Kenwood KA-94 repair

[ricothetroll]

Hi !

I took this hi-fi amplifier out of my basement after it has been there for a few years. When switching it on the first time, I heard the protection circuit relay clicking, indicating that things were going OK in the power amp, at least from a protection circuit point of view.

I left it on during an hour or so while I was fixing the speaker I wanted to connect it to, and then when I finally connected everything the amp was not working anymore. And there is no relay click anymore after switching it on and off again.

I got the service manual and checked a few DC voltages. It appears that the bottom channel (on the joined schematic) has correct voltages, while the top channel has a wrong biasing of the whole driver/power amp circuit: the base voltage of Q13, Q15, Q17 and Q19 if around -17 VDC instead of the expected +/- 1.2V and +/- 0.6V.

I have desoldered and tested all the transistors and diodes highlighted in green on the joined schematic, but they all tested good. I'm running out of ideas...

Would anyone here have a clue about what could have gone wrong in that amp ?

Thanx in advance !

Best regards

Eric

 

[JohnRoberts]

look at q9 and q11 they should be pulling up the pre driver stage.

JR

 

[ricothetroll]

Hi JR,

Thanks for your answer !

I did suspect this stage but I have tested Q9 and Q11 and they are good. I also checked that their base voltage is the same as on the "good" channel and it is.

Or would it be possible that the transistors are good on the multimeter (Vbe = Vce ~ 0.6V) but their DC gain is gone bad ?

Best regards

Eric

 

[JohnRoberts]

Looking a little closer q11 looks like a cascode stage so q9 is doing the work, q11 just passes the current through.

Does the q5-q7 long tail pair (LTP) look in equilibrium.... if amp output is pegged low q5 may be cutting off (base lower V than q7).

Generally the node voltages will reveal what isn't right.

JR

 

[ricothetroll]

Hello,

I finally found some time to do some further measurements on this amplifier. I added the measures on the joined pdf file as comments.

• The only major voltage difference starts at Q11 collector (-18V instead of +1.2V).
• The voltages on the resistors R107, VR1 and R27/29/31 seem coherent: none of them is open or shorted.
• The voltages accross the drivers base series resistors (R35/37 and R36/38) does not show an excessive current leak through the bases of Q13/15 and Q14/16
• D5/7/41 have been tested good so they shouldn't interfere with the bias voltage

My conclusion is that Q9 or Q11 (or both) do not provide enough current through R107, VR1 and R27/29/31 to make the voltage rise to 1.2V. What I do find strange though is that I pulled them out and tested them good. Could it be possible that they are not 100% faulty but only lost their ability to drive enough current to bias the circuit properly ?

Thanx in advance !

Best regards

Eric

 

[abbey road d enfer]

There's something weird right from the start. Ca. -110 mV on the bases of both Q1 & Q3.
If the amp is pegged down at -20V, it should reflect as about -4V on the base of Q3.
And 110mV on the base of Q1 is not correct. It should be a fraction of mV.
Have you checked C5 & C1?

 

[JohnRoberts]

I still have the same questions as last time...

what is status of base drive to q9 telling it to pull up (q11 is just a cascode stage passing the q9 collector current through it).

If q9 is not being told to turn on, check status of LTP (long tail pair) Q7 and Q5.

Node voltages for all those devices could help understand what is or isn't happening.

JR

 

[ricothetroll]

Hello,

Thanx a lot for answering me !

@JohnRoberts I actually put all measured voltages on the previous pdf as pdf comments which should display on mouseover, but here is a version with the voltages directly on the image. Vbe(Q9) = 0.7 V so Q9 should be on (BTW same voltage on Q10 the "working" channel).

@abbey road d enfer those negetive voltages on the bases of the differential stage is weird indeed, though it is the same on the "working" channel. I'll remove those caps and check for leakage.

Best regards

Eric

 

[abbey road d enfer]

]@abbey road d enfer[/USER] those negetive voltages on the bases of the differential stage is weird indeed, though it is the same on the "working" channel. I'll remove those caps and check for leakage.

Do you confirm there's about -20V on the West side of R15?

 

[JohnRoberts]

Good question about R15... schematic says -0.114V on end connected to base Q3, presumably the other end (output voltage) is following the ballpark -20v at driver stage. If that was open the base would not be at good nominal voltage, so unclear what is going on with the output stage.

[edit so many of those voltages look correct... what is the final output voltage /edit]

JR

 

[walter]

What is going on with VR1? R107 is -20.39 @ VR1 and -19.6v @ Q43. I'd look at R25,R27,VR1 and what kind of connection to the Q43 module?

 

[JohnRoberts]

What is going on with VR1? R107 is -20.39 @ VR1 and -19.6v @ Q43. I'd look at R25,R27,VR1 and what kind of connection to the Q43 module?

Those voltages look valid/possible...

typical Vbe multiplier... normally thermally bonded to heat sink.

JR

 

[walter]

Looking at it some more, the -18.13v at J8 suggests an open downstream. Can you measure the collector lead of Q11? Or measure continuity between the collector leads of Q11 and Q43, should be close to zero.