Just picked up an Adcom GFA-5800 (broken)

mmadden28 · October 2009

megasat16 wrote: »

I am thinking it would be OK too but you really need to check and make sure the driver stages are OK. At minimum, I would check to make sure there is no high DC V (should be only a few mV) at the Gate of the MOSFET. It seems like a good start without having oscilloscope.

Is the driver stage anywhere in the schematics I have posted? or would they be in that middle section of the amp above the toroid (shared board for both left and right channels)??? Or is it the section on the board I have pictured but before the outputs?

I do have an Oscilloscope (actually I have 3

) if that makes any difference in how I approach any testing.

Without the MOSFETS or Gate resistors in place, should I still be able to measure the voltage at the gate? or would the Gate resistors at least need to be in place?

It would be nice to be able to test the thing out somehow ensuring no other issues exist prior to putting the new transistors in play if that's possible.

mmadden28 · October 2009

mmadden28 wrote: »

Because its FUN to have lots of TOOLS!!

I see that there are tons of them on ebay, but what about these new models? OK or Stay away?

http://cgi.ebay.com/Metered-Variac-Variable-AC-Output-Transformer-20-Amp_W0QQitemZ220493162905QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item33566aa999

http://cgi.ebay.com/Metered-Variac-Var-AC-Output-Transformer-5-Amp-NEW_W0QQitemZ220493162683QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item33566aa8bb

megasat16 wrote: »

...
Variac is very nice tool to have if you don't mind spending a few coins. I would get one myself I don't have one yet.

http://www.circuitspecialists.com/level.itml/icOid/9441

Find yourself a nice one and get a gift from them when you spend over $50 bucks (surely in your case).

It appeasr that the ones I linked to on eBay are the same company you linked too. :cool:

megasat16 · October 2009

ben62670 wrote: »

I am not disagreeing with what you are saying, but for people who do this once in a great while they become dust collectors. Again if the fuses are low amp pieces that cost next to nothing to use, and they don't blow with a true power on with (like flipping the power switch) then you are good. So say you do get a VARIAC and you power it up slowly I would still use lower amp fuses till truly tested good. Again you still should use some low amperage fuses for testing. Those output trannies are not cheap. Bottom line when it comes to amp repair listen to Chris(anatech) over anyone here.
Ben

I know you meant well and that's why I explained the fuse does not function the same as Variac and the light bulb in series for the DUT.

I would trust Anatech at DIY for many things. But I also have great respect for a lot of polkies here. I don't need to name them since you already know who they are too. They are highly regarded and come forward to help when needed.

DIY have many talented people like NP roaming to help. Even though I don't post there a lot, I like DIY a lot. Also, there are many people on AK with down to earth attitudes for people in need of help.

Here in CP, I trust there are a lot of highly skilled and qualified persons exist.

And I believe these lot of qualified polkies are kept quiet or have no will to speak or simply abandon being polkies. (I may have to do that one day for personal reasons.

)

megasat16 · October 2009

mmadden28 wrote: »

Is the driver stage anywhere in the schematics I have posted? or would they be in that middle section of the amp above the toroid (shared board for both left and right channels)??? Or is it the section on the board I have pictured but before the outputs?

I do have an Oscilloscope (actually I have 3 ) if that makes any difference in how I approach any testing.

Without the MOSFETS or Gate resistors in place, should I still be able to measure the voltage at the gate? or would the Gate resistors at least need to be in place?

It would be nice to be able to test the thing out somehow ensuring no other issues exist prior to putting the new transistors in play if that's possible.

I don't see the schema of driver stages you posted here before. You really should get the full Service Manual with all parts list for this amp. If you are lucky, there will be instruction for biasing adjustment in the service manual too.

If you have a scope and a probe, you can do a lot more trouble shooting than doing without it. You can observe waveforms which is very important after the amp is fixed but under testing before put it into use.

From this point on, I really want to suggest you need to read and follow Anatech Post in DIY. I followed your thread in DIY this morning and I found his suggestion to be very informative and very helpful.

You need to check every MOSFETs and put every parts (resistors, diodes) back into place. Also, you should replace the Zener with the proper replacement as NP suggested.

I wouldn't recommend you turn the amp on without all parts installed. You should install every parts in the amp before attempting to troubleshoot while the unit is power on. There are many reasons for it and Chris (Anatech) has said it already.

megasat16 · October 2009

mmadden28 wrote: »

It appeasr that the ones I linked to on eBay are the same company you linked too. :cool:

Yep! Same company. You get a gift if you order from them directly.

I wish you luck in fixing the amp successfully! It costs a lot of money for parts and a lot of frustrations to repair and tinker a bit but if you can do it successfully, the joy for your achievement is wordless.

mmadden28 · October 2009

megasat16 wrote: »

...You really should get the full Service Manual with all parts list for this amp. If you are lucky, there will be instruction for biasing adjustment in the service manual too....

Yeah, seems its not so easy to get a copy of one. I just sent an email to Adcom asking if they can provide.

I do however have a copy of the section of the SM that describes the process for adjusting the bias (attached).

mmadden28 · October 2009

Something's not right with this schematic. Its in two pages and the left does not match up with the right. Seems there are 8 connections shown on the left side, but only 7 connections on the right.
At first I though it was just an illusion because of the page scales being different (and it may still be an illusion), but i did resize them and they still aren't right.

I need to go an meter the traces out to find out what's what.

I attached the two copies. I wonder if there is a small section missing where they join.

Systems · October 2009

mmadden28 wrote: »

but only 7 connections on the right.

I see 8.

megasat16 · October 2009

I concur with GV. There are 8 connections between Schematic A and B. Also, Part A is the driver stages for the power MOSFETs. Did you post it before?

disneyjoe7 · October 2009

I see 8 also.

ben62670 · October 2009

MM did you contact Adcom about a service manual? That is where I got all mine.

mmadden28 · October 2009

Yeah, I see 8 now that I looked at it printed--I don't know its an illusion-I looked at it quite a few times but...--I'm not sure whats going on with my head--well lack of sleep is one thing....

I sent Adcom an email tonight(last night) to see if they can help with either parts or the SM.
Newark never responded about my query for matched parts. I was able to catch them in live chat tonight, but they said I had to login and ask a tech, but then they were closed again. Well maybe Monday.

Man, not even one avail on eBay.

ben62670 · October 2009

You have to call them to get a manual;) I have tried to get them via email a few times, and waited and waited. When I called it was usually in my email in 15 minutes or so.
Ben

mmadden28 · October 2009

megasat16 wrote: »

I concur with GV. There are 8 connections between Schematic A and B. Also, Part A is the driver stages for the power MOSFETs. Did you post it before?

I posted a combined image of A & B in post #51.

I'm glad to know that's the driver stage-I was really beginning to wonder what else was wrong with my head--what with the input and the output being on this same board and very few parts on the shared board in the middle. At least now I know most if not all of the parts in the driver stage are at least not shorted. I decided I am just going to replace all the transistors and Zeners in the driver stage as well-just in case-cheap enough to do, only problem is it seems the IRF610 is defunct--I'll keep looking

So the middle section, with the separate transformers for L & R channels...what's it called? What is its purpose? Is that the Power supply section?

mmadden28 · October 2009

So when all is said and done on this amp--Should I put Dynamat back in this thing? On the PC boards? On the walls?
Woudl that make a difference in an amp?
I mean the former owner had peices of dynamat and another type all over this thing. You can see some of it along the top where the lid rests, some peices on the row of blue caps on the middle section. A peice coveringt he Adcom logo in the right channel board. Several peices in the driver sections There was more but I had already removed some of it before the pic.
I'm guessing he didn't use the dynamat with the foil on the PC boards for obvious reasons.

ben62670 · October 2009

That would be your input and VAS stages to drive the output boards. The PS is located under that board. The PS is very simple in older Adcom's, Parasound's... The purpose of that board is to increase the voltage, and current capabilities of the signal to drive all those outputs. The dynamat may have dropped down onto the electronics after going through a bunch of heat cool cycles that could have caused the short in the first place.

mmadden28 · October 2009

I don't get how that's the input--the input lines go directly from the rca/xlr to the bottom left corner of the riaght channel board, and shows the same on the schematic. And all the parts on the schematic are on that board (well with the eception of the missing transistors and caps as previously mentioned) and all the parts on that board are on the schematic.

The only area in the schematic that shows the other connections is the output which goes directly to the binding posts, and the +/-85V and +/-70V sections. There has to be another part of the schematic for that middle section.

mmadden28 · October 2009

ben62670 wrote: »

That would be your input and VAS stages to drive the output boards. The PS is located under that board. The PS is very simple in older Adcom's, Parasound's... The purpose of that board is to increase the voltage, and current capabilities of the signal to drive all those outputs. The dynamat may have dropped down onto the electronics after going through a bunch of heat cool cycles that could have caused the short in the first place.

I didn't see any loose peices floating around--and the only kind he had on the inside was without foil. Besides I'd expect to see some charring where the short occured if that were the case, no?

megasat16 · October 2009

Do not attempt to swap out the working parts. I wouldn't swap out any transistors on driver boards unless I absolutely need to. They need to be checked for proper operation and leakage. If they works fine, put it the way as they came out. Remove and Check one transistor at a time and solder it back that one before removing the next one since they need to match very closely and factory has matched them. That way you'll not upset any matching transistor pairs.

If you need to replace, you'll need to match them properly. Either checking or matching transistors are better done with a good transistor tester. It'll cost more money, I know. It's not cheap buying all the tools and that's why the pro repairs are costly.

Anatech is frank and honest with you and discourages you from fixing it since there could be many things wrong. It will also be very expensive if you make a costly mistake.

There is a chance you can use this amp as a single channel amp. If this amp is dual mono amps and the transformer has two secondary windings, you can cut the wires from the secondary of the transformer at the rectifiers and tape up each ends individually. Just check which wires need to be cut before actually cutting it. You don't want to cut the transformer wires for the good channel.

I am not saying you to give it up but you have options and choices before spending money on parts and tools. It may be an easy fix with all MOSFETs replaced or it may be more than that. It's your money. Think it slow and do what you may.

It's essential to have service manual even before starting to troubleshoot this amp any further.

I can't tell what exactly is in the center portion of the amp but I am guessing there are DC protection circuits and may be part of the Voltage Gain Amplifier Stage (before the MOSFETs). There is also power supply section but I don't know wherether it uses bridge rectifiers are or it uses full wave bridge diodes.

I would trust Ben with Adcom amps since he worked on them before.

ben62670 · October 2009

mmadden28 wrote: »

I don't get how that's the input--the input lines go directly from the rca/xlr to the bottom left corner of the riaght channel board, and shows the same on the schematic. And all the parts on the schematic are on that board (well with the eception of the missing transistors and caps as previously mentioned) and all the parts on that board are on the schematic.

The only area in the schematic that shows the other connections is the output which goes directly to the binding posts, and the +/-85V and +/-70V sections. There has to be another part of the schematic for that middle section.

My bad. It is a PS for the input stages that are located on the left and right output boards. The 5800 and 5802 use separate PS for the input stages. You would definitely see some material burnt on there if it did cause a short.

ben62670 · October 2009

megasat16 wrote: »

Do not attempt to swap out the working parts. I wouldn't swap out any transistors on driver boards unless I absolutely need to. They need to be checked for proper operation and leakage. If they works fine, put it the way as they came out. Remove and Check one transistor at a time and solder it back that one before removing the next one since they need to match very closely and factory has matched them. That way you'll not upset any matching transistor pairs.

If you need to replace, you'll need to match them properly. Either checking or matching transistors are better done with a good transistor tester. It'll cost more money, I know. It's not cheap buying all the tools and that's why the pro repairs are costly.

Transistor matching is a huge pia. If it ain't broke don't fix it. If there is a fault that happened to blow the transistors instead of a short then you may end up blowing a whole bunch of new parts. I don't remember if there are any protection circuits in the 5800. I know that in the older Adcom's that I had there were none. I had a 5800, but never really played with it. I wish I did so I could be of more help. It is not easy via pictures over the internet.

mmadden28 · October 2009

megasat16 wrote: »

Do not attempt to swap out the working parts. I wouldn't swap out any transistors on driver boards unless I absolutely need to. They need to be checked for proper operation and leakage. If they works fine, put it the way as they came out. Remove and Check one transistor at a time and solder it back that one before removing the next one since they need to match very closely and factory has matched them. That way you'll not upset any matching transistor pairs.
...

I only made that decision based on Anatech's comment:

You need to be able to check for leakage currents. Heathkit IT-18 is excellent for that job, measured beta as well. To measure without a meter means that you must have a full and complete understanding of the parts used. Replacing them is less expensive. I don't normally agree with doing that (called "shotgunning"), but in your situation I don't see any other way to be sure.

But what transistors do those need to match with?
Well I'll at least do the zeners once I figure them out.

I'm not buying any parts until I get the Service Manual-But I'm still ordering a Variac in the meantime.

No matter what happens, at the very least I am definitely going to replace the known bad parts and give it a whirl.

ben62670 · October 2009

The HeathKit is for matching the transistors(mosfits in this case). I have a 121 IIRC.

disneyjoe7 · October 2009

mmadden28 wrote: »

So the middle section, with the separate transformers for L & R channels...what's it called? What is its purpose? Is that the Power supply section?

Not sure from the picture not clear enough to see clearly, but those square looking things I think could be DC to DC converters. I never seen them in an amp before but could be used for the rail voltage of 85v to the driver voltage of 70v. I'm a bit concerned over the Dynamat on the caps in the power supply. The problem with this amp shorting out could be right there if those caps are used in the 70v supply are where shorted out.

Systems · October 2009

mmadden28 wrote: »

only problem is it seems the IRF610 is defunct--I'll keep looking

The IRF 610/9610 are the driver transitors .Digi Key stock's a newer version with the a different suffix.
http://search.digikey.com/scripts/DkSearch/dksus.dll?vendor=0&keywords=irf610

There is a site called Techdiy (or somthing like that)that sell's matched jfet's/Mosfets' etc.They may be able to help sourcing and matching your Mosfet's but ofcourse at a price,but then you would only have to order exactly what you need.

Systems · October 2009

disneyjoe7 wrote: »

Not sure from the picture not clear enough to see clearly, but those square looking things I think could be DC to DC converters. I never seen them in an amp before but could be used for the rail voltage of 85v to the driver voltage of 70v.

They are transformers and looks like one for each channel to supply 85 VDC to the input and predriver stage.The big toroid supplies the 70 VDC to the output stage.

I'm a bit concerned over the Dynamat on the caps in the power supply. The problem with this amp shorting out could be right there if those caps are used in the 70v supply are where shorted out.

If you are reffering to the Dynamat across the blue caps,no shorting will occur because the cases have an insulated cover and the leads are at the bottom .On the otherhand IMO the dynamat is serving no useful purpose.

disneyjoe7 · October 2009

I don't know need a better picture of the markings on the square things to say for sure, but I can say they do make such a thing for sure.

As far as the caps tops I seen caps which are not insulated.

mmadden28 · October 2009

They are transformers. here is a snippet from the UM explaining their purpose which is in line with what GV said.:

The dynamat used inside the case (if it is Dynamat) is non conductive, like a hard opencell foam, at least not conductive to the point of shorting anyway-I'll meter it to be sure. The only foil lined dynamat is on the top where the top lid screws down and on the outside.

mmadden28 · October 2009

GV#27 wrote: »

The IRF 610/9610 are the driver transitors .Digi Key stock's a newer version with the a different suffix.
http://search.digikey.com/scripts/DkSearch/dksus.dll?vendor=0&keywords=irf610

There is a site called Techdiy (or somthing like that)that sell's matched jfet's/Mosfets' etc.They may be able to help sourcing and matching your Mosfet's but ofcourse at a price,but then you would only have to order exactly what you need.

Thanks!
it's www.tech-diy.com (looks like somebody else has techdiy.com which just says "coming soon"

mmadden28 · October 2009

OK, I got an email back from Adcom, they game me and email addy and phon to call for the Schematics and SM--I sent an email and left a VM.

I was also referred to Affiliated Electronics in NJ for Adcom repair parts--I had already sent them an email the other day.

I sent an email to Tech-DIY asking if they have what i need since they didn't have it listed separately ion their site but they do have them in some Pass amp kits.

I haven't heard back yet from Newark.

Question--just curious--since the output FETS are setup parallel, I'm assuming they chose 8 for each N and P channels to achieve a certain output power.
I think I know the answer but...I wonder if I could test with just lets say 3 N and P FETS installed, obviously not driving the output for normal use, but wouldn't that at least let me know if the input and driver circuits were functioning normally?
Or does the rest of the circuit expect 8 of each to balance out something else?

I'm just thinking like there's a 3 pack of 9240s on ebay for like $12--If there were something wrong with the driver or input circuits and these got shot , the I wouldn't be out much and I would know before killing the good matched set.

Thoughts?

Just picked up an Adcom GFA-5800 (broken)

Comments