Connect with us

Kenwood Receiver Heating Issue

Discussion in 'Troubleshooting and Repair' started by KilgoreCemetery, Sep 17, 2018.

Scroll to continue with content
  1. KilgoreCemetery

    KilgoreCemetery

    231
    9
    Apr 12, 2017
    I've been trying to figure out this Kenwood KR-70 receiver. The guy I bought it from said that the volume drops to zero after you run it for a while and, although it hasn't done it to me yet, there are some other signs that something is wrong with the receiver. There's one power transistor that heats up way faster and way hotter than the other three. Also, I'm getting about 100mV at the speaker connects with the volume at zero. That being said, this receiver powers on just fine and will play music, but I'd like to fix whatever is going on before it becomes a bigger problem.

    I've already spent quite a bit of time checking stuff over and making notes. I've swapped just about all of the transistors in the amp section from one channel to another with very little change. I've recapped the amp section after finding some that were way out of tolerance, and I've lifted the legs to test a bunch of resistors. I even tried marking and adjusting the trim pots. Nothing seems to affect the imbalance between the channels or prevent the one transistor from heating up faster than the others.

    The full schematic pdf is too large to upload here, but I do have an enlarged picture of the amplifier section with a lot of my notes on it. My biggest concern is that no matter what transistor I swap into the Q5 spot (not to be confused with Qe5), it gets hot within a minute of being powered on even though it has a lower base and emitter voltage than the other channel. Q4 gets warm within that same time frame, but it may just be radiating heat from Q5. The other two transistors are cool to the touch pretty much all the time. It seems like there is good heat transfer to the heatsink as well based on how hot it gets.

    The only transistors that I haven't swapped from one channel to the other are Qe1 and Qe2, which are the only ones that have near-identical voltages. All the electrolytic capacitors in the amplifier section have been replaced.

    I really don't know what else to check at this point. Could somebody give me an idea of where to look?
     

    Attached Files:

  2. Ylli

    Ylli

    268
    70
    Jun 19, 2018
    Let's start by measuring the voltage across R112, R113, R212, and R213. Put the meter + probe on the transistor emitter end of the resistor and the - on the other end of the resistor. Then let's get the voltages (measured to ground) on collector, base and emitter of Q2, Q3, Q4, Q5. Please measure these carefully - at least 3 significant figures.

    The voltage across R212 appears to be the wrong polarity. This would indicate that current is being driven in to Q4 instead of Q4 supplying current.

    Do you really have + 10 volts on the base of Qe9, or is that 0.1 volts?
     
  3. KilgoreCemetery

    KilgoreCemetery

    231
    9
    Apr 12, 2017
    These kept dropping gradually even after the receiver had been on a couple of minutes
    R112: 91.9 mV
    R113: 101.3 mV
    R212: 13.9 mV
    R213: 13.6 mV

    Q2: C = 43.6V, B = 21.36V, E = 20.81V
    Q3: C = 20.77V, B = 0.555V, E = 13.2mV
    Q4: C = 14.67V, B = 0.684V, E = 95.2mV
    Q5: C = 43.6V, B = 15.42V, E = 14.85V

    I should have mentioned in the original post that the power transistors are all 2SD180's. I'm not sure if that helps or not. I wish I knew more of how the amplifier section is supposed to work, but I've had a hard time finding those sort of explanations lately. At first glance, this doesn't look like your standard push-pull circuit, but, to be honest, I don't know if that's true or not

    Ah.. That's an error. This board has transistors with different pinouts and there are no markings for the legs. It should be:

    C = 10.65V, B = 167.8mV, E = 0.1mV
     
  4. Ylli

    Ylli

    268
    70
    Jun 19, 2018
    This is actually the same circuit as your last unit. The protection is a bit different, but the audio path and config is the same. I've seen some very good articles describing various audio amps, but I'll be darned if I can find one now. Hopefully someone will have a good link for you.

    There are some discrepancies in those voltage measurements. For example, you measured 0.014 volts across R213, but measured Q4 emitter to ground (the same points) as 0.094 volts. That is the difference between 0.014/0.47= 30 mA idle current and 0.094/0.47 = 200 mA idle current. 30 mA might be about right, but 200 mA is much too high. Might be seeing an offset due to where you are connecting for ground. Let's be sure to use the amplifier board pin 1 for our ground from now on.

    The biggest issue I see at this time is that the voltage at Q4's collector should be very near half the supply voltage. It is much lower than that. Q5 is running hot but Q4 is OK? If Q5 is being driver harder to try to bring this voltage up, that may be why it is running hot. I would suspect there is some parasitic DC load on the output line (Q4 collector, R212, C211 junction). Double check the polarity of C211. Is there a DC voltage on the output side of C211? Compare that to any DC voltage on the output side of C111.

    Monitor the voltage across R212, adjust VRe4 and see if the voltage changes (you are adjusting idle current here). Now monitor the voltage at the collector of Q4 to ground, adjust VRe2 and see if the voltage changes (here you are adjusting centering).

    Enough for now.
     
    Last edited: Sep 18, 2018
  5. KilgoreCemetery

    KilgoreCemetery

    231
    9
    Apr 12, 2017
    I apologize for the discrepancies. Some of these values are from before I replaced the capacitors. Also, as the heat builds, some of them change quite a bit. Whenever I spend a significant amount of time checking voltages and stuff like that, I have a small fan that blows on the heat sink to keep Q5 cool.

    I'll work on updating the voltages and get a complete list all at once to help rule out variables. Finding pin 1 may actually prove difficult since all the connections between boards are haphazardly soldered together on the bottom of the unit, but I could certainly ground to the same spot on the chassis every time.

    Come to think of it, I may have to recheck the voltage measurements for the resistors as well since the power transistors are on top and the resistors are on the bottom. I'll have to make sure I didn't check them in the wrong order.

    I agree about Q4's collector voltage being too low. Q5 gets hot, and Q4 is slightly warm to the touch, but I don't know if it's warming up as the heatsink warms up or if it's generating its own heat.

    I've tried swapping Q5 with Q2 and Q4 with Q3 previously. The voltages and fast warm up stay with the sockets, not the transistors themselves.

    I've actually been looking for some sort of DC load along the Q5 emitter/Q4 collector junction (go me!). I tried swapping out C211 and C111 with brand new radial capacitors, but it made no difference. After testing the original, axial, ones, I put them back since they were surprisingly good. From there I started tracing back towards Qe6 and Qe8. Stupid question.. if the isolator pad under Q4 is cracked or has a hole or something would it be able to create a small voltage drop, or would it just bridge the gap and be a 100% dead short?

    I had also adjusted VRe4 and VRe2 previously, but I don't think I actually monitored R212 since the trim pots are on the top and the resistor is on the bottom, but I did notice a change with the collector voltage of Q4. However, it never got much higher than about 15V

    I'll get you some proper, updated measurements on this and the other stuff you requested.
     
  6. Ylli

    Ylli

    268
    70
    Jun 19, 2018
    I think you swapped the reported voltages across R2xx and R1xx. If I swap the numbers back, it makes much more sense. We appear to have 30 mA idle current in the Q2/Q3 branch, and more than 200 mA idle current in the Q4/Q5 branch. 30 mA is good, 200 mA is excessive.

    Q4 and Q5 have the same current through them, but Q5 has a greater voltage drop across it so it gets hotter than Q4.

    It **may** just be out of adjustment. Let's adjust VRe2 to get 0.4 volts on the emitter of Qe4. Now monitor the voltage across R213 and try to adjust VRe4 to get a drop of 15 mV. There may be some interaction, so go back and readjust VRe2 for 0.4 volts on the emitter of Qe4 and then back to VRe4 for 15 mV. If you can't get there, report back with what you can get to.
     
  7. Ylli

    Ylli

    268
    70
    Jun 19, 2018
    Simplified schematic. Should look pretty familiar.
    Capture2.PNG

    Whoops, the resistor labeled R649 should be labeled R213
     
    Last edited: Sep 18, 2018
  8. KilgoreCemetery

    KilgoreCemetery

    231
    9
    Apr 12, 2017
    Yep. I totally goofed that up. The top of this receiver is really clean-looking, the bottom is a point-to-point mess where they only had 4 different wire colors to choose from. Let's try that numbering system again:

    R212: 91.9 mV
    R213: 101.3 mV
    R112: 13.6 mV
    R113: 13.9 mV

    The transistor voltages to ground are still accurate though.
     
  9. KilgoreCemetery

    KilgoreCemetery

    231
    9
    Apr 12, 2017
    The lowest I can get it to go with VRe2 is 496mV before it starts climbing back up. If I get VRe4 involved, I can get it down to 447mV.

    That gets Q4 collector voltage up to 16.29V
     
  10. Ylli

    Ylli

    268
    70
    Jun 19, 2018
    With VRe2 set to minimum resistance, that should take the base of Qe4 to ground, or 0.0 volts. With the base at 0.0 volts, the emitter should certainly be at 0. Again, adjusting VRe2, how low can you get the base voltage of Qe4?
     
  11. KilgoreCemetery

    KilgoreCemetery

    231
    9
    Apr 12, 2017
    Hmm.. well, I started by spraying all the trim pots with some deoxit and working them back and forth a bit. I realized that the resistance on VRe2 had a sweet spot in the middle and higher resistance on both ends of a full rotation. Now the resistance changes gradually.

    If I set VRe4 back to where it was marked previously and turn VRe2 to the lowest position, I can get the emitter of Qe4 down to 434mV. Also, now Q4 has a collector voltage of 19.41V and the power transistors seem to all be heating up at the same pace
     
    Last edited: Sep 18, 2018
  12. Ylli

    Ylli

    268
    70
    Jun 19, 2018
    Progress, but it seem that pot (VRe2) is still a bit flaky. Really should be able to get Qe4 emitter down to 0.4 volts and lower (really should be able to get it down to zero). Do you have any fixed resistors available? If we put a fixed resistor in parallel with VRe2, we will change the point that it needs to be adjusted to and maybe get it away from the flaky area. If my number crunching is right, a 15K resistor would be ideal, but 15K - 30K would be OK.

    Otherwise, you might consider replacing that pot. The 'service' manual I found is a one pager, showing just the specs and the schematic - no layout pictures or parts list.
     
  13. KilgoreCemetery

    KilgoreCemetery

    231
    9
    Apr 12, 2017
    I ended up looking at the service manual for the Kr-77. It's not identical, but it is similar.

    I have a whole mess of 13k and 24k resistors. Also, I have a KR-4400 that I can pull parts off of. I'll have to check the specs, but it has the same type of VR's.
     
  14. KilgoreCemetery

    KilgoreCemetery

    231
    9
    Apr 12, 2017
    Ok, I tested a few of the 13k resistors and found one at 13.5k. Popped it in and we now have sub-400mV adjustment with room to spare. It hasn't really warmed up yet, but the lowest measurement I've gotten is 311mV.

    At 390mV, I'm getting 21.30V on the collector of Q4 and none of the output transistors are getting hot. Seems like a win to me! I'm sure I could dial both channels in to where they need to be now.

    Thanks a lot, Ylli. I appreciate the help!
     
  15. KilgoreCemetery

    KilgoreCemetery

    231
    9
    Apr 12, 2017
    If you don't mind me asking, how did you come up with the 15K resistor solution?
     
  16. Ylli

    Ylli

    268
    70
    Jun 19, 2018
    Well, the 30K adjustable was obviously not adjusting down to zero as it should, but the higher resistance end seems to be OK (you said the voltage dropped and then started to rise again). So we want to use the high end of the VR.

    My number crunching told me that the resistance needed between Qe4 base and ground was about 7.8K. If we set the 30K adjustable to 30K and put a 15K in parallel, we get a 10K resistor. Now we can reduce the value of the adjustable and get to the required 7.8K while still in the high portion of the VR.

    Don't forget to tweak VRe4 to get the right idle current. You want 14-15 mV across R213.

    Did the supply voltage come up any? You have measured it at 43.6 volts, but the spec is 49 volts. That concerns me a bit. If it is still low, start by measuring the *AC* voltage between the red wires of the power transformer (PT). Be sure to get the secondary winding and not the one red wire on the primary. Should see around 72 VAC. !!CAUTION - 72 volts can give you a dandy shock!!
     
    Last edited: Sep 18, 2018
  17. KilgoreCemetery

    KilgoreCemetery

    231
    9
    Apr 12, 2017
    I did finally get around to setting the idle current! The supply voltage is now up to 49.3V!

    Thanks again for your help, Ylli. This thing sounds way better now too.
     
  18. Ylli

    Ylli

    268
    70
    Jun 19, 2018
    Great!
     
  19. jp55

    jp55

    1
    0
    Dec 16, 2019
    I have a Kenwood KR-70 (that had been idle for 30 years) and experienced the overheating issue. It was sounding great but after 1-2 hours, the volume would cut out for a few seconds--then return. Back of the unit got very warm. I cleaned, tried higher resistance speakers, still got the cut out. Took the mechanical solution and installed a 70mm fan at the back, on the switched outlet. Now runs cool and has no issue. Sounds great paired with speakers of the same vintage. pair_sm.jpg pair_sm.jpg o
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-