I've have an GFA-555ii that has the dreaded DC output on one of the channels. Took out one of my JBL 4412a's with it. I'm not sure if the output transistors are fried yet. I did replace all of the with the On Semi MJ21193G / MJ21194G in 2005 since it was too hard to find real original Toshiba's. And I replaced all the small electro caps to back then.
Worked fine for a few years until this DC issue. Part of it I think was the bias adjust trimmers being flaky which I want to replace as well.
I've read some of the older posts regarding the Sanyo 2SC2912/2CA1210 parts crapping out and causing the DC. I did replace one back in 2005 on the input board although it didn't cause the DC. I've also read about the problems with fake ones being sold now.
Echowars posted about substitute parts here in 2007: http://www.diyaudio.com/forums/solid-state/106748-adcom-555ii-part-sources.html
So the On Semi MJE15032 and MJE15033 are driver substitutes for the 2SA1380/2SC3503? But if the On Semi's are a TO-220 and not a TO-3PB will that work?
Since that post was close to five years ago the 2SA1380/2SC3503 have been superceded. What's the best pre-driver available to sub for the 2SA1210/2SC2912 combo?
I have a pair of 1/4" Vishay 2k ohm trimmers that are rated at 250v / 0.5 watts. Is this high enough? http://www.vishay.com/docs/51016/t73.pdf
My service manual doesn't say what the originals are rated at.
I'm trying to get all the parts from Mouser if possible as I'm repairing some other stuff too.
Worked fine for a few years until this DC issue. Part of it I think was the bias adjust trimmers being flaky which I want to replace as well.
I've read some of the older posts regarding the Sanyo 2SC2912/2CA1210 parts crapping out and causing the DC. I did replace one back in 2005 on the input board although it didn't cause the DC. I've also read about the problems with fake ones being sold now.
Echowars posted about substitute parts here in 2007: http://www.diyaudio.com/forums/solid-state/106748-adcom-555ii-part-sources.html
So the On Semi MJE15032 and MJE15033 are driver substitutes for the 2SA1380/2SC3503? But if the On Semi's are a TO-220 and not a TO-3PB will that work?
Since that post was close to five years ago the 2SA1380/2SC3503 have been superceded. What's the best pre-driver available to sub for the 2SA1210/2SC2912 combo?
I have a pair of 1/4" Vishay 2k ohm trimmers that are rated at 250v / 0.5 watts. Is this high enough? http://www.vishay.com/docs/51016/t73.pdf
My service manual doesn't say what the originals are rated at.
I'm trying to get all the parts from Mouser if possible as I'm repairing some other stuff too.
The MJE parts are for the driver/output amplifier, it sounds like you have the predriver/driver/output version (newer).
I think you could use the NJW series 0302/0281.
http://www.onsemi.com/pub_link/Collateral/NJW0281-D.PDF
I think you could use the NJW series 0302/0281.
http://www.onsemi.com/pub_link/Collateral/NJW0281-D.PDF
The NJW's are in TO-3 packages. The 2SC/2SA are TO-126. This is maddening trying to find subs for the Sanyo's
I did see that On- Semi Has acquired Sanyo Semiconductor. ON Semiconductor: Customer Integration
I did see that On- Semi Has acquired Sanyo Semiconductor. ON Semiconductor: Customer Integration
I saw in another post that Adcom sent these for replacement.
http://www.datasheetcatalog.org/datasheet/sanyo/ds_pdf_e/2SC3502.pdf
It looks like the Ic is lower than the originals by 40ma
http://www.datasheetcatalog.org/datasheet/sanyo/ds_pdf_e/2SC3502.pdf
It looks like the Ic is lower than the originals by 40ma
"The NJW's are in TO-3 packages"
TO-3P, for the drivers using TO-3P parts.
Can't help you with the predrivers, sorry.
TO-3P, for the drivers using TO-3P parts.
Can't help you with the predrivers, sorry.
vas sub
Take a look at the Fairchild 2sc3503/KSC3503 for your Vas sub. The part number for the 2k trimmer from Mouser is 652-3329P-1-202LF.
Take a look at the Fairchild 2sc3503/KSC3503 for your Vas sub. The part number for the 2k trimmer from Mouser is 652-3329P-1-202LF.
Take real care when trouble shooting that BIAS stage. Check ever single part in it including the thermal switch. It is a UP-62 norm closed 85C part, you should have <.1 ohm cold. You can use the 90C part if you can not locate the orginal. Make sure you reinstall the copper spring holder for the thermal switch so it sit's firm to the heat sink. Also make sure you also check Q307/257 located on the heat sinks including the wiring too and from the boards. You still may find the orginal Sanyo 2912/1210 parts if you look hard but as we all know there are fakes. The fake's for these parts are easy to spot as they are all plastic with no metal backing like the orginal Sayno's. The 1210 can be had from B&D enterprises but they are out of the 2910's as far as I know. Let us know how it works out.
Sorry, that last line should be 2912's, I think I posted that at 2 am. I have a limited supply of the 2912's in stock you can contact me through PM if you would like a few. I can also provide matched sets of the diff pair. In the event the outputs are gone make sure you replace ALL of them , not just the one's that are bad. The .22 emitter resistors you can use the Mills MR05-.22. It takes some art to form the leads just right to fit but they are very good parts. The MJ21192/93 are still around they just moved the fad oversea's from Mexico. The power base ON-Semi's are also an opt. I just finished a MKII that had been to another shop, the tech gave up on it. The amp hit the ground from 3ft, yes it was on and blasting. As a few know here it was the amp from Hell, but she now lives another day. There are a few member here that have vast experance with the Adcom line, help is only a click away. Good luck.
OK, I'm really starting to think it's the 2K bias trimmers that must be bad. Keep in mind the amp has been sitting on a shelf for over a year after it blew out the JBL $$$$$$$$. I have an old Yamaha M-80 that I virtually stole from a pawnshop schmuck and put that in it's place and forgot about the Adcom.
First I tested the six predrivers doing the ECB dance on the Input Board with the diode check on my meter (BK Precision 5390) and they all checked out. Also checked the diodes, resistors etc.
Measured the outputs on each side for shorts etc., all readings were the same both sides.
Time to breakout the Variac and slowly crank up the amp to 120VAC. Measure for DC on the channel the wasted my JBL's. Around 1.5mv, no huge DC found???? Same on the other channel.
Let the amp warm up and breakout my trusty Laser Thermometer. One side is at 95 degrees, the other is at 78. OK, obviously one is biased hotter so now I decide to check the bias. On the side that "blew" and was running warmer I read about 30mv. On the other side 0.00mv.
I adjust the trimmer to the center of it's range and turn the amp back on and it's a 2mv. If I just bump the trimmer just slightly it jumps way up. And when I say just slightly I mean that when I remove the screwdriver from the trimmer it will jump.
I had adjusted the bias to 10mv a few days before it blew.
embrown057, Mouser it OOS on the 2K trimmer you recommended.
I'm going to pop one of these Vishay T73 1/4" 2K trimmers in and see if the bias is still so touchy. I just have a feeling they got cooked like the small electro caps C104/C154 did from the predrivers. I might take you up on the Sanyos. I had replaced one of them before with a substitute non Sanyo part.
I'd like to put some clip on heatsinks on the TO-126's. Couldn't hurt.
I'll report my results after the trimmer swap.
First I tested the six predrivers doing the ECB dance on the Input Board with the diode check on my meter (BK Precision 5390) and they all checked out. Also checked the diodes, resistors etc.
Measured the outputs on each side for shorts etc., all readings were the same both sides.
Time to breakout the Variac and slowly crank up the amp to 120VAC. Measure for DC on the channel the wasted my JBL's. Around 1.5mv, no huge DC found???? Same on the other channel.
Let the amp warm up and breakout my trusty Laser Thermometer. One side is at 95 degrees, the other is at 78. OK, obviously one is biased hotter so now I decide to check the bias. On the side that "blew" and was running warmer I read about 30mv. On the other side 0.00mv.
I adjust the trimmer to the center of it's range and turn the amp back on and it's a 2mv. If I just bump the trimmer just slightly it jumps way up. And when I say just slightly I mean that when I remove the screwdriver from the trimmer it will jump.
I had adjusted the bias to 10mv a few days before it blew.
embrown057, Mouser it OOS on the 2K trimmer you recommended.
I'm going to pop one of these Vishay T73 1/4" 2K trimmers in and see if the bias is still so touchy. I just have a feeling they got cooked like the small electro caps C104/C154 did from the predrivers. I might take you up on the Sanyos. I had replaced one of them before with a substitute non Sanyo part.
I'd like to put some clip on heatsinks on the TO-126's. Couldn't hurt.
I'll report my results after the trimmer swap.
OK. swapped the trimmer. It's still a little jumpy when adjusting, meaning it will jump up about 0.4mv when I remove the screwdriver. But now it seems stable with -1.2mv DC offset on the right channel and 0.5mv DC on the left. I'm not sure what's going on or why it did what it did.
Well more weirdness. I put the new trimmers in, got the bias to stabilize around 10mv,
the amp was still just showing around 1.5mv on the outputs.
So I then tried the 10ohm 1/4 watt "witness resistors" mentioned in another Adcom thread across both output pairs and made some shorted RCA plugs for the input. Slowly cranked the Variac up and both resistors smoked.
Same thing without shorting the inputs. Removed the resistors and measure the outputs, back to around 1.5mv. I did note that when I had the meter hooked up to the outputs to measure DC and turned up the Variac from zero it showed several volts. As I turned it up to full the DCV dropped down to the 1.5mv.
What gives?
the amp was still just showing around 1.5mv on the outputs.
So I then tried the 10ohm 1/4 watt "witness resistors" mentioned in another Adcom thread across both output pairs and made some shorted RCA plugs for the input. Slowly cranked the Variac up and both resistors smoked.
Same thing without shorting the inputs. Removed the resistors and measure the outputs, back to around 1.5mv. I did note that when I had the meter hooked up to the outputs to measure DC and turned up the Variac from zero it showed several volts. As I turned it up to full the DCV dropped down to the 1.5mv.
What gives?
You lost me totaly with the "witness resistors". 1.5mv dc is not not bad , you should see it change from .0021mv~.0015 mv if the servo is working. What voltages do you have at the base of each set of outputs? You should have around 1.70vdc~1.77vdc. You can mesure at the input board at term #5-6 and 7-8 referance to speaker ground. This will tell us what your driving the base voltage at. I asume your not blowing any fuses at this point, just have a stedy 1.5mv at the spkr terms?
I'm not quite please with what you have if it's just staying at the 1.5mv you discribed. The servo should have it swinging not sitting at one voltage, not that I have ever seen, not in the MKII's. I have seen one of the Adcom 3A go out but not two at the same time. That realy dose not explane why you blew out a speaker, if the bias went very high the current also would have gone sky high and the 7amp fuse would have opened. It's a fast blow fuse if you have the correct one's installed. I use 6 amp in all my repairs just for insurance. The Bias pot can be a little tricky. You have to do the course agjustment @ 10mv then run the amp up to 66wrms for around 10 minuets. Let the amp settle againg then re-adjust the bias voltage. You may have to do that severial times to get it right. I always use the adjustment tool, not a screwdriver. You can damage the slot of the pot and it's never just the right fit either. The tool can be had for just a few dollars. Then after adjusting the pot I press down on it a few times to make sure it settles. Remember you have a +/- 1mv on the adjustment. Temp makes a big change also so leave the cover on while you are running the amp before adjustment.
Last edited:
This is from "anatech" from this thread:http://www.diyaudio.com/forums/solid-state/47001-adcom-555-ii-channel-failure-6.html
The DC offset does fluctuate, it's not a static reading with 1.5 being the "average"
Before when the speakers were hooked up it did blow fuses. I have 7.5a in there now. It's not blowing fuses now though. I'll measure the outputs tomorrow, too sleepy to be poking around in an amp.
I appreciate you help too, thanks.
Maybe he'll chime in on this thread and shed some light on the witness resistors.
The DC offset does fluctuate, it's not a static reading with 1.5 being the "average"
Before when the speakers were hooked up it did blow fuses. I have 7.5a in there now. It's not blowing fuses now though. I'll measure the outputs tomorrow, too sleepy to be poking around in an amp.
I appreciate you help too, thanks.
I might be mis-understand you but from what I read you installed the "witness resistors across the output transistors" I belive what Chris said is across the output spkr terms. Is that what your saying? Chris is one of the best sources of Adcom info here as he has worked on them for decades. I belive I did say to check that thermal switch also. I just had one (thermal sw) that was intermident, along with other massive problems, and the LED never came on duing that falure. As far as the voltages he said to check, he is spot on as always. Chris has been my mentor for some time now I'm proud to say. He has given me severial tools in the way of information through out the last few years, public thank you Chris. Well, if the offset is changing and it's only a few Mv. then connect a 8 ohm load , send in a 1k sine wave (50mv) start low, and take a look on your scope and see what you have. If in fact your doing this repair with the proper equiptment you can post some attachemnts of your readings for all to see. The numbers will tell us all what's going on.
O', last thing I would advise. If in fact you do find the vas or drivers bad I would not try to just de-solder and remore them, depending on how hard the amp has been run. The board get's very hot in these area's and it's very easy to have pad or trace damage. I find there is less risk of further damage by cutting the part from the transistor side, heating the connection and just pulling out the lead one at a time, with care. I belive you said you wanted to add heat sinks also, not a bad idea, some may differ. The only problem you will have with that is finding a sink 9.5mm x 19mm that fits in that area and that's cutting it close. It's late here also so I'll post a picture later of the MKII I just finshed with heat sinks installed.
Nice amp porn. Wow you really went to town on that one. Are those red caps Wima's?
Well, I don't own a scope, wish I did but I just have wing it with my meters and Variac.
I'm going to start over tomorrow and double check the 2912/1210's again and the thermal switches too like you recommended. It's strange that it's shows no huge DC with the meter hooked up but toasts those resistors pronto.
Someone mentioned in yet another Adcom 555ii thread about leaving the legs long on the Q's so they would sit up higher off the PCB to help with the heat. Mouser has some heatsinks that will work. Yep, definitely clipping them out is better.
Well, I don't own a scope, wish I did but I just have wing it with my meters and Variac.
I'm going to start over tomorrow and double check the 2912/1210's again and the thermal switches too like you recommended. It's strange that it's shows no huge DC with the meter hooked up but toasts those resistors pronto.
Someone mentioned in yet another Adcom 555ii thread about leaving the legs long on the Q's so they would sit up higher off the PCB to help with the heat. Mouser has some heatsinks that will work. Yep, definitely clipping them out is better.
Yes Wima cap's some of the Mills and Vishay/Dale and Caddock resistors. Cardas Liz wire, Cardas RCA's and some of the Multi-Caps for the by-pass and Zoble cap's. New fuse holders and fuses and the PTE wire covering.There is more, too much to list. I had to take the amp down to the to bare metel and have a body shop strighten out the left side of the case, I refinished the case in wrinke black. The amp hit the floor from 3 or 4 feet while running, poor guy! It went to a shop in Tampa for 3 weeks just to tell the guy it's not woth repairing and if I was not the type to take hard luck cases I would have told him the same. But it's more than the time involved, it's knowing you repaired an amp no one in there right mind would have done. Some of the guy's here realy don't like the Adcom line of amp's, no names and then there are the other that love them, I'm the latter. But now, this is about your amp and getting it back up and running without blowing any more speakers. Let us know what you find. ?
OK, I went back and used the diode function on my meter to test the junctions instead of the resistance. It looks like Q107,Q157 and Q108,Q158 are toast. Q105 was the one I replaced years ago with an NTE from Fry's
2SA1210
Q105
B+/C- OL
B-/C+ 0.654
B+/E- OL
B-/E+ 0.668
E+/C- OL
E-/C+ OL
Q107
B+/C- OL
B-/C+ 0.620
B+/E- 0.723
B-/E+ 0.592
E+/C- OL
E-/C+ 0.771
Q155
B+/C- OL
B-/C+ 0.651
B+/E- OL
B-/E+ 0.666
E+/C- OL
E-/C+ OL
Q157
B+/C- OL
B-/C+ 0.617
B+/E- 0.724
B-/E+ 0.621
E+/C- OL
E-/C+ 0.767
2SC2912
Q108
B+/C- 0.672
B-/C+ OL
B+/E- 0.675
B-/E+ OL
E+/C- 0.850
E-/C+ 1.118
Q158
B+/C- 0.527
B-/C+ OL
B+/E- 0.580
B-/E+ OL
E+/C- 0.850
E-/C+ 0.569
2SA1210
Q105
B+/C- OL
B-/C+ 0.654
B+/E- OL
B-/E+ 0.668
E+/C- OL
E-/C+ OL
Q107
B+/C- OL
B-/C+ 0.620
B+/E- 0.723
B-/E+ 0.592
E+/C- OL
E-/C+ 0.771
Q155
B+/C- OL
B-/C+ 0.651
B+/E- OL
B-/E+ 0.666
E+/C- OL
E-/C+ OL
Q157
B+/C- OL
B-/C+ 0.617
B+/E- 0.724
B-/E+ 0.621
E+/C- OL
E-/C+ 0.767
2SC2912
Q108
B+/C- 0.672
B-/C+ OL
B+/E- 0.675
B-/E+ OL
E+/C- 0.850
E-/C+ 1.118
Q158
B+/C- 0.527
B-/C+ OL
B+/E- 0.580
B-/E+ OL
E+/C- 0.850
E-/C+ 0.569
- Status
- Not open for further replies.