Need Help Fixing the Output on a KA-8100

[ampman]

Hi,
In the middle of restoring my KA-8100 when discovered the right channel output is basically half of the left.
I'm measuring the output over the audio spectrum (20-20k Hz).
The source input is on the AUX input @0.150mV to both R&L. This results in good voltages on the pre out terminals resulting in the same input on pins 1 and 2 into the output PCB (from the pre out/main in).
The expected output on the left channel is 1.054V@20 to 0.987V@ 20KHz.
The right channel is half of expected 0.567V@20Hz to 0.488V@20kHz.

I've traced the signal (1kHz) through the output PCB and verified voltages are within specs as spec'd on the schematic. I really don't see anything that stands out as bad yet see the drop after Qe8
Unable to obtain the expected bias voltage on test point 8 & 10. It's low.

I need help from those more knowledgeable about the Kenny line of amps from this ear.

Thanks

 

[Jose Jimenez]

When you say drop after Qe8, are you referring to a waveform using an o-scope? Bias improves gain, especially at low level signals. I'll look for your reply.

 

[sregor]

Any distortion (or clipping)?
Gain is set by Re38 and Re18.
Is there a pre/main connector switch on back? dirty switch can cause problem, and easy to overlook.
AC signal at collector of Qe8 is pretty much the same as the output.
If clipped or distorted, maybe problems with current limiter, (Qe 14-20)

 

[ampman]

"When you say drop after Qe8, are you referring to a waveform using an o-scope?"
Yes, referring to the sine wave at the output of Qe7 in reference to Qe8. See the following


Cold start up.

Volume at infinite
Balance in mid-position

Output at speaker terminals -- L/side = ~700uV -- R/side =~347uV

Bias on VR3 = 14.1mV
Bias on VR4 = 13.9mV

Bias on VR3 increased to 16.4mV over two minutes
Bias on VR4 dropped to 5.4mV over two minutes
------------
Volume reset to lower mid-range

Input to base of Qe5 = 256mV
Input to base of Qe6 = 220mV

Output @Qe7 = 344mV
Output Qe8 = 248mV

Removed both transistor pairs - both pairs tested good with basically the same hFE
-----------------
"Any distortion (or clipping)?
Is there a pre/main connector switch on back? dirty switch can cause problem, and easy to overlook."

No. No distortion. Very clean sine wave throughout the signal path and square wave is good also.
The pre/main output/input connectors are RCA type, connected by a solid metal rod bent to form a 180° U.

Kenwood switches are known to form oxidation over time, but all switches have been cleaned. Faders were cleaned & lubed also.

 

[ampman]

Sorry for the double post

 

[ampman]

Hi,

Any ideas?

 

[sregor]

Confusion.

Output at speaker terminals -- L/side = ~700uV -- R/side =~347uV

uV? should be mV? measuring with meter or scope?

Input to base of Qe5 = 256mV
Input to base of Qe6 = 220mV

Output @Qe7 = 344mV
Output Qe8 = 248mV

The signal at the bases of Qe5 and Qe6 should be very small - AC signal at collectors should be about the same as the output. Larger signal should be easier to see.
Problem is that everything is inside a feedback loop, so is not so easy to try and isolate signal by looking at different points.
First check the signal coming in at the gates of the input FET pair (Qe2) Signal should be the same on both, regardless of the overall gain. Compare the input signal with the other channel.

After looking at the schematic, I noticed something which I believe is unique to this amplifier ( a WTF moment). The gain should be set by R18 and R38, but calculates out to be about 40, vs. the gain diagram which show 26 dB (gain of about 20). Then noticed a second 22K ohm resistor, Re40, which goes back to the protect relay. SO...My take is that the low channel is correct, and bad relay contact (or wiring issue to protect PCB) is not connecting the resistor (Re39) after the relay engages, causing higher gain. No idea why they did it this way. Never seen it done this way in other amps ((Kenwood or other companies). Right channel is okay. Left is too high gain. Simple way to check would be to unsolder one side of R40, and see if gains match (both too high) Good luck.

 

[ampman]

Sregor,
Thanks for the reply.

"uV? should be mV? measuring with meter or scope?"
The volume control was set to minimum for the first set of measurements. The second set of measurements are taken with the volume control set to 30-40% of max.
The output measurements are from an Audio Precision Sys1 unit using the data editor list.

“First check the signal coming in at the gates of the input FET pair (Qe2) Signal should be the same on both, regardless of the overall gain. Compare the input signal with the other channel.”
If I remember correctly the signal coming in is the same. Compared to the other channel it’s quite close.

It will be a day or two before I am able to make the measurements you suggest. I’ll post results then.

Thanks again for helping me learn.

 

[sregor]

 

[ampman]

Hi all,
First -- Thanks for your help. Appreciate it.

Progress is stopped for the present. I removed R39 & R40 to measure them. One was open and the other was off spec. Replaced them both.

One channel operated and had output, the other did not.

Traced the issue to the IC which is bad. Not producing output. Removed them both, swapped them and the lower output followed the device.

Have ordered replacements from Japan.

Will update as I can.