interesting behavior while repairing and servicing a Kenwood KM-105

[davy rocksta]

i know, i know is a very simple amplifier that might not worth the time to fix...
it's ok i'm learning a lot with this one
here is the schematic
https://elektrotanya.com/kenwood_km105_schematic.pdf/download.html#dl

ok so this unit had the output transistors shorted in just one channel, the other one is perfectly fine... it was very strange 'cause the main fuse was working fine and once i desoldered the output transistors the rele worked but it shouldn't since there was a short there... i was so confident that i only needed to change the output transistors that i just did.... AND BANG!!! the rele closed even when there was a short!!! 20 bucks to the garbage....
ok so i been doing a lot since then:

- the other short was in Q26 and Q30 "current protection" transistors (i'll try new ones as fast as i buy them)
R 34, 32 38 & 36 where burned out so i replaced them...
- i thought Q16 & Q18 "class B driver" transistors where bad, but i checked them, and they seem to work properly
- the BIAS Q14 seems to work fine as well
- since the beginning, i noticed a 3.5V at the speakers out (the good channel was very close to 0V) i tried to change a lot of things including the driver transistors, i even replaced the Q26 and Q30 with equivalents and still 3.5V at the speaker output (right now i'm about to get the clossest matches for these transistors so i'm testing without them)
- i thought those 3.5V were caused by a short so i began to take measurements live...
BUT if there is a short why the rele still closes??
ok i began to notices difference vs the working channel in Q10, Q16, Q18, Q24, Q28 and Q13 but i could't find where those 3.5V where generated
and long story short i began to measure the voltage on resistors and guess what? while i was checking the R20 a big spark showed (ofcourse, one tip was to chasis ground, and i'm sure the other tip wasn't touching other terminal) i don't know what happened so i turned off the amp, and checked the resistor and it is working fine and well as the near by components
when i turned on the amp again the rele is not working anymore (the voltage across the coil is 0V so the amp is in protect mode)

So i can't find any short, should the rele work again once i install the Q26 and Q30??

what do you recommend to do next?
and what that short might occur while i was testing the R20 if i was measuring voltage??
and why the protection mode began to work after this event and not before??
so far, this behavior is very strange to me
any help will be appreciated

 

[alschmucker]

Thought that board looked familiar I am presently working on a 209 looks like same circuitry or very close. Left channel blown on mine also. It has many cracked solder joints inspect very closely and resolder as necessary have seen bad solder joints do weird things. Will it work after replacing Q26 Q30 don't know just learning this also. Will be watching your thread, you can teach me something. After thought if there is DC on the speaker terminals would that not be a bad cap?

 

[rickl]

i've only fixed a 208 and a 209. i had some assembly glue corrode some jumpers and components. look for green fuzzy leads. I don't recall the problem with the 209 but the 208 had blown outputs.

I'd power up on a DBT and start comparing with the schematic. Also look closely at any cracked solder joints. I don't know if the 105 has a removal bottom plate.

there are several ways to troubleshoot: compare each side until you find differences, start at the speakers and work backwards, inject a signal and work forward through the signal path, ... Pick one

outputs can be difficult to find but seems like OP found a source.

 

[rickl]

I looked at the schematic link above and compared to a 209. the amp sections look identical. if you look at a 209 service manual, the schematic has voltages listed. you can use the voltages as reference when looking for the problem. since you are using a DBT (~75-100w bulb), the voltages you measure will be a bit lower than the schematic.

 

[davy rocksta]

ok guys thank you very much for your time...
i did it... i'm that kind of man that likes to do things by brute force, but still the KM-209 helped me to corroborate the values at the class B drivers... and also very important, to know where to calibrate the BIAS
so, this is what i did:
- the protection was on, so a short was there yes or yes...
- i checked out every component (but i realiced that values on Q9 vs Q10 were right, and the differences began in Q11 vs Q12), and yes, the short was there in one 100mF capacitor (C4 i think)
(i knwe about open capacitors but never about shorted capacitors i never thought the problem was there)
- also a 10 ohm resistor was burned out and with very high resistance (R81)
(i think this burned resistor caused problems with how the protection circuit worked improperly)
- i manage to get the exact type of transistors so i put them in place

i think those were all the changes i did, now the unit turns on properly and no more 3.5V in, in this case, right channel terminals i did an input test without speakers and the right channel showed more power delivery (i think this is 'cause there are no output transistors yet)
i tried to check the BIAS on the good channel, but the max value i manage to get was 3mV and that's far away from the 9 to 18mV recomended in the manual (at 0 ohms) i think i need to install the other output transistors before i check, but i'm afraid that i might change those driver transistors as well, i really hope not...

 

[davy rocksta]

ok so far the output transistors are installed (and i replace the other resistors so the unit looks symmetrical), and the unit works, but the issue now is that i noticed, even before installing the output transistors, that the power metters were not balance
once i started the testing the actual sound is not balanced but i think this issue is near the signal input...
and i noticed that the voltage in C1 and C2 are different 0.08 vs 0.110 while it should be 0V i already replace those caps but the values remain the same
and at R5 and R6, where ther's the 18V rail, the voltage that goes from the resistors to the Q1Q3 pair emisors and Q2Q4 pair emisors is 0.67V and 0.71V while it should be 0.6V
i don't know what should i do now, but to remove the resistors and transistors and check them out

 

[davy rocksta]

i think those 0.08V and 0.110 might be due a transistor leakage?

 

[elnaldo]

Unbalanced signals in old amps is often caused by differences at the volume potentiometer. You can measure the signal level at the potentiometer input and output. Or set it to MAX to discard a difference ar the potentiometer traces.

 

[davy rocksta]

hi elnaldo, this unit is a power amp, so it does not have tone controls nor a volume/gain knob

i realice that there ir leakage in the driver transistors Q17 and Q15, but they are in good condition (i replace them with similar models, it presents the same behavior) and also the bias transistor seems to work fine so i don't know what is causing that leakage, maybe caps 15&16 or a faulty resistor, or the current protection transistors...

i tested the unit without output transistors and the metters exhibit the same issue, so those transitors are working fine

still if tested with non efficient speakers/headphones both channels sounds the same is a very tiny difference that can be fixed with the preamp balance knob, but still i want to fix that issue

 

[davy rocksta]

and well i don`t know but i'm getting heating issues and i think is due to the 2SC3419 which is in the BIAS temperature compensation, how could i check that out?

 

[Oldsansui441]

and well i don`t know but i'm getting heating issues and i think is due to the 2SC3419 which is in the BIAS temperature compensation, how could i check that out?

Hi, if your bias is fluctuating and/or hard to set and you suspect the 2SC3419, a suitable replacement would be a KSC2690. I have had a few bias temp compensating transistors faulty that will test fine out of circuit but start to act up under real conditions in circuit.

A BD139 should also be suitable.

 

[davy rocksta]

Hi, if your bias is fluctuating and/or hard to set and you suspect the 2SC3419, a suitable replacement would be a KSC2690. I have had a few bias temp compensating transistors faulty that will test fine out of circuit but start to act up under real conditions in circuit.

A BD139 should also be suitable.

oooh that's nice to know, i did a test with my heating iron and i think once the temperature was high enough the rele didin't close, but i didn't saw any pops or distortion doing this... it seems that the problem was that i used the computer's DAC output directly, and also using a DSP to simulate a loudness effect.... very bad idea, that way the input signal was saturated and i think those pops were actually a capacitor discharging into the line...
once i used another amp as a preamp that problem stoped and i was doing a lot of noise with this thing (also balanced the outputs maybe that was part of the issue here), really decent audio quality

 

[davy rocksta]

ok, so... my friend tells my that it's ok if it is working fine, he doesn't care about that issue with the unbalanced channels but.... i do
here are my readings, i think the problem should be within those transistors... and BTW i think the driver transistors are working fine maybe that leakage i read hwen they are in place is the way they should work
i wonder if i replacer all initial state transistors (1 to 12) fix the issue... who knows?

i hope the image is understandable