Oscilloscope Repair

[unclejed613]

you really won't find much on the web about it, it just happens to be the closest description i could think of for what this amp does..... a more accurate description might be something like....
"fully differential-flat response from DC to daylight (well... 20Mhz, at least)-cascaded-cascode-hybrid-high slew rate-low current-voltage amplifier....." or something like that ....
this amp has a lot to do just to accurately drive a pair of deflection plates.... i would have stuck "high frequency" in the soup somewhere, but after all it is only a 20Mhz amplifier

 

[pierreretief]

Great, guys, seems we got some smart guys thinking about this. Just so by the way, I did measure at the terminals, so that rules out the hairline crack...

Im going to be taking the other transistor measurements and also the cap voltages tonight. will keep you posted

 

[pierreretief]

Oh yeah, I might get my hands on another scope. but its going to cost a few bucks, but if it would make this job easier, like very very easy, then I will get it... Or is it going to be pretty much just as hard even if I had another scope? I think its 40MHz

 

[unclejed613]

to actually see what's going on in the circuit in real time would be extremely useful

 

[shimniok]

I think having another scope would be immensely helpful. The multimeter will still be needed, but I think a scope would more readily allow you to explore the circuit to see where the AC is coming from.

It'll also let you get a better realtime view of what is happening as you're turning the position knob -- you'll see not just high/mid/low but where the AC first starts to appear, what that waveform looks like, etc.

If you get two probes, you can view both sides of the circuit at the same time. Or, any two points in the circuit. I'd recommend doing that. If should be possible to get 100MHz probes on ebay for ~ US$15 each + shipping.

And what we'll also be able to do is look for where the AC appears in the circuit AND you'll be able to show us the exact waveform on the plates because a 2nd scope will be able to trigger off that waveform.

And you can always sell the 2nd scope later. That's precisely what I did with a Tek scope I used to track down problems on my Hitachi. I ended up selling the Tek for a tiny profit in fact.

Michael

 

[pierreretief]

Ok, I will try for the scope, but I am quite broke at the moment. But I will see what I can do. I am going to take the other measurements now and post them. Hope it helps

 

[pierreretief]

Q5:

DC - Vbe - Down - 0.740
DC - Vbe - Wonky - 0.741
DC - Vbe - Middle - 0.768
DC - Vbe - Up - 0.785

DC - Vce - Down - 6.49
DC - Vce - Wonky - 6.48
DC - Vce - Middle - 4.6
DC - Vce - Up - 2.38

AC - Vbe - Down - 0.58
AC - Vbe - Wonky - 0.01
AC - Vbe - Middle - 0
AC - Vbe - Up - 0

AC - Vce - Down - 0
AC - Vce - Wonky - 0
AC - Vce - Middle - 0
AC - Vce - Up - 0

 

[pierreretief]

Q6:

DC - Vbe - Down - 0.778
DC - Vbe - Wonky - 0.777
DC - Vbe - Middle - 0.760
DC - Vbe - Up - 0.732

DC - Vce - Down - 2.38
DC - Vce - Wonky - 2.39
DC - Vce - Middle - 4.16
DC - Vce - Up - 6.48

AC - Vbe - Down - 0
AC - Vbe - Wonky - 0
AC - Vbe - Middle - 0
AC - Vbe - Up - 0

AC - Vce - Down - 0
AC - Vce - Wonky - 0
AC - Vce - Middle - 0
AC - Vce - Up - 0

 

[pierreretief]

Q19:

DC - Vbe - Down - 0.618
DC - Vbe - Wonky - 0.650
DC - Vbe - Middle - 0.641
DC - Vbe - Up - 0.646

DC - Vce - Down - 5.59
DC - Vce - Wonky - 5.49
DC - Vce - Middle - 4.86
DC - Vce - Up - 4.87

AC - Vbe - Down - 0
AC - Vbe - Wonky - 0
AC - Vbe - Middle - 0
AC - Vbe - Up - 0

AC - Vce - Down - 0
AC - Vce - Wonky - 0
AC - Vce - Middle - 0
AC - Vce - Up - 0

 

[pierreretief]

Strange that they dont mirror... why is that?

Q20:

DC - Vbe - Down - 0.690
DC - Vbe - Wonky - 0.679
DC - Vbe - Middle - 0.651
DC - Vbe - Up - 0.663

DC - Vce - Down - 5.42
DC - Vce - Wonky - 5.42
DC - Vce - Middle - 4.88
DC - Vce - Up - 4.89

AC - Vbe - Down - 0
AC - Vbe - Wonky - 0
AC - Vbe - Middle - 0
AC - Vbe - Up - 0

AC - Vce - Down - 0
AC - Vce - Wonky - 0
AC - Vce - Middle - 0
AC - Vce - Up - 0

 

[pierreretief]

There is something I dont quite understand. I am still not sure what we are looking for, because if you look at Q13 on the previous page, it seems we have our culprit. All the funny things happen when its wonky, so why does that not mean that it MUST be the problem. Or is it not Q13, but something before Q13. If so, then why measure Q19 and Q20, since that is after Q13...

Maybe you can just clarify this for me. I will post Q13 again:





Q13:

Anyway, here follows Q13's results. You will see that I indicated a few more tests with the DC Vce, like (3/4) which refers to a straight line, approx 3/4 way up the screen. And note AC-Vce for the wonky part, quite a jump just below and above... which tells me thats quite close to the prob...


Q13:

DC - Vbe - Wonky - 0.624
DC - Vbe - Middle - 0.029
DC - Vbe - Up - 0.012

DC - Vce - Down - 0.01
DC - Vce - Wonky - 39.2
DC - Vce - (3/4) - 101.8
DC - Vce - Middle - 93.6
DC - Vce - Up - 96.8

AC - Vbe - Wonky - 0.012
AC - Vbe - Middle - 0.185
AC - Vbe - Up - 0.237

AC - Vce - Down - 0.18
AC - Vce - Wonky - 22.1
AC - Vce - Middle - 0.25
AC - Vce - Up - 0.29

 

[pierreretief]

C25:

Down - 16.94V
Wonky - 16.97V
Middle - 16.97V
Up - 13.55V

 

[pierreretief]

What other caps should I measure?

 

[shimniok]

C26 and C33 were the other two we wanted to look at early on.

We needed both AC and DC for C25, C26, and C33. The reason is we want to make sure they are doing their job filtering out AC.

These are all on the reference voltage line for Q13/15 (and Q14/16) -- the schematic appears to have a couple of errors. It's my understanding that Q13/14 should be tied only to the node right next to C25. The schematic shows another junction at R77 which I think is incorrect. Likewise, the schematic shows a junction between the +12VA and Q13 emitter, which I don't think should be there.

Will respond to other questions shortly after thinking about things for a bit.

 

[pierreretief]

So we dont want to see AC on the caps, since a cap cant have voltage over it changing instantly?

 

[shimniok]

The caps are suppose to reduce or eliminate AC -- if they are shorted, they will fail to do so. If they are open they will also fail to do so. Caps resist voltage change. Another way to look at it is that they buffer voltage / energy. Also relevant is that smaller caps only buffer higher frequency AC. That's due to the time it takes to discharge.

 

[shimniok]

We can probably get this done just with the multimeter.

The reason I wanted to see Q19 and Q20 was they are ultimately driven by the Y+/Y- plate signal (between Q13/15 and Q14/16) after amplification by the boosters, Q17/18. I figured oscillation in Q17/19 area could affect Q13/15.

And, I wanted to see the whole picture. That's why I wanted to get measurements on Q5-9.

3. Both Q13 and Q15 show ~22V for Vce when trace is wonky. We already know the Y+ signal is ac -- so we know Q15 Vc and Q13 Vc are varying. For Vce to vary, Ve must vary *with respect to* Vc. Pretty significantly, at that.

What are we looking for... the source of the AC. Something, somewhere, is oscillating. If it's Q13 or Q15 themselves, then why? What other components are making this happen? Transistors cannot themselves oscillate without external components like resistors and capacitors.

oscillator basics

So what I plan to do is go thru every transistor measurement and see what logical deductions I can come up with based on what I know of transistors and electronics. For example, if you have a transistor with emitter tied to ground, but Vbe is AC, you know that Vb is ac because Ve is stuck at 0V. Likewise if a collector is tied to a voltage source directly, but Vce is ac, you know Ve is AC because Vc cannot alter. Vb will probably equal Ve + 0.6V or so. It may vary a little bit if the transistor current varies so you might see a little AC on Vbe. And so on.

I'll be looking for stuff that either shouldn't be possible or is unlikely to be wanted, like Q13 having a tiny Vbe with trace up, but mostly I am going to focus on when the trace is wonky unless I can determine a relationship between the trace being up and what we are seeing with the trace wonky.

That last part said-- one thing that has been nagging me the last few days is... there's a balance adjustment for the circuit... we need to go back and look at the service manual to see *where* the balance potentiometers are my printout of the vertical amplifier description references VR1,3,4 and 7,9,10 but I am also curious as to the purpose of VR6.

I can't help but wonder if the circuit were severely unbalanced, if it might start to oscillate on only one side. And if the trace would seem to be oddly "offset" up or down, or some such. And if some devices that should be a little bit off (Vbe=0.5, say) are wildly off (Vbe=0.012, say) and other devices are seeing weird drive voltages (Vbe=2.68V? Or -2.68V?)

A severe unbalance could be due to defective potentiometers, cold/broken solder joints, incorrect adjustments, broken traces near the pots, broken components near the pots, etc.

In case I don't get a chance to look into all this later... could you do me a favor and investigate VR6?

If it were me, I would first eyeball its position -- is it cranked all the way to one side or the other? Then I'd measure ohms resistance between the middle leg and each outer leg. It's acting as a variable voltage divider. So there should be some resistance on each side. If it seemed to be offset significantly, I'd remember it's original position, but try varying it to see what if anything that does to the trace when it is wonky.

Michael

 

[pierreretief]

great, thanks

 

[unclejed613]

when i tweak pots on a board, i always mark the original position with a fine tip permanent marker, so i can get them back to their original positions.

 

[shimniok]

So far here's what I notice.

. Q5/Q6 voltages mirror each other fairly well
. Q19/Q20 voltages remain fairly constant and similar to each other
. Q9/Q10 sort of mirror but not very well
. There is an odd change in Vbe on Q17; otherwise, Q17/18 also appear similar and relatively constant
. The fact that the wonky trace dies down over time suggests to me an oscillation that is slightly damped
. Even Q14 and Q16 have some AC across C-E which I think is really interesting.
. With trace UP, there are voltage problems on the ref voltage, at C25, as well as Q11, Q13, and to some degree, Q17 Vbe.
. Several transistor "pairs" don't mirror each other correctly: Q13/14, Q15/16 ... we can't tell about Q11/12 because we need Q12 measurements.



The reason I want to rule out (or discover) problems with VR6 is because it seems the circuit is unbalanced and the transistors that aren't balanced/mirroring each other (#7) seem dependent on VR6 in some way. Q17 Vbe, Q13/15 Vc, Q11 Vb, Q9 and Q10 Ve.

Now it could be another component causing the circuit to be out of balance, but let's start with VR6.

What we still need to look at:

0. Verify VR6 is approximately 'in the middle' and resistance from middle leg to each outer leg is approximately the same, about 1/2 470k or 235k.
1. Q11 Vbe wonky -- with Red on B and Black on E is the voltage + or - ?
2. Q12 Vce, Vbe, UP, MID, WONKY, DOWN
3. C25 AC, up, mid, wonky down
4. C26, C33 AC and DC, up, mid, wonky, down
5. Emitter AC and DC voltages with respect to ground for Q5 and Q6 -- at up, mid, wonky and down.