• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Bias problem on a KA-7100, new trimmer does nothing

I do have a scope, I got it intending to learn how to use it but I'm afraid I don't know how. That means I don't know how to intepret what I'm looking at. I have traced a sine wave part way through a circuit, but didn't understand how to set ranges to make some sense of the trace. Maybe I have a good opportunity here?

How does an amp oscillate? What exactly is doing the oscillating? need to learn more about that.

I'm sure the DMM is on DC, i took many measurements that seemed within specs on the s/m. You are right - that is a good tip!
 
Have a look on YouTube for how to use an oscilloscope videos. There are lots of them. Bad small value (100pF or less) caps can cause the amp to oscillate. They are put in there to roll off the frequency response at some point above audio frequencies but below rf frequencies. The reason is complicated to explain and would require more electronic theory knowledge on your part.
 
The upshot of the "maybe it is oscillating" is the amp might be working too hard and amplifying signal _above the audio range_.

You've got a great amp, the transistors doing the amplification are capable of amplifying far more than just 20Hz-20KHz, and doing so often produces hot power transfomrers and odd behavior. Which you typically cannot hear.

The "bad, small value cap" is in reference to very small-value caps added before/after transistors to filter out higher-than-audio frequencies that are being passed along and amolified. If one of these caps is not performing its job, you may be passing a high frequency signal and the amp is doing a very nice job of amplifying it... but you wont hear this.

Just my hobbyist $0.02 interpretation of the previous posts.
 
thanks for chiming in, Doug. And just how do you go about finding this little culprit? I have a scope, although as noted i don't pretend how to use it yet. Today I removed several tiny caps and measured them with the adapter on the DMM, but everything I measured compared similarly to those on the right side. I still have a 000.0 reading on the left bias, and I've tried to go over every sinlge component on the left to make sure it was identified and installed properly. I am sad to say they seem to be. What about the differential input? Is there a way to test it in place? It's the only thing I can't "see".

So - more questions. Does the unstable voltage at Qe15 Base and Qe21 C. mean oscillation? The hot heat sink would suggest that.

Could it be that having all new (modern) transistor equivalents, (except the differential pair,) creates new gain or voltage situations not suitable for the original schematic values? I have'nt read about anyone replacing every tranny yet. Maybe it's too much?

Where else can the fault be hiding? I used 6800uf mains, not 10,000uf. I used AK approved components, not 'think-it-will-work's, and I work one piece at a time, comparing the original to the replacement to the schematic to the notes. So it's frustrating to think it is correct and something is very wrong. I want to find a backwards transistor or electrolytic. I want to find a bad solder joint. I want to find silkscreen errors - oh, I did! Under the main power board the pins 7 and 8 are labeled backwards. It's the 28v B+/-. My wiring was correct, the silkscreen printing was wrong underneath.

We can overcome this! There must be some systematic approach to use that will find the issue. I just don't know what to do next. Back to the scope videos...(thanks dr!)
 
Last edited:
What I know about small caps comes from working on electric guitars, tone controls, effects foot pedal and speaker crossovers. I manged to modify a few tube amp tone circuits by changing the little caps, so I know tiny values are the higher frequencies and larger values get lower in frquencies. I usually used them as filters to pass the high frequency to ground. i can understand the need for that in this amp.
 
Interesting issue :idea:.. this was discussed in a pass forum regarding bias voltage and zeroing.

Just will say again my experience with a 9100 model of the Kenwood. I had always wanted to adjust any bias issues although I never found the amp to be operating out of spec.. now over 25 years old. So I used the procedure in the "users manual" for the amplifier, followed the procedure and found the bias to be non adjustable :dunno:.

Faced with going back and forth with the procedure, I decided to zero the meters (on the amplifier), and then get an idea of how far off the bias really is at that signal function. The meters did zero and after input of my "auxiliary in" seemed to operate to spec.. and sounded great.

Still not sure why the bias did not function to spec or if it really did anything to improve the sound but the amp works well and I am happy :thumbsup:.

Anywaysssssss... there have been numerous experiences with the rebuild of the Kenwood amps and their operational specs as noted by the manufacturer. I suspect some of these anomalies have to do with the DC design which is notably different when you troubleshoot the innards.

But as is often said in football.. "A win, is a win"! :banana:
 
One way you could go:

Your oscilloscope can be used like a signal tracer. Use a signal generator to input a known signal, e.g. 1 khz sine at 0.75 V ptp, and then follow that signal through the circuits, comparing right channel to left channel. Watch for where the wave or amplitude you see on the scope differs from the other channel.

There is a simialr, kinda cool way to do the same: http://solorb.com/elect/testcirc/octopus1/index.html

Anither way, what i do most often, is pull each component and measure it with a cheap $20 component tester: http://audiokarma.org/forums/index.php?threads/cheap-component-tester.749729/

I hear you and understand tour frustration. Im an ametuer so i run into this frustraion often. This is where i pull up my before images, compare every component to the photo, look for empty PCB holes, reflow the solder joints, question the silkscreening again, etc.
 
Getting out the scope, a Tektronix 2213, 60Mhz. I understand a scope is a "visual voltmeter", but like modern art, I don't know what I'm looking at. For a signal I'll use my Kindle Fire and an online tone generator. Maybe not perfectly clean, but effective enough.

That Octopus looks useful, I might get to that later, thanks. I have several more amps in the repair pile. I have been using that cheap 9v Chinese transistor tester, and if nothing else it gives relative comparisons. It seems to be working just fine, and has helped confirm pin outs in every replacement transistor.

I'll try to start tracing at the input and work forward, comparing right against left. All I know is to look for anomalies on the left.

Thanks for your input, 50nstillhifi. I never worked on a 9100. Every 3500 and 7100 has always adjusted properly for me, so I haven't seen this issue before.

...and the mad scientist retreats to the dark room with the flashing blue glow...
 
Oh, sorry, guess my subscription ran out, so I upped the status to Subscriber again. This forum is definately worth it!
 
Scope is up and running, I was able toget a stable sine wave onscreen. I have been testing the amp in full ac, no load, no inputs. For this test should I go back to using the DBT? In other words, would my results be meaningful enough at the reduced voltage? I hesitate to leave it powered up very long at full AC because of the apparent possibility of overheating, and I know I won't be very quick on this test.
 
Set the scope to AC coupled, 1V per division. Set the sweep to 2mS/ Div. With the amp OFF connect your scope probe to one of the emitter resistors on the left channel. Doesn't matter which lead. Connect the ground lead to the chassis. Turn the volume all the way down and set the input to AUX. Turn the power on, direct , no dbt. You will probably see a large deflection on the scope at first. but it will settle down to 0. You should see a flat line, no signal at all. If you see a sine wave, the amp is oscillating.
 
Flat line. 1v/div on 10x scale, 10x probe. Only kept it on about 1 1/2 - 2 minutes. No change during that time. Scope did wobble when I hooked up the ground, power off, so I know it was functioning. Was the aux input supposed to have a test signal? I didn't think so, so there was none.
 
No test signal. It doesn't appear to be oscillating. I think you have bias on the left channel and are just measuring it at the wrong place or aren't making good contact. Try this.
Power off, measure across the bias pot on the left channel with your meter on Ohms. measure TH2 on the right channel and TH1 on the left. Then measure R3 and R4, the bias pots. Tell me what you measure.
 
dr*audio said:
Try this.
Power off, measure across the bias pot on the left channel with your meter on Ohms. measure TH2 on the right channel and TH1 on the left. Then measure R3 and R4, the bias pots. Tell me what you measure.
Click to expand...
L bias pot = 299 ohms. R bias pot 291 ohms. I had set the L pot about 290 ohms after measuring the original Lpot and also the right side. 299 is measured in place in circuit just now.

TH1 = 645 ohms TH2 = 647 ohms. Both measured in place.

Voltage:
VR3 = -345mv VR4 = 1.2v Clearly something wrong here.

collectors on the main outputs were +/-50.5v and +/- 50.1v

I measured bias across one emitter resistor, then two, R69, R71 on the left side. Any millivolts seen quickly drop to 000. I tested with clips at the CP points, I tested underneath at the solder points of R69,70, and I tested from the emitters of Q27 to 29. All 000. The right side currently measures 45mv across both emitters, as I haven't kept resetting it without letting the amp warm up.

I'm going to change Qe13 again, using a KSA992. Maybe I'll get lucky!
 
Last edited:
I just thought of where the current might be going. Measure, power off, from either emitter on the left outputs to ground with your meter on ohms.
 
I've been looking for any weird grounds too, because to my analytical brain "something is grounding" (see post #5,#7) I again noticed the rail voltage dropping to 44.5v on the left channel after it was powered up several minutes.. This doesn't seem right to me, and might suggest some weird grounding too. I have gone through the PS carefully, as I have changed almost everything but the rectifiers (i have parts for that, too).

I had an oscillation problem on a KR-5600. I had put in 10,000 uf filter caps and heavy regulators, but somewhere they overdosed the current. I changed the regulators and the amp worked fine. This time I just used replacement 6800uf caps, and the approved MJE15032 / 33 pair. But this heating problem behaves differently, it is slower, suggesting some buildup of tension as voltages change over time. Just my mechanical brain! I have not been able to catch it in the act.
 
You must log in or register to reply here.
Back
Top Bottom