Need some help troubleshooting amp in my KR5600

jasoncauthron

Active Member
I bought this little KR-5600 in working condition, needing the tuner dial cord restrung. I checked it out briefly before I took it apart to restring it. It had some DC on the output about 65 mV on one side and 15 mV or so on the other. I listened to it for a little bit, tested out all the functions and it sounded pretty good. I didn't think it had ever been worked on before.

Upon taking it apart to restring it, I saw several 1W power resistors that had been hot (which is not unexpected) and had discolored the board below them. That power supply board looked pretty rough, actually. So I figured I'd work it over a little, replace some capacitors and resistors, put new thermal compound under the output transistors, since it seemed to run a little hot, and maybe replace those differential amp transistors at the input to try and lower the offset.

Once I got access to the main amp board, it was apparent that someone had worked on it before. There were several caps removed from the component side of the board and tacked onto the solder side. Q5, the voltage amplification transistor on the left channel had been changed to a TO-220 device with no visible numbers on it, and having a pin configuration that didn't match the original. The drivers had been at least removed and reinstalled, and might or might not be original. But, like I said, the amp worked and sounded okay to me in the brief time that I listened to it.

Since I had it all apart at this point and couldn't easily hook it up to re-evaluate it, I started replacing capacitors and all the parts that I had planned on replacing anyway. Replaced all the electrolytics, replaced the diodes and zeners. I changed both VAS transistors, Q5 and Q6 with KSC2690s and went ahead and swapped out the driver transistors too with the KSA1220/KSC2690's. For the differential inputs, I matched up pairs of KSA992's.

When I put it all back together, I fired it up on a DBT. Bulb dimmed, relay clicked and I felt pretty good about it. I checked the bias by reading mV across the output emitters: Right side was right side was just about right, reading 35mV. Left side was a little higher reading 59mV. I adjusted the left side down, stopping on 30mV to err a little on the low side. Then I checked the offset: Right side now had 10-14mV (had been 65mV :thmbsp:) Left side still read 15. Or at least that's what I thought until I tried to hook up a speaker and realized I was testing on the wrong speaker terminal. I'm 90% sure I had done that the first time too, but regardless, I actually had in excess of 200 mV on that channel... I hooked up a speaker anyway and verified that it worked. Still sounded okay, actually, though I didn't listen long and kept the volume low.

Now, of course, I don't know for certain that I didn't "cause" the problem, but I'm pretty sure that it was actually like that before I started working on it, especially seeing that that channel had been operated on before. The outputs are original and test fine. 200mV is too much though isn't it, indicating that something is actually wrong? I started checking my work, double checking cap polarity, solder joints, etc. No problems found, so I started checking voltages. This is a simple amp, but I can't figure out what's going on in the left channel. Here's what I measured, relative to chassis ground:

Q1: b) .003V c) -35.0V e) .485V b) 0.08 c) -35.0 e) .652

Q3: b) -.09V c) -35.9V e) .485V b) 0.08 c) -36.2 e) .652

Q5 b) -34.7V c) -1.38V e) -35.3V b) -34.7 c) -1.18 e) -35.1

Q7 b) .238V c) -1.38V e) .870V b) .485 c) -1.18 e) 1.13

Q9 b) .442V c) -.242V e) .868V b) .662 c) -0.014 e) 1.12

Q11 b) -.23V c) .448V e) -.240V b) .008 c) .662 e) -.014

Q13 b) -8.8V c) -.24V e) -1.38V b) -.690 c) -.015 e) -1.18

Q15 b) -.24V c) -8.8V e) -.24V b) -.035 c) -.69 e) -.015

Q17 b).875V c) 37.7V e) .32V b) 1.14 c) 37.2 e) .585

Q19 b) -1.38V c) -37.7V e) -.832V b) -1.19 c) -37.5 e) -.636

Q21 b) .32V c) 37.7V e) -.230V b) .59 c) 37.7 e) .010

Q23 b) -.83V c) -37.8V e) -.240 V b) -.635 c) -37.5 e) -.033

I think this is something simple and it may be obvious to someone who knows more than me. If anyone would be willing to take a look, here's a shot of the relevant portion of the schematic. Thanks!

Jason
 

Attachments

  • Kr5600 amp.jpg
    Kr5600 amp.jpg
    136 KB · Views: 28
Last edited:
up to now i worked out the offset is positive ..
just a thought and nothing to do with it being positive ,was just an observation ..it might be worth putting an original pair of diff transistors back in and see what happens .
 
Thanks for the observation and the reply. IIRC, though, the offset at the speaker terminals is negative... I can't check right now, but I'm pretty sure that's what I measured. It starts out around -.245mV and then settles closer to -200mV after it's been on a bit.

I am going to swap back to the original diff-input transistors. They're 2SA640's and they don't seem to be particularly well-matched or high gain according to my meter-- out of the four of them, they range from 198- 275. I thought about using BC560's for the inputs, since they're more similar to the SA640's, but they have c-b-e pinouts that would be a little more trouble to fit. Is there any reason why my sub of KSA992's would not be appropriate? Or, along those lines, are the KSA1220/KSC2690's good subs for the SA850 and SC1735 drivers? I figure they might be overkill for this amp, but better over than under, right? Would oscillation make it show an offset on one channel? If I'm going to keep doing this, I guess I need an O-scope.

On another note, it appears to me that transistors Q9, Q11, Q13, and Q15 are not essential to the function of the amp. In other words, it looks like I could remove them and the amp should still work. I'm thinking about trying that to see if it changes anything.
 
Last edited:
If your right channel is working why not compare voltages? It might help to narrow you're search before making further changes. FWIW I've had success with the BC560 in instances where the offset is a little high with original the A640s. I'm not opposed to working the pins so long as its done with thought and care. There's too few good, low noise transistors still stocked anymore to let that derail you IMO.
 
Okay- Verified that the offset at the speaker terminals on the left channel is, in fact, negative.

I went ahead and reinstalled the original diff-input transistors on the left channel and the offset actually went down a touch, but is still in excess of 200mV.

Then I removed Q9, Q11, Q13, and Q15 and tried it without them, but there was no change.

The right channel is working properly and the offset is quite low on that side, ~10-12 mV. I have checked voltages on all the right channel transistors as well, but I didn't post them. When I find the sheet where I've got them written down, I will. There are some differences, mostly on the drivers and the outputs, but I'm not able to make sense of it myself.

Thank you for the reply and the suggestions!
 
I added voltages for the right channel to my initial post. They're in red after the corresponding left-channel transistor, in base, collector, emitter format.

Hope that's not too confusing. Thanks again for any help!
 
Put the original drivers back in the left side this morning- Still no change. I don't think my transistor subs are the problem. I think there's a drifted or open resistor somewhere, or maybe bad solder joint, but I can't find it.
 
Q17 b).875V c) 37.7V e) .32V ...base is low so not driving correctly ...
Q7 b) .238V c) -1.38V e) .870V ......base is low ..
just making observations and noting them down ...
in the meantime if you could check the emitter resistors for drivers and outputs ..
also check bias pot and thermistor ..
 
A couple of shots from the outfield: To me, looks like the diff amps are running on too little current. Try changing r7 (check its value) (10k). Also there is a strange -8.8 volts on base of q13 (also q15 emitter). Perhaps something is spilled on board, and needs a good cleaning. Hard to tell from here. Good luck
 
Last edited:
That 8.8V must have been an error on my part- I think that I meant .88, as that's about what I'm seeing tonight. Sorry about that-- Didn't mean to send anyone on a wild goose chase, though I kinda sent myself on one because that's the main thing that prompted me to remove those transistors and test it without them. I reinstalled them this evening. So as of right now, the left channel has the original diff-input pair, the original driver transistors at Q17 and Q19.

Voltage at junction of R66 and R7 is 14.1 Volts, as it should be according to the schematic. R7 reads 9.8K Ohms. I removed it and replaced it with a 7.5K but it didn't change anything. Voltage on the emitters of Q1 and Q3 remained low, less than .5V

I found that R51 is shown to be 330 ohm but there was a 390 installed. Looks original. I changed it out to 330 just to see if it made a difference, but it did not.

I checked out the emitter resistors on the drivers and the outputs, and everything checks out good. Bias adjusts normally on that side. I changed out Q7, substituting a KSA992. Re-adjusted bias and maybe that caused my voltage readings to change a bit.

Q1 base= .004 collector= -35.1 emitter= .485
Q3 base= -.009 collector= -35.9 emitter= .485

Q5 base= -36.2 collector= 1.4 emitter= -36.8

Q7 base= .286 collector= -1.38 emitter= .929

Q17 base= .930 collector= 38.3 emitter= .365
Q19 base= -1.39 collector= -38.7 emitter= -.800

Q21 base= .380 collector= 37.9 emitter= -.189
Q23 base= -.821 collector= -38.2 emitter= -.225

Offset at the L-channel speaker terminals is roughly -220mV

The power amp in the KR-5600 appears to have a lot in common with the amp in the KA-3500.. Just an observation.
 
Last edited:
One more shot, then I'm off to bed. Voltages on q1 and q3 indicate (to me that is) that q1 is being shut off, and q3 is turning on harder. Shutting off q1 should decrease the base voltage on q5, but its not. (q5 pulls the output negative so reducing the drive should make the output more positive. Since it doesn't seem to be getting drive from q1, either the transistor is leaky, or cap c5 (22pF) is leaky, or something on the board (cleaning again...) ??? good luck
 
I'll throw in one more thought. It looks like the input pair is working unbalanced - specifically q1 is running at low current, but nothing is shut off or real far off. Perhaps r5 is open or out of spec. This is probably a more likely scenario than my last post. The issue seems to be that q5 base is getting more voltage than current and resistor accounts for. Good luck
 
Again, I sure appreciate the help! I honestly didn't expect this one to be very hard to track down.

I pulled C5 (the 22pF between the base and collector of Q5) and checked it. My meter probably isn't too accurate in the pF range- measured it as 40pF. It's a little ceramic disc capacitor and I happened to have a bunch of them so I replaced it (mine all read about 40 also), but it didn't have an effect on the offset or the voltages I measured.

I've been leaving the receiver on for about 5 minutes or so before I took voltage readings. This morning I checked it just a few seconds after turning it on and everything was about the same except that I noticed that the voltage on the emitters of Q1&3 was .51V, but dropped a little as the unit warmed.

I'm sure this problem existed before I ever worked on it: I just flubbed the measurement and didn't catch it. Like I've mentioned before, it works and actually sounds okay just casually listening. Listening closely I can tell that the L channel is not quite as clear as the Right and maybe a little muffled-sounding. But it really thumps on the left channel whenever the speakers are switched in. There's bound to be something wrong for there to be that much DC on the Left side. Going from threads where people were troubleshooting the KA-3500, some people suggested that the offset could be "adjusted" by varying the resistors that would correspond to R9 and R10. Should I try that? I'm feeling like I'm running out of places to look. Could the output transistors themselves cause an issue like this without them actually being "bad"? I removed them from the heatsink and I'm not certain that I reinstalled them as the same "pairs". I did check the gain on them and I think it was the NPN's that seemed rather low. Can't remember the numbers because it didn't really occur to me that it could be a problem.

The other thing I've noticed is that when I fire this thing up on the DBT, using a 100W bulb, the bulb goes completely out like there's very little current draw. I'm more used to working on higher-powered amps, so I guess this may be normal for this little amp w/ a 100W bulb.

I think I've already checked R5, but I'll check it again.
 
Last edited:
carefully check the voltages you wrote down as this looks wrong to me ...
Q19 b) -1.38V c) -37.7V e) -.832V b) -.636 c) -37.5 e) -1.19
its one thing looking at correct values and a total different thing to be working out what is written down incorrectly ..
 
You're correct. On that string of numbers, the second set is for the corresponding transistor in the right channel, for comparison, and it looks like I swapped the base and emitter numbers. Sorry for the confusion

I just pulled R5 to check it and it's 2.2K. I stuck a 1.8K in there to see if that might change anything, but it appears to have not. I re-checked all the left channel voltages just now. And double-checked my notes.;) So this is with the original 2SA640's as the input pair, a KSC2690 for Q5, the original drivers 2SA850 and 2SC1735 at Q17 and Q19 respectively, a KSA992 for the bias transistor at Q7.


Q1: b) .002 c) 34.6 e) .486

Q3: b) -.09 c) -35.5 e) .486

Q5: b) -35.1 c) -1.41 e) -35.6

Q7: b) .285 c) -1.39 e) .929

Q17: b) .940 c) 37.2 e) .368

Q19: b) -1.41 c) -37.6 e) -.816

Q21: b) .368 c) 37.1 e) -.190

Q23: b) -.811 c) -37.3 e) -.230


And some right channel numbers for comparison:

Q2: b) .08 c) -35.1 e) .659

Q4: b) .08 c) -36.5 e) .659

Q6: b) -35.2 c) -1.2 e) -35.7

Q18: b) 1.11 c) 37.1 e) .565

Q20 b) -1.18 c) -37.2 e) -.623
 
we need to get the drivers base q17 up some more to turn that output transistor q21 on ..
i need to give this some more thought ..
you say it biases up ok ? i cant really see how ...compare voltages across each emitter resistor ..you will likely see a large imbalance ..
base emitter junctions should see a 600mv drop ... so the drivers need 1.2v to get the 600mv turn on for the outputs ..
 
Last edited:
It does indeed bias up okay. And there's not actually a large imbalance across the emitter resistors. I just checked it. I'm reading the bias by measuring voltage from emitter to emitter directly at the two output transistors, because that's a lot easier to get to than the emitter resistor legs. I'm reading 37.2 mV on the left (the channel with problems) and 38.5 mV on the right (the good) side. Across the emitter resistors I'm seeing 18.4 on one, 18.7 on the other. On the "good" side I have 19.1mV and 19.3mV.

Thanks again for taking the time to look at this!
 
maybe its time to swap the outputs .. if its relatively easy or if you have some subs there all the better.
..
i just find it odd that it seems to be drawing the current ok on the outputs when that one should really be turned off ..
unless i am missing something here .:scratch2:
as a matter of interest what voltage is between the drivers emitters ?
 
One more stab - from the other side of the circuit. What kind of voltages are on r41 and r43 (both are 3.3k) Should be around 18 to 20 volts at where they meet. If low, could be problem with c29, as well as the resistors being off. These are the other side - they are pulling the voltage at the output up, which feedback through q5 should bring to normal. If either were bad, would cause output to stay negative. My only thought problem is that they should affect bias indirectly.
 
Across R49, emitter resistor for NPN driver L channel = .586V

Across R51, emitter resistor for PNP driver, L channel= .614V

Across R50 emitter resistor for NPN driver, R channel= .588V

Across R52 emitter resistor for PNP driver, R channel= .611V

Schematic shows these all to be 330 Ohm. R51 was originally a 390 Ohm. I changed it back to a 330 early on and left it, as it didn't seem to make any difference.

I am tempted to swap the outputs left-to-right and see if the problem follows them. I think I can do it by desoldering them and then removing the heatsink. If I have to pull the whole amp board again, then that's a pain. I hope they're not actually the problem... because there aren't any real good replacements. I'm sure something could be made to work, but...

I'll check those voltages on R41 and R43 shortly. I appreciate the help!

EDIT: Voltage at junction of R41 and R43 is 19.25V The end of R43 that goes to the collector of Q17 has 37.8V to chassis ground and the end of R41 that goes towards the emitter of Q9 and Q7 is .953V where the schematic says it should be 1.16V
 
Last edited:
Back
Top Bottom