Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Oscilloscope Repair

Status
Not open for further replies.
Correct.

We'd been measuring Vce and Vbe which highlighted some issues, but by setting the trace in the true middle of the screen and adjusting VR6 and measuring Vc and Vb and Ve this will help us figure out what parts of the circuit are unbalanced. E.g., is it limited to the amp section or is it on the input side too? Probably the imbalance is causing the oscillation.

Glad to hear you are getting the scope. I think that will help. Most scopes have high voltage input limits, higher than the 120V we have in this ckt. But to visualize the higher voltage waveforms you need a probe with 1x/10x setting, set to the 10X which attenuates the signal by 10 and effectively increases the volts/div by 10. Ie, my scope has 5v/div max, 8 divisions, so 40Vpp waveform that I can see. If I set the 10x switch, it becomes 50v/div, total of 400Vpp. If you are DC coupled, the 10X switch may help see both the DC and the ac on top. If you are AC coupled the max Vce is ~30V ... but maybe Vc and Ve are higher so you could use the 10x switch.

Would love to see a picture of the wonky waveform stabilized. That may help. But also you can use the scope to more easily see and write down where the waveform is and isn't in the ckt. Use AC coupling, btw, to eliminate the DC component.

Good luck!

Michael
 
Right, here are the transistor voltages. It seems to my untrained eye that Q7/Q8 and Q13-Q16 look funny, and why 0V on Q9 and Q10... anyway, maybe the you can spot the fault from this data..

Q5
E - 4.1V
B - 4.8V
C - 8.5V

Q6
E - 4.1V
B - 4.8V
C - 8.5V

Q7
E - 10.1V
B - 11.8V
C - 10.7V

Q8
E - 6.9V
B - 4.8V
C - 6.3V

Q9
E - 9.2V
B - 8.6V
C - 0V

Q10
E - 9.2V
B - 8.6V
C - 0V

Q11
E - 11.8V
B - 10.8V
C - 16.4V

Q12
E - 11.8V
B - 12.4V
C - 16.2V

Q13
E - 16.4V
B - 16.8V
C - 119.9V

Q14
E - 16.2V
B - 16.8V
C - 53.7V

Q15
E - 119.4V
B - 118.4V
C - 119.4V

Q16
E - 119.4V
B - 118.8V
C - 53.7V

Q17
E - 11.6V
B - 10.9V
C - 6.2V

Q18
E - 13.2V
B - 12.4V
C - 5.3V

Q19
E - 119.8V
B - 120.3V
C - 125.5V

Q20
E - 119.4V
B - 120.4V
C - 124.8V
 
i'm a little busy this weekend, but i'll try to take a look at it tomorrow (sunday). if i get a chance, i'll print out the schematic and mark the voltages on it and see what i can figure out........
 
Collectors for both Q9 and Q10 are tied to ground, no problem there.

  • Q5/Q6, Q9/Q10 are perfectly balanced.
  • Q7/Q8 are not even close. -- but these two aren't symmetrical/mirrored
  • Q11 base is slightly off versus Q12, the others are balanced.
  • Q13/14 collectors are way off, B and E are close.
  • Q15/16 collectors are way off, B and E are close.
  • Q17/18 are not quite the same
  • Q19/20 are pretty darn close.

I think we can rule out the current limiter/switcher circuits. The imbalance (and presumably the oscillation) appears to be an issue only in the amp.

So whatever is causing this appears to be affecting Q11 Base, the midpoint between Q13/15 (their collectors / Y+) and to a small degree, Q17 Base. Basically that whole feedback network. And Q17 might affect Q19 to a small degree (although the Vb of Q19/20 are very, very close).

Btw, it shouldn't be possible for Q19 and Q20 collectors to have different voltages; in the schematic they are tied together to 120V source! Your measurements show them off by 0.7V I think that is one area worth investigating.

Also, it finally jumped out at me... my understanding is that a Bipolar Junction Transistor, even when conducting in saturation mode, should drop a few tenths of a volt across C and E but Q15 shows 0 volts. We saw this in prior measurements but it didn't quite hit me. Then I thought Q13 looked iffy as well, measuring 0.01VDC, but 0.18VAC at its lowest point which maybe is normal. Q14 is iffy measuring 0VAC/0VDC at the up position. Q15 shows 0VDC and 0VAC at the up and middle positions. Neither should be possible for Q14/15. But at least Q14 is dropping the appropriate voltage at the midpoint. Really only Q15 looks super odd at this position.

If Q15 is shorting under certain circumstances, it could be doing other weird stuff. What I don't comprehend is why the Q11 and Q17 base voltages are *lower* than their counterparts if the feedback / Y+ is too *high*.

And Q15/16 have the same base voltage, which rules out the voltage source and the base resistor (22Ω). Q13/14 also have the same base, ruling out problems with their respective base resistors and their shared voltage source.

I also notice that, as expected, Q19 Ve is about 0.4V higher than Q15 Ve, due to R95 between them. However, Q20 Ve = Q16 Ve. That shouldn't happen due to R103. These readings may also be due to measuring at different times due to effects of warm up.

Q14 Ve = Q12 Ve and likewise Q13 Ve = Q11 Ve. This should happen due to voltage drop across R75 and R74 unless the current through here is very low resulting in a drop of < 0.1V due to current < 4.5mA This may also be due to measuring at different times due to effects of warm up.

All told, I suspect Q15 is bad. Maybe there is something else wrong in the circuit, too, but I can't help but wonder if Q15 isn't misbehaving causing Q11 and Q13 to misbehave. I don't know for sure if it is causing the oscillation but I can't help but think it is causing the imbalance.

When the trace goes wonky, where is the position knob? 1/2 of the way down? Or 3/4 of the way down? 7/8ths? More? Less?

That question is key to me understanding if there is some weird non-linear voltage drop behavior in just Q15, or also Q14, or all the transistors.

I would like to check into the following just to make sure of what we're seeing. With trace and VR6 set at midpoint:

1. Measure voltage across R74, R75, R95, and R103 would probably be in the < 1V range. From the looks of it. Also measure ohms, should be 22Ω within 5-10%. I say do both just to be sure there is both current and correct resistance. Shouldn't take too long, hopefully.

2. Doublecheck Vc for Q19/Q20 -- the best way I can think to do this accurately is to attach the black lead to Q19 C and the red lead to Q20 C and hten power up and let it warm up for 10 min. You should see 0V. Set on milivolt range and it should be really really close to 0V. If not, we'll see if we can figure out what is going on.

I think if the above two sets of measurements indicate no issues, then I would strongly suspect Q15 as the main cause of all of this, because I can't see how it can have the same base and collector voltage as Q16 yet have a totally different voltage drop, nor does it make sense that Q15 goes to Vce=0 at mid and up points.

I am suspicious of Q14 so IIWY and if it isn't too expensive, I would get replacements for Q13, 14, 15 and 16 in case Q14 is problematic, too. And especially, if you can't get original transistors it might not be a terrible idea to replace pairs (Q13/14 and Q15/16).

My theory to explain how up/wonky positions are related is if Q15 shows misbehavior in the mid/up positions where it is either fully conducting (up) it seems plausible it could be acting non-linearly in other areas of its operating range, resulting in oscillation.

Michael
 
When the trace goes wonky, where is the position knob? 1/2 of the way down? Or 3/4 of the way down? 7/8ths? More? Less?

Now, with VR6 set to the middle, the trace is not wonky anymore, but with the position knob all the way down (where the wonkyness usually happened) the trace is a straight line one div above the middle

And sorry, but what does IIWY mean??
 
Ok, I ordered all 4 transistors. they are more expensive, but I guess it is since they are high voltage transistors? Or why do you think? They also look different to the normal BJT black ones. They have a heatsink also...
 
R74 Voltage = 0v Resistance = 22ohm
R75 Voltage = 0.04V Resistance = 22ohm

R95 Voltage = 0.228V Resistance = 100ohm
R103 Voltage= 0.235V Resistance = 100ohm
 
Now, with VR6 set to the middle, the trace is not wonky anymore, but with the position knob all the way down (where the wonkyness usually happened) the trace is a straight line one div above the middle

Interesting!! Hm. The imbalance, then, must be the main problem which isn't surprising since that's what you first noticed.

And sorry, but what does IIWY mean??

IIWY == If I Were You :D

Ok, so, hopefully the 4 trannies aren't TOO expensive. I had to pay like $20 for the matched pair on my amp. But, could've been worse. They were big honkin transistors (to supply 70wpc and a crapload of current to the speakers). Make the TO-220 power trannies in that scope look like pipsqueaks.

Anyway... I mentioned I'm not 100% sure it's Q15 but 90% sure. Vb checks out. It just doesn't drop voltage like it is supposed to. Q19 upstream and Q13 downstream both seem matched to their counterparts reasonably well. I just can't imagine how Vce would be 0V unless Q15 is jacked up

Ok, so Vc checks out on Q19/20 -- excellent!

I'm glad to see R103/R95 measure out ok. You're seeing .228/.235V at 22Ω so that is around 11mA ... since they are similar it is probably a-ok.

But R74 concerns me because there doesn't appear to be any voltage drop across the resistor. At least you are seeing R75 dropping *some* tiny voltage, 0.04V. That's about 2mA.

It is possible that R74 has a problem at high voltages -- someone had suggested that and I mentioned it recently, I think. If it were shorting out at high voltages, that would cause a little too much current, and maybe that is what is throwing off Q15 -- or maybe not. Might be causing the buzzing I thought you mentioned.

Finally -- I think it's worth replacing a couple things now to see what happens. So here is my suggestion for next steps. You want to change one thing at a time and see what it does, if anything, to the behavior of the circuit. If you change a bunch of stuff and it gets worse, we're really screwed. But if you change one thing at a time and report the results we can get through this even if it gets worse before it gets better.

1. a. If your multimeter has a separate milivolts setting, check R74 at that setting one more time.

b. If 0.00V drop -- or if you don't have a separate millivolts setting, replace R74 with a resistor of similar type -- R74 is a Carbon Film, 1/4 Watt, 22Ω (see service manual, 2-C10-013). They should be dirt cheap, like US$0.10 or so. Get 5 or 10 since there's loads of these in the amp and others may go bad. These resistors should be fairly readily available -- from digikey, mouser, newark, jameco, etc.

c. After replacing, see if there is a voltage drop across the resistor now? Let us know how much before doing #2. Hopefully it will match R75.

d. Check function of the scope and see if it seems fixed or not.

2. If replacing R74 didn't fix everything, then replace Q15 -- rather than all four -- and see what happens. Let us know before #3. I don't see too much voltage or any suggestion of a short that would kill Q15 so I think it is safe to put in the new one.

3. If that fixes it outright, great, if there are still problems which there might be, we can decide what to do at tha time. Don't get impatient and replace everything at once :D

Keep your fingers crossed, hopefully between R74 and Q15 we have it nailed.

Oh, these transistors are probably slightly unusual as well as high gain, high voltage BJTs or something, and they are larger TO-220 packages rather than the smaller TO-92 packages of most of the others, so that is probably why they cost more.

Michael
 
Last edited:
i just did a double check here, and i think Q15 is shorted or breaking down. R85 has 4mA flowing through it, but R103 has little or no current flowing through it. the deflection plates should both be sitting at approximately half the rail voltage of 120V (i.e. at about 60V, 53 V would probably be close enough). yet the Y+ plate is stuck at the rail voltage, while the Y- plate is at about a normal level. (btw, the difference in voltages at the transistors furnished the info to find the resistor currents) i would also suspect Q13 as being open, which is why it's Vce is so high, and the current through Q15 is only 4mA despite it's being shorted or saturated, as well as the fact that the excess current doesn't pass through R74.
 
Q13's Vce is so high because it has to make up for the voltage *not* dropped by Q15. I could be full of it, but -- on my own, working P3502 -- Q13/15 Vce's seem to vary inversely proportional to each other (likewse Q14/16), based on trace position.

So I think Q15's breaking down, yes. And actually, all the excess current is probably NOT going through Q13 because it is basically shut off at this point, but instead diverts thru the feedback network to Q11/Q17, perhaps...

Anyway, sounds like we are on the right track with checking/replacing R74 and swapping out Q15 to see what that does.

Eager to hear what happens...
 
2 things. How would replacing 74make a difference? Since it is measuring exactly 22ohms, doesnt that mean that it must be working?

and secondly. I couldnt get the exact transistors, but NTE substitutes. SHould I still only replace one, or will they all work better together...
 
if you get NTEs, go with replacing the whole quad of Q13,14,15,16. granted, the circuit is pretty much Hfe independent, but you want all of the devices to have the same Ft, so they will all amplify equally well at high frequencies. the Ft is the Transition frequency, which is where the Hfe drops to 1, very similar to an op amp and it's Gain Bandwidth Product. in fact some transistor manufacturers have begun listing Ft labeled as GBW (i'm not so sure the terms are interchangeable). while Hfe changes with collector current, Ft changes with Vce. it's a good idea to get all 4 transistos at once, because NTE probably uses a different part with slightly different characteristics to replace a range of different parts, and you want all 4 of the transistors to be as similar as possible. for a given range of Vceo, Ic, Pd. Hfe and Ft (those are the most important specs when selecting a replacement transistor BTW), NTE usually picks a device that exceeds all of the parameters for a whole range of parts, and buys a lot of them and re-labels them with an NTE number. so if you have a transistor rated at 50V, 1A, 40W, Hfe of 150, and an Ft of 50Mhz, the NTE part is likely to be a part rated at 60V, 1.2A, 50W, Hfe of 170, and an Ft of 60Mhz. this is fine if you're replacing a single ended amp to go ahead and just replace the one transistor, but the oscope circuit here isn't a single ended amp. you have a quad of transistors here that should have their characteristics as close as possible to each other. they don't have to match perfectly, but they should be within 10% of each other.
 
How would replacing 74make a difference? Since it is measuring exactly 22ohms, doesnt that mean that it must be working?

But it is dropping 0V. There aren't many explanations for that happening.

Either your meter isn't reading the voltage drop, or there's no current, or the resistor is shorting.

It is possible that Q15 is dropping no volts (shorting) causing Q13 restrict current going through R74 so low that you aren't able to read it.

Or... as I mentioned above:

shimniok said:
It is possible that R74 has a problem at high voltages -- someone had suggested that and I mentioned it recently, I think. If it were shorting out at high voltages, that would cause a little too much current, and maybe that is what is throwing off Q15 -- or maybe not. Might be causing the buzzing I thought you mentioned.

I'll let unclejed weigh in as to whether this is possible. It's a cheap component and would take little time to replace, so can hurt to try.

Michael
 
it's been my experience that only wirewound resistors develop short circuits, so, no it's not likely that a resistor is shorting. Q13 has some, but not much forward bias B-E (about 0.4V) so it may be in cutoff. you won't measure much voltage across a 22 ohm resistor, even with the circuit operating normally, but there should be a few millivolts across it at least. your meter resolution didn't go that low, so it's possible there's a few millivolts there, but we're not seeing it, just because most DMMs have only 1 or 2 digits beyond the decimal point to work with. your measurements resolve in 100mV steps, so a difference of even 50 mV could go unnoticed (above 50mV, you might get the last digit alternating between 2 readings). since the load being driven by this amp is purely capacitive, there isn't a whole lot of current here to begin with. if i were working on this circuit, i would begin measuring the voltages across the various emitter and collector loads to find out just where the currents are or are not going, especially around Q13 and Q15 and comparing them to Q14 and Q16. so the immediate priority would be to look at R74 compared to R75, and R95 compared to R103. then i would check the base currents by measuring R76, R77 and what looks like R83, R84. then use ohm's law to find the currents......
 
hey guys... Just got the transistors and replaced all 4, like you said unclejad... but alas. same problem persists...

But the hiss I hear does change if I move the position knob. and also as i move it up, the line "flickers" sometimes and changes very slightly in brightness, but not sure if thats new or not, since its very slight.

Also the hiss was gone at first, and took a few seconds to start up again, I cant remmember exactly if that normally happened, Il let it sit for a day or so and switch on and listen if the hiss is there immediately.

Ai, seems this thing just doesnt want to work!
 
also, very strange, if I move the position knob at a certain speed, i make the hiss much softer. but only at a certain speed, not too fast, not too slow...

mmm???
 
sorry, i must have missed it somewhere, but you say it's hissing? this is a key piece of information........ a hissing noise could come from only a very few sources. a wire or connection is arcing somewhere, or the CRT anode (actually a cathode supply in this case, more in a minute...) supply has a problem. the anode supply is a switching regulator made up of Q402, Q403, and Q404. in this scope the tube supply is kind of "backwards" instead of putting a high positive potential on the anode of the tube, a relatively low voltage is on the anode. the voltage required to force electrons toward the screen is actually applied as a very high (somewhere around 1.9kV according to the block diagram). this power supply could be the source of the hissing. i usually use a piece of carburetor hose as a "stethoscope" to find hissing and chirping noises. as i get older, i have also made high frequency microphones attached to pieces of carb hose, because my hearing now rolls off at 10-12Khz, and a microphone going into a scope input is the only way i can detect high pitch noises above 10khz. in any case, a piece of carb hose is a very good tool for localizing noises. if you find that what i'm assuming is marked T402 (in which case the power line input transformer would be T401) is the source of the hissing, you may want to check the electrolytic caps in the power supply, as some of them (especially the smaller ones) may be dried out. if you find the hissing coming from somewhere else, especially from a connector or wire bundle, you probably have something arcing. i suspect that if you find the source of the arcing, you are probably very close to the source of the vertical problem as well. another possibility here (worst case scenario) is that the tube has an internal short between the Y+ plate and one of the grids.
 
You know... I asked a couple of times about the hiss after you mentioned it in passing ... but my questions went ignored :D

So, are we still sure that R74 resistor can't possibly be shorting/arcing ?? :D (now you know why I asked about that possibility) :D

I'd be thinking about what the hissing has to do with the unbalanced circuit... I don't see how the Z supply has anything to do with it.

You might think about doing some of the testing unclejed mentioned to identify where the problem is in the amp circuit. I still think there's a good chance that the amp circuit is where the problem lives. If nothing else, you now know that it is something driving Q15 that is bad... I still believe that Q15's non-linearity is very close to the root cause. So the next step is to find what is causing Q15 to behave so oddly.

Michael
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top