your (socketed) 5532's, which is easy, quick and cheap enough...

Jakob E.
 
I couldn't properly understand the proposed orientation change of the caps. Could you explain a bit more?
I advised him to get some extra 5532's btw in case.
 
On original schematic and pcb, the output capacitors (100uF) are oriented with minus facing towards the output xlr's, and plus facing the opamp.  This polarity is chosen arbitrarily at design, as we don't expect DC across them.

What killed them was the +48V on their negative side.

When replacing these four 100uF's, use 50V types (or 63V, if they fit in there) - and REVERSE POLARITY to the opposite of what is shown in schematic/pcb-layout. This puts +side of electrolytics towards the XLR's, and helps somewhat on protection. And, as mentioned above, polarity on these was not important in the first place, so there will be no difference in performance.

Jakob E.
 
The compressor is ok now and with the modification gyraf advised it is fixed now.
Thanks everyone.
 
Back
Top