My parts should arrive tomorrow. Should I retain the now-fixed circuit mod from a few posts back before I power up again?
I would remove the jumpers from Q101 and reinstall Q108/109 & R120 and the rest of the new front end parts. Then power it up to see what voltage you have on the bases of Q116/117. Hopefully you'll be sitting at around +/- 1V there. If not remeasure the voltage at the bases of Q108/109.
On both channels, I removed the jumpers to Q101, reinstalled the leg of R120, reinstalled Q108 and Q109 (and rematching these, as I had the match wrong previously). Then I replaced Q112 with new parts (the old parts leads were a bit mangled). Then I replaced Q110 and Q111 with KSA1381s.
In the left channel I replaced the fried R103 and R104. Then I replaced Q102 and Q103 with KSA992s matched as closely as possible. Then I replaced Q104 and Q105 with KSC1845s.
Can you remind me which reference designators I replace with the 1N4148? I don't see that P/N in the parts list.
In the left channel I replaced the fried R103 and R104. Then I replaced Q102 and Q103 with KSA992s matched as closely as possible. Then I replaced Q104 and Q105 with KSC1845s.
Can you remind me which reference designators I replace with the 1N4148? I don't see that P/N in the parts list.
Sounds like you're making some headway.
Replace D101, 102, 103, 104 with the 1N4148's. They're originally 1ss176 diodes.
Replace D101, 102, 103, 104 with the 1N4148's. They're originally 1ss176 diodes.
D101-104 replaced. I powered up and the relays clicked as normal.
Q116L base: -74.2v
Q116R base: -74.0v
Q117L base: -75.5v
Q117R base: -75.3v
So, not at +/-1v. 🙁
Q108L base: 65.27v
Q108R base: 65.28v
Q109L base: 75.0v
Q109R base: 75.0v
So, there still seems to be issues with the Q108 base voltage.
Q116L base: -74.2v
Q116R base: -74.0v
Q117L base: -75.5v
Q117R base: -75.3v
So, not at +/-1v. 🙁
Q108L base: 65.27v
Q108R base: 65.28v
Q109L base: 75.0v
Q109R base: 75.0v
So, there still seems to be issues with the Q108 base voltage.
Well that sucks! I know you're getting tired of throwing parts at that thing.
Lets step back and take a big picture look at this for a bit. The fact both channels are doing the exact same thing is still a big factor we need to think about.
Because the resistors are all measuring close to spec. I'm beginning to wonder if there's a trace somewhere that opened up. Maybe there's a small weak spot in the traces that has gone unobserved and that spot opened up (or just went to a high resistance not fully opening) under a momentary high current situation. The in-circuit voltages you posted early on appear to show the appropriate voltages. But what you may need to do is get that thing under a bright light and do a close inspection of all the traces. Also It would be prudent to actually ohm point to point in circuit. I'm especially interested in the second stage continuity. This being R109 down through Q108/109 and all the way down to the negative rail. It seems that there is no current flowing through here. Also check the continuity of q110 collector to Q114 base, this should be shorted.
Can you post some high res pictures of the different sides of one of the boards. Maybe one picture that shows one half of the board and another that shows the opposite end of the board. This may give enough detail for myself or another person to make an observation. Its always good to have some more eyes looking at a problem in case you overlooked something.
Lets step back and take a big picture look at this for a bit. The fact both channels are doing the exact same thing is still a big factor we need to think about.
Because the resistors are all measuring close to spec. I'm beginning to wonder if there's a trace somewhere that opened up. Maybe there's a small weak spot in the traces that has gone unobserved and that spot opened up (or just went to a high resistance not fully opening) under a momentary high current situation. The in-circuit voltages you posted early on appear to show the appropriate voltages. But what you may need to do is get that thing under a bright light and do a close inspection of all the traces. Also It would be prudent to actually ohm point to point in circuit. I'm especially interested in the second stage continuity. This being R109 down through Q108/109 and all the way down to the negative rail. It seems that there is no current flowing through here. Also check the continuity of q110 collector to Q114 base, this should be shorted.
Can you post some high res pictures of the different sides of one of the boards. Maybe one picture that shows one half of the board and another that shows the opposite end of the board. This may give enough detail for myself or another person to make an observation. Its always good to have some more eyes looking at a problem in case you overlooked something.
BTW there are a few fail safe resistors in this amp. Which means fusible, so they'll open if they are hit with a really high current. Most are in the output stage but there are a few in the front end. They have the designation RF if you go to the section of the manual where the parts are listed you'll see them. R103/104/109 are a few of them. I know you've checked those multiple times but keep an eye on them. If anything weird happens along the way a quick spot check to make sure one of them didn't open would be prudent.
Also pull Q139 and put it on the tester. I don't see where this could cause this issue, but you should test it regardless.
When you get to the point you're putting the output stage back together you'll need to recheck every one of those fusible resistors around that section.
Also pull Q139 and put it on the tester. I don't see where this could cause this issue, but you should test it regardless.
When you get to the point you're putting the output stage back together you'll need to recheck every one of those fusible resistors around that section.
I agree that we should take a step back at this point and question some of our assumptions. I think the problem very much might be induced by something I did incorrectly. Keep in mind that I always check for shorts or lack of connectivity with my DMM after installing a part. But of course it’s still possible that I installed something backwards, or that perhaps a silkscreen is incorrect. There are a few pads that have been damaged that I’ve rectified with jumpers. I have a desoldering gun but the pads on this thing are pretty delicate. You’ll see that in the photos. I’ll post some photos of one of the boards tomorrow during daylight. After I post the photos I can take a look at Q139, the RF resistors, and the continuity test you’ve kindly listed.
A part installed wrong is something else that I was thinking about earlier today.
What I was thinking was maybe something you pulled and reinstalled early in the process was installed backwards by accident and even though you've replaced those parts since then, they're still oriented wrong. The fact both channels are doing the same thing points to something like this, either a part installed wrong in both channels or an open circuit due to something that burnt or was damaged.
One other thing to look for is that sometimes I've see where a manufacturer will have a pair of pads next to each other on the board and to connect them they'll lay a leg of one component over to contact the adjacent pad. Its like a board error where they forgot to add the connecting trace in during the design. If you remove a part like this without noticing that the leg is supposed to be laid over then continuity is lost. Its a remote possibility, but something to be aware of when you start inspecting.
What I was thinking was maybe something you pulled and reinstalled early in the process was installed backwards by accident and even though you've replaced those parts since then, they're still oriented wrong. The fact both channels are doing the same thing points to something like this, either a part installed wrong in both channels or an open circuit due to something that burnt or was damaged.
One other thing to look for is that sometimes I've see where a manufacturer will have a pair of pads next to each other on the board and to connect them they'll lay a leg of one component over to contact the adjacent pad. Its like a board error where they forgot to add the connecting trace in during the design. If you remove a part like this without noticing that the leg is supposed to be laid over then continuity is lost. Its a remote possibility, but something to be aware of when you start inspecting.
One other thing to be aware of is don't trust the printed silkscreening on the boards. I've seen silkscreening errors where they put the polarity wrong for an electrolytic capacitor or even silkscreened the transistor in backwards. I generally take lots of pictures before I disassemble so I can go back and compare.
If you happened to take some pictures back at the start you may want to reference them.
If you happened to take some pictures back at the start you may want to reference them.
Sorry I just noticed your earlier comment about the silkscreen.
So yes I concur with you on that, don't trust the silkscreen. 🙂
I can take a look through my photos to see if I have something. There was definitely one case where the PCB didn't have a trace between two parts, and one board had a solder bridge to R113 and the other didn't. I mentioned that on one of the earlier posts.
I took some photos. Take a look, you’ll probably have some questions.
Nakamichi PA-7AII - Google Photos
Nakamichi PA-7AII - Google Photos
Nice pictures. I'm digging through them now! The traces are wider than I expected, so burning a trace shouldn't have been a problem or at least a remote problem anyway.
I can see you've had some problems with the smaller trace pads in the front end lifting. Its hard to tell but it looks like the Q112 emitter pad has lifted. It may be an optical illusion. Anyway make sure its connected to the trace.
The traces are so dark that its really hard to follow them in the tighter areas because they blend in with the green background all too well. Overall I really don't see anything. I tried to place the transistors based on what they're connected to in order to ensure they're oriented correctly and they appear to be.
Here's what I'd suggest. Try probing through it and inspecting to see if you can find anything that measures open or looks out of place as we talked about prior. If you're still coming up empty then consider packing the bare boards (well protected) in a box and shipping them to me. The plain boards minus heatsinks should pack easily into a smaller flat box and they don't weigh a lot so shipping from NM to Tx should be minimal. I've got some power supplies here I can power them up with on the for bench testing and I've got a lot of parts if needed. I won't charge you for anything as long as you're not in a hurry and fine with me tinkering with them over a week or two period after work.
I hate to admit it but sometimes there's just no replacement for having something sitting on the bench in front of you.
The traces are so dark that its really hard to follow them in the tighter areas because they blend in with the green background all too well. Overall I really don't see anything. I tried to place the transistors based on what they're connected to in order to ensure they're oriented correctly and they appear to be.
Here's what I'd suggest. Try probing through it and inspecting to see if you can find anything that measures open or looks out of place as we talked about prior. If you're still coming up empty then consider packing the bare boards (well protected) in a box and shipping them to me. The plain boards minus heatsinks should pack easily into a smaller flat box and they don't weigh a lot so shipping from NM to Tx should be minimal. I've got some power supplies here I can power them up with on the for bench testing and I've got a lot of parts if needed. I won't charge you for anything as long as you're not in a hurry and fine with me tinkering with them over a week or two period after work.
I hate to admit it but sometimes there's just no replacement for having something sitting on the bench in front of you.
Yeah, there was definitely some pad damage to Q112, but I get 0.3 ohms from each pin to the other end of its trace, so I feel pretty comfortable with it.
Thank you for your very kind offer to take a look at it. Give me a few more days to make some more measurements and I'll get back to you. I've been pretty busy lately, so it's been hard to find much time with the amp.
Thank you for your very kind offer to take a look at it. Give me a few more days to make some more measurements and I'll get back to you. I've been pretty busy lately, so it's been hard to find much time with the amp.
I checked the RF resistors in the removed right channel PCB and found something interesting. All were close to spec except for R109, which is supposed to be 75 ohms but measured 42k in-circuit and 35k with one leg lifted. With the left channel still installed, I also did an in-circuit measurement of R109 and it was 6M there. What do you think?
Aha you found it.... By that I mean the source of the lost current in the second stage.
Definitely bad. I'd just use a standard 75ohm 1/2W or 1W if it'll fit into the same location.
Definitely bad. I'd just use a standard 75ohm 1/2W or 1W if it'll fit into the same location.
Ok, unfortunately I don't have any 75 ohm 1/2W or 1W resistors on hand, so I'll have to do another order. I decided to take a look at the left channel RF resistors before I order. All of those except R109 were also ok. With the board removed, R109 was now OL.
Also, do you know why Nakamichi installed the RF resistors spaced a bit above the board? Was it for heat mitigation or something?
Also, do you know why Nakamichi installed the RF resistors spaced a bit above the board? Was it for heat mitigation or something?
Yes they usually space them above the board to allow for better convective cooling. It also keeps the board from browning as much due to heat over time.