Adcom GFA-555 II with a bad channel

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello,

I know there are several threads here on this amp but many started ten or more years ago. A clean start seemed appropriate. However, I have read through most of them in large part. I picked up an Adcom GFA-555 II at GW yesterday. The guy there said it powered up fine with no pops or hums. Hmmmmmmm...... It had 149.99 on it. I took it to the test station and indeed the power light came on but so did the thermal protection. I pointed that out and offered them fifty bucks. Worth a shot I thought.

I got home and turned it on. I took readings on the dc offset at the speaker terminals and forgot to write them down but the left channel was fine but the right was high (IIRC). Opened her up and did a visual inspection and it is really clean inside. I noticed that the thermal protection no longer turned on so I put a pair of cheapo 8 ohm speakers on and a FiiO player with line level output control. I got good clean audio from the left channel and nothing from the right.

Checked the fuses next (probably should have at first) and the negative side of the right channel 7a fuse was melted but not blown. All other fuses were fine. Went to a local electronics store got a new fuse and put it in and started it up slowly with my variac and it immediately started chugging and humming so I shut it down.

I have a variac, a good DMM and a borrowed O-Scope that I do not know how to use. I have figured out how to dismantle it to get to the power transistors.

I have seen where some folks do an in-circuit test on the transistors but do not believe it to be very reliable. Is there a good way that will at least get me started on locating the fault? I have also read that usually it is one transistor that goes but have to believe that the rest that did not fully go have been damaged. At least I have one functioning channel.

I am studying the service manual and doing an eyeball trace following wires and circuits to get a sense of where things are. Interestingly, the big caps look new and are 125 v vs. 100v. And, the power TO-3 transistors are mounted on silicon/plastic thin wafer but I cannot see any thermal paste at all. Is that normal?

Last thing I fixed (with a lot of help here) was an old Nikko NA-850 and the culprits were little transistors that were hidden under wires. The symptoms there were similar to this amp.

I am willing to take out each transistor and cap and test out of circuit if needed. I learned last time that powering up too soon just blows what you just put in. I want to understand where the fault is before I replace anything.

All I hope to get at this point is a starting point. Thanks James
 
Hello. Progress report. I disassembled the amp so I could easily get to the underside of the pre-driver board. Since I had a good channel I just went side to side measuring things without power looking for differences. Q157 (2SA1210) gave me a hard short. And, Q156 had different readings that its counterpart on the good side.

I pulled both and tested (just a diode test) both and Q156 seemed okay but Q157 had a hard short - full continuity.

Would that explain the huffing/buzzing/puffing of the amp when I brought up the variac? I think it is part of the story for sure. The rest of the pre-driver board seems fine. Someone has replaced several transistors on the good side with originals. I surmise that since they are originals it had to have been some time ago. That said, I am not going to fix something that is not broken just yet.

So, I will have to find a modern equivalent for this transistor. Thoughts anyone? Thanks in advance.
 
I hope to get some guidance as I am at a crossroads it would seem. I had one good channel as described.

1. I fully tested/compared good side/bad side components. Resistors/Diodes/transistors.
2. I confirmed I had a bad 2SA1210. Ordered and installed replacements per another post and replaced both sides with transistor/complimentary transistor. Replacements were KSC3503DSTU / KSA1381ESTU
3. Given the unknown age of the electrolytics, I replaced all small electrolytics, including the four on the output board.
4. I checked for input ground to chassis ground and got 100 ohms and 99.5 ohms on the channels.
5. I checked each power transistor in circuit with the C-E short test and all passed.
6. I disconnected the smaller transistors on each side of the power boards and tested them. All fine.
7. Tested and confirmed the variable resistors for setting bias were working. Turned them down a bit - more resistance to avoid sending too much bias to the power transistors at startup. (wrong thing to do?).

I plugged in the variac and brought up the power and where I got hum/chuffing/buzzing before I now got silence. I let it sit for awhile at 90v and then slowly brought it up to 120 with no apparent issues. I let it sit for about five minutes and put my hand over it and felt a moderate amount of heat. Then, I powered down.

I put my dummy load on each speaker and began to bring the power up and it immediately started humming and chuffing.

My thoughts are that I need to go back in and pull each power transistor and test it out of circuit. PITA it is.

In studying the schematic I really thought getting the replacement for Q157 would do the trick because it was completely opened. I believe it is involved with turning on the power transistors. This coupled with the positive testing of the smaller power board transistors seemed to be enough. Clearly I am missing something.

Thoughts please?
 
Hello again. I am studying the schematic to try to figure out the topology of this amp. I am not an EE - big surprise I know - but am committed to figuring it out. Q157 was shorted and in tracing the schematic to see what it connects to I came across something I do not understand.

What does Q257 do? It does not appear to be connected to anything on the power board.

I know I am missing something because I am not the afore-mentioned EE. Thanks in advance.

James
 
Administrator
Joined 2004
Paid Member
Hi James,
You've come a long way on your own. The Fairchild (now On-Semi) devices are excellent replacements. For outputs you can use either the MJ15022/23, MJ15024/25 or MJ21193/94, MJ21195/96. The first transistor family I mentioned would probably be better if you have a choice. Buy only through approved distribution channels.

I lost all my schematics thanks to the Spora Ransomware (everyone should get "Malwarebytes", the paid version (which is pretty cheap). It stopped that one dead in it's tracks on other systems.

Once I get schematics I can try to help you out.

Old capacitors that leak invisible fluid, no I'm not kidding. The easiest way to detect that stuff is to hit the area near the capacitors with a hot soldering iron. It gives off a peculiar odour that you will remember. If you detect that stuff, you're going to have to depopulate the PCBs in the area of the capacitors (all of them) and clean the heck out of the boards and any parts you might re-use. I use Simple Green, others have used dish washing detergent (works well). Follow that up with methyl hydrate and maybe lacquer thinners. I have had to do one board 5 times to get rid of that stuff. That stopped being a profitable repair the second wash.

Testing transistors reliably can only be done with the part removed from the circuit! So you're correct there. Don't forget that resistors can go open without any obvious marks. Also make sure the traces are continuous, especially if you have had the capacitor fluid problem.

-Chris
 
Thanks Anatech

Hello,

At the risk of sounding obsequious - I was delighted to see you had responded.

I tried to attach the service manual but it is too big. So, I am attaching only the schematic. I have tried the solder iron on the boards and do not get the obnoxious smell. I recently repaired an old Carver TFM-25 that had obvious spillage so I am familiar with the smell.

There is obvious evidence someone else was in here previously and they replaced only the pre driver transistors on the good side. Since they were original specs it must have been a long while though.

I remain puzzled but will keep poking around.

I have already pulled the negative rail power transistors on the bad side and they have all diode-tested fine - indeed they look pretty new.

Is there a clue I should look for pertaining to the change caused by adding a dummy load? To me, in my little non-electronics expert attorney brain it means by inserting the dummy load I have completed a circuit that allows excess current to flow. But, all the components in that circuit seem to be testing fine. Ground issue?

I think I have checked all the ground connections.

Again, I appreciate your input and all others too. I constantly remind myself to use baby-steps.

James
 

Attachments

  • Adcom-GFA-555-II-schematic.pdf
    423.3 KB · Views: 109
Administrator
Joined 2004
Paid Member
Hi James,
Yes, connecting any kind of load to an amplifier that has a DC offset will probably cause excessive current to flow. That can confuse you by having you look for a current fault in the output stage when you actually have an offset fault in the voltage amplifier section. From the sounds of it, you have a DC offset fault.

Thank you for the schematic. It's a good, clean copy. The voltage amplifier section is that bit on the PC board.

Q157 shouldn't have been shorted. It is current limited by Q156 and provides current to Q257, which is the bias control transistor, then down through Q158, which is a constant current source (sink actually). If Q156 went defective, then I can see Q157 getting hot enough to short. If in doubt, take it out and replace it. Those parts aren't expensive anyway. Also check / replace R162 & R163. As well, check R170 & R171 (100 ohm) for opens.

A defective Q157 will certainly cause some kind of DC offset problem. In addition, the diff pair would have suffered a reverse breakdown and should be replaced with a matched pair. The closer the match is, the better the amplifier will sound and operate.

-Chris
 
Update again

I pulled the rest of the right side power transistors for the positive rail and they diode-tested fine. By diode-test I mean I test c-e/c-b/e-b/e-c/b-e/b-c. I noticed that not all of the transistors had the little tubular insulator that goes in the hole through the heatsink for the screws that hold the TO-3 transistors. And, I found this little guy pushed through a hole in the power circuit board. Pulled it and it tested fine. Hmmmm.........

Gotta go get some work done and ponder all of this.
 

Attachments

  • Transistor1.jpg
    Transistor1.jpg
    695.7 KB · Views: 251
Administrator
Joined 2004
Paid Member
Hi James,
You found Q257! Sneaky little devil was exactly where he should have been. Q257 usually only fails if you have shorted output transistors and / or drivers.

All the screws for the output transistors need insulating tubes. Better check the others too!

-Chris
 
Thanks for the tip!

Hi,

Okay, I now have a path forward. I am pleased that it looks like I do not have to replace the power resistors and save some money.

I have already tested all the components you mentioned but some of them were in circuit when I did. I just tried to pull Q156 and it was pretty brittle. Despite being gentle one of the legs fell off. Replacement recommendation?

Vintage baby! Back to Digikey I guess.

Forgive me but in your thorough reply before you mentioned a differential pair - are you referring to Q151 and 152?

Thanks.

James
 
Schematic - What I did with mine was printed out all three pages and then took a straight-edge and razor and cut it to fit into one big sheet and matched up the lines. A little Scotch tape and it is easy to work with. It ends up being about 23" x 11" but is so much easier to use with my aging eyes.
 
Just had a Wait A Second moment

Hi,

If Q157 was the original culprit and I have replaced it and its equivalent why am I getting the fault?

Not much of an issue at the moment because I have broken Q156 in removing it - poor little thing - but that indicates to me something else going on.

Would poorly adjusted variable resistors for setting bias cause the same excess current symptoms? I did move them when I tested them.
 
Found the answer I think to my question in the prior post. I just began looking around and under and over everything. Touching and poking around. Q157 which I replaced and tested for continuity on install now seemed wiggly. It must have been moved a bit because it lifted the old fragile traces underneath. I had to build a little solder bridge to reinstall it. That may be why after I replaced things it still did not work.

Sadly, in the process I broke Q156 and am now awaiting parts from Digikey.
 
Hello Chris,

Thanks for the confirmation. My schematic skills are getting better as is my grasp of the lingo. With the discovery that my Q157 replacement had lifted the trace underneath and these new parts I am hopeful. BTW, I went ahead and tested the existing Q151/152 in circuit and they were fine.

That said, that does not tell me what their respective HFE is or if they are adequately matched. Hopefully Digikey (they usually do) gets my stuff to me soon.

Thanks,

James
 
Administrator
Joined 2004
Paid Member
Hi James,
Yes, the diff pair would normally test fine. However, when the emitter - base junction is reverse biased to the point where it breaks down (about 6~7 volts), there are changes in the junction that can usually increase noise and also change the beta ( current gain). You wouldn't catch these changes in normal testing. Matching these parts is a critical thing in any amplifier because it is here that the distortion is subtracted from the original signal. So not only would DC offset suffer (vary from its calculated value, 0 usually but not always), so does the distortion the amp generates.
The differential pair should be in close physical contact - with shrink tubing - to maintain equivalent temperature?
Yes, exactly. I normally put some thermal compound in between the parts too. Not a lot.
Could the shrink tubing act as an insulator and inhibit cooling?
Yes, exactly. The diff pair should be insulated from ambient temperatures and air currents. They only run warm at most in a good design. A hot running pair would either signal a poor design or something not working right.

Matching these parts requires that they be at exactly the same temperature when the reading is taken. If you've read the other threads, you will find a discussion on this and PCBs some members have made for this test jig. Any amplifier that has gone DC needs a set of new diff pair transistors unless they are protected. Some amplifier designs do prevent the emitter - base junctions from becoming reverse based enough to break down. This is even an issue with voltage regulators and the single transistors used as an error amplifier. If the pass transistor(s) fail short, it can damage that transistor as well. Some regulators use a diff pair for the error amp, and they are subject to the same issues if they fail.

-Chris
 
Thanks yet again

Hi Chris.

I really appreciate you sharing your experience. I just got done reading the Nelson Pass DIY Op Amp article and had an "Ah Ha" moment or two on some of it but not all. I understand what you have said above - which is a minor miracle for me.

I think I actually felt some new synaptic connections happen in my brain. It is actually pretty exciting to be learning new stuff at 58 years of age for a guy with a BA in Philosophy and a law degree. Different worlds for sure.

My kids find it amusing as I research, read and study this electronics stuff but have admitted they are impressed with what I am learning. All of them have benefited because as I work through stuff I have built or resuscitated they end up with it.

I have ordered 16 resistors total to replace the 4 differential transistors and 16 transistors for Q156/106. I will try to match as close as possible. I hope to get a good transistor tester for my upcoming birthday so I can accurately compare HFE on these.

I forgot to order the tube insulators for the power resistors and wonder if it would sufficient to use shrink tube on the bolts?

Thanks,

James
 
Last edited:
Administrator
Joined 2004
Paid Member
Hi James,
Heat shrink tubing will work. It's not perfect but will do the job.

Matching transistors by hand with a transistor tester doesn't work well. To have any chance of success you need to maintain the parts at the exact same temperature. The jig takes care of that by keeping the two parts at the same temperature and the test current stable. I use tweezers with a transistor tester for my pre-sort, then use the jig for the actual match. It never ceases to amaze me how poor the matches with the meter are compared to the jig. So aside from pre-sort, the only reason I use the transistor testers these days is to test a part for go - no go, and to know what gain range a part is in.

So remember, when matching the transistors, you can't touch them, you can't breathe on them (hard not to do) and you must do it where there aren't any air currents or direct sunlight. Even a desk lamp can heat parts unevenly. You will end up with some "okay" matches, and others that aren't really matched at all. You can't tell between them. That is the reality of using a transistor checker to match. However, it is better than just grabbing a pair of them and sticking them in without any attempt to match them.

-Chris
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.