Kenwood CS-6040 oscilloscope fault: severe lack of vertical amplification

[fezder]

Haven't used this 'scope lot, but now few days ago I noticed this annoying problem. I feed same signal on channels 1 and 2, and channel 1 shows MUCH less than channel 2. Channels 3 and 4 work as intended, so problem is only on channel 1. Error on channel 1 is so huge that I can't calibrate it to eliminate it. When feeding 1khz/40mVpp (value stated in service manual for calibration purposes), channel 1 shows only 4mVpp. Like tenth or so what should be. I don't have 50ohm termination, but confirmed with working 'scope that I indeed feed 40mVpp. Input attenuators work as intended when changing ranges. but even right after attenuators signals are way different in amplitude.

Thankfully, I do have service manual, which suggests fault lies in vertical secttion, more precisely in attenuator part. Power rails are fine (no excessive ripple).
There are couple components that had me wondering if this repair is doomed right away;
DN1901 twin J-fet in metal package
datasheet not found...
UPA101G npn transistor array
(https://www.ic72.com/pdf_file/u/105603.pdf)

Those components got me thinking, how to test them, and finding suitable replacement..
That's all info I can give now.
I can take photos in case someone wants to lend hand and stuff. But, any opinions?

 

[fezder]

I just rewrite this instead making new post...
So, here's problem, attenuated channel 1:

and now with proper XY settings, xy shows attenuation too:

 

[JimB]

Fezder, why don't you just attach the pictures to the post rather than using some third party picture hosting site which defaults to some blurb about privacy, cookies, etc and wants me to accept their various conditions.

JimB

 

[fezder]

Oh, didn't think about that. Edited as per request.

 

[fezder]

Here is the attenuator pcb. Those metal boxes are for channel 1 and 2. Those work outof circuit themselves just fine (attenuations 1:/, 1:/10, 1/100, 1/000).
However, neither did transistors or jfet showed being defective.

Better closeup.

 

[JimB]

Thanks Fezder, it is much better to have the pictures here on ETO rather than is some third party site.

The first thing that jumps out at me, the Variable gain controls on the two channels do not appear to be set the same.
The dots on the knobs are at different positions. Or is it that the knobs are not set to the same position on their shafts?

Rambling a bit and letting my mind wander.....
I am guessing that your signal source is a function generator which synthesises a sine wave from a triangle wave.
The trace on channel 2 has a distinct transition point at the crests of the sinewave.
The trace on channel 1 does not show this transition. This could be due to the lower amplitude on ch1, or, it could be that the frequency response of ch1 is acting like a low pass filter.
You are doing this test at 1kHz, what happens if you change the frequency? How is the trace at say 50 Hz ?

JimB

 

[fezder]

Variable knobs are at cal-position; in normal position. Knobs just were is wrong position since I just placed them back after putting case together.
Channel 2 needs to be calibrated again (both 1mV/5mV attenuation as well as variable attenuation/balance), since line on channel 2 moves when GND-coupling is used. But I doubt channel 2 calibration issues would crosstalk so much on channel 1. That said, it must be calibrated.
Also, according to service manual, gain is calibrated before variable balance., but good point still.
Fixed knobs now to proper place.
I did test now with 1kHz and 50Hz, no remarkable difference in performance:
1kHz/10Vpp.

50Hz/10Vpp

Due CRT slow sweep needed for 50Hz (or my skills with analog 'scope...) and camera framerate picture is what it is.
Voltage amplitude makes no difference, tried as low as 40mVpp.
As for my function generator, It's hooked like this. Not best way due improper termination.

All of my 'scopes are 1M and function generator is 50 ohm impedances.

 

[fezder]

Also, I tried with 9v battery in case this is AC problem, and problem is even with DC coupling.
And, here's schematic of this pcb problem seem to loom. I could isolate problem further, but I need to know which component to lift to prevent oscillations and magic smoke...

 

[fezder]

Slight progress; managed to narrow faulty area further. Took all transistors and that Jfet out of circuit, and attenuation in attenuation PCB is now gone.
Also, took wire going from vertical amp section to final amp to eliminate possibilty that something in that section causes havoc (turned intensity to 0 for this test just in case.

I managed to narrow with these tests fault most likely between R151 and R169. I measure both sides of this mirrored design as well as compare it to working side.
That HIC (U102) ic got me worried again. According to schematic, it only does something with variable balance, but It's just random PCB with some epoxy on top of it; something special.
EDIT: since I mentioned variable, variable knob signal 0-1Vdc works same way in both channels. This signal comes to that HIC (U102) via 10ohm resistor.
HOWEVER, variable volts/div, doesn't work at all on channel 1, whereas it works in channel 2.

 

[fezder]

OK, now I tested variable knob function, in same places where I ended attenuation search earlier; Channel 2 variable volts/div knob functions as intended, but channel 1 is locked. In fact, channel 1 is as much attenuated now as channel 2 variable knob fully used.

 

[fezder]

Even more indication now that variable balance in channel 1 is broken/stuck. Now, variable balance trimmer (on pcb for calibration purposes; VR104) doesn't effect variable attenuation at all, but waveform vertical position. Trimmer that controls vertical position is after that (VR503, left in schmetic from ch1 position)

Also, already tested transistors in problem area that they are working correctly, out circuit.

 

[fezder]

Yup, HIC (U102) pin 1 is stuck in very small amplitude. Due this fact now known, I took that HIC (U102) out of circuit since it seems to act greatly with variable attenuation. And, that did the trick! According to schematic, it fuctions as some sort of error apmlifier, or something like. Even package suggests some sort of that.
This is now situation, obviously I can't calibrate variable balance now since HIC is out.

And this is the HIC:

Which seems like package that cannot be repaired....any thoughts about making own custom one for replacement? I really doubt these exists in market pre-made...

 

[electechnow]

Hi Fezder.
Still haven’t heard back from admin re sorting signup.
so hope you don’t mind me posting here.
I have read the above info and wonder if you know what the the value of the electrolytic capacitors on the (U102) and (U202) are as mine have been leaking. I took U102 , U 202 out of circuit and lost my trace so hoping to be able to replace the caps and see if the traces return.

 

[electechnow]

Morning Fezder.
I have found out their value by cleaning with acetone. 10uF, 16 V.
But still have not signal pass on CH 1 and CH 2.
Thinking of bypassing Attenuating input modules.