Newbie needs a little help with first repair

[injb]

Hi, I'm trying to figure out what's wrong with this Kenwood FG-273 function generator I bought on ebay. I got it years ago and I was going to fix it back then, but it got shelved for a while. (https://www.eevblog.com/forum/beginners/is-this-function-generator-broken/new/?topicseen#new)

Anyway, I'm giving it another shot, and hoping to learn something as I go. But I have no experience with this stuff and I could use some help, if someone doesn't mind taking a look. I have a 4 channel scope, and I have the schematic, but I'm struggling to understand how it's supposed to work, and what I should be looking for.

The issue is that the output signal is tiny and badly distorted. Turning up the amplitude knob doesn't increase the amplitude, it just increases the distortion. The first thing I found is that the actual signal generation is working fine. The signal coming in to the amplitude pot is perfect. The problem happens in the final amplifier stage.

Following the advice from my old thread linked above, I've looked at the signals in various places, but it's hard to tell if what I'm seeing is the problem or not. The part that's confusing me right now is that the input signal to the push/pull transistors (Q12 and Q13 in the schematic) is a square wave, regardless of which signal is being generated. For example, I captured the attached waveforms with the sine generator:

* Blue (channel A) is the from R84, just after the amplitude potentiometer, and looks perfect
 * Red (channel B) is the collector of Q10 (approx 30v p-p)
* Green (channel C) is the collector of Q11 (approx 30v p-p)
* Yellow-ish (channel D) is the final output (only a few hundred millivolts, should be 20v p-p)

I don't really understand the function of Q10 and Q11. Do they act like Darlington pairs with 12 and 13? I get the feeling that this square wave isn't my problem, or at least not the whole story, because I don't see how it could cause the weird signal shape I'm getting at the output, or the tiny amplitude. I checked the voltage at the output stage (R102 and R103) and it's around +/- 30v, so a bit higher than the 25v in the schematic (in fact all the voltages I checked in this area are a little higher than it says in the schematic - not sure if this is a clue or not).

So assuming this square wave on the base of Q12 and Q13 is not the problem, how it it supposed to work? If everything else was working correctly, how could you ever get anything other than a square wave out of the final stage?


Thanks!

[tautech]

Typically output stages of AWG's/FG's get 'hurt' when asked to inject a signal directly into a circuit with a higher voltage than the breakdown of their power output stage.
Work on that premise and suspect any device that might not withstand such a reverse voltage.
Passives are generally OK but they're quick and easy to test......resistors should always measure =or+ their marked value (in circuit) unless there's a parallel path. Caps are most always OK.

[Mr. Scram]

Typically output stages of AWG's/FG's get 'hurt' when asked to inject a signal directly into a circuit with a higher voltage than the breakdown of their power output stage.
Work on that premise and suspect any device that might not withstand such a reverse voltage.
Passives are generally OK but they're quick and easy to test......resistors should always measure =or+ their marked value (in circuit) unless there's a parallel path. Caps are most always OK.

I guess a bias tee would prevent that from happening.

[injb]

Thanks for the relies, I'll keep all that in mind.

But now that I've been looking at the schematic, I really want to understand the function of Q10 and Q11. I don't understand why they have no signal input to their bases. They just have a biasing voltage from the supply rails, and that's it. What's the point of that?

Edit: I think I figured it out. Q10 and Q11 are configured in common-base mode, so the emitter is the input and the collector is the output. From reading about this configuration, it seems it gives very high voltage gain, which makes me think that the 'square' wave I'm seeing is actually just the sine wave being clipped. Looking at the waveform again, it's actually not a square, it tapers a little, because it's really the base of a sine wave.

So now I think what is supposed to happen is that the gain of those Q10 and Q11 is supposed to be controlled by negative feedback from the output, and obviously that's not happening because there's practically no output. So the 2 output transistors are broken, and if I replace them then then I'll probably see my normal input signal at the collectors of Q10 and Q11.

Does that sound like I could be on the right track? Even if I'm right it still leaves me completely in the dark as to the purpose of Q10 and Q11, but one step at a time

[Shock]

If you look at -0.84V and +14.0V, 0v, -13.9V it appears that is critical in setting the bias for the base of Q12 and Q13. You want to check these values but I think there is something to the supply reading high, it's almost 20%. Check it's on the correct line voltage and for noise/ripple on each of the voltage rails.

I'd also switch the db pads to bypass so you know what your output voltage should be and reduce the chances of a bad signal. So make sure they are making good contact and it's a low resistance path. Next I'd check Q12 and Q13 and the caps for noise/ripple, plus R98 and R99 for drift.

If you look in the amplitude section the caps there are up to 25V above ground, so a cap acting like a resistor there would have quite a adverse effect on the signal. Is noise/ripple seen there?

It can sometimes help with the scope connected when you manipulate controls such as pots and switches to see bad and dirty contacts.

[injb]

If you look at -0.84V and +14.0V, 0v, -13.9V it appears that is critical in setting the bias for the base of Q12 and Q13. You want to check these values but I think there is something to the supply reading high, it's almost 20%. Check it's on the correct line voltage and for noise/ripple on each of the voltage rails.

I'd also switch the db pads to bypass so you know what your output voltage should be and reduce the chances of a bad signal. So make sure they are making good contact and it's a low resistance path. Next I'd check Q12 and Q13 and the caps for noise/ripple, plus R98 and R99 for drift.

If you look in the amplitude section the caps there are up to 25V above ground, so a cap acting like a resistor there would have quite a adverse effect on the signal. Is noise/ripple seen there?

It can sometimes help with the scope connected when you manipulate controls such as pots and switches to see bad and dirty contacts.

There is only some ripple (120 Hz) on one side of the big resistors R102 and R103. On the other side it's gone so the caps seem to be doing their job.

I did spot one capacitor that has a very slight bulge. Is this bad? It's really very subtle - I tried to capture it with a few pics (attached), but it's hard to tell. Anyway, it's definitely slightly convex at the top, whereas all the others are dead flat. This one is part of the 5v regulator circuit at the other side of the board.

I also removed Q12 and Q13. My multimeter has a hFE tester and diode tester so I tried them in that, and from what I can tell they're both dead. I tried another good transistor to make sure I was using the tester correctly, and I get hFE of 130-140. I also get ~1K ohm between the base and each of the other 2 leads on that good transistor using the diode test (and nothing if I reverse the leads, which seems right).

But on the 2 I removed from this board, I get nothing for the hFE test. I tried the diode tester, and on one of them I get infinity between the base and either lead, both ways around. On the other one, I get 0 ohms between the base an one lead (can't remember which) and infinity from the base to the other.

I think these transistors are just dead! I don't have any prototyping stuff whatsoever, but I have a basic breadboard kit arriving soon so I think I'll be able to do a more definite test then.

[tautech]

...............
I did spot one capacitor that has a very slight bulge. Is this bad? It's really very subtle - I tried to capture it with a few pics (attached), but it's hard to tell. Anyway, it's definitely slightly convex at the top, whereas all the others are dead flat. This one is part of the 5v regulator circuit at the other side of the board.

It's shagged showing any sort of bulge.

I also removed Q12 and Q13. My multimeter has a hFE tester and diode tester so I tried them in that, and from what I can tell they're both dead. I tried another good transistor to make sure I was using the tester correctly, and I get hFE of 130-140. I also get ~1K ohm between the base and each of the other 2 leads on that good transistor using the diode test (and nothing if I reverse the leads, which seems right).

But on the 2 I removed from this board, I get nothing for the hFE test. I tried the diode tester, and on one of them I get infinity between the base and either lead, both ways around. On the other one, I get 0 ohms between the base an one lead (can't remember which) and infinity from the base to the other.

I think these transistors are just dead! I don't have any prototyping stuff whatsoever, but I have a basic breadboard kit arriving soon so I think I'll be able to do a more definite test then.

Yep, sounds dead Jim !
Always have the datasheet on hand to get correct leadouts so your hFE tests results make sense.
Just occasionally they'll test like what seems OK but it's outside the datasheet spec.

[Shock]

As well as those transistors I'd say that cap is going, the heatsink next to it is not helping. Use a 105C rated version though as a replacement in situations like this.

The 25V supply comes from the 15V section marked on the schematic so that is C31, C33 and as well as C74, C75 I'd probably just replace the lot to be sure. I'd typically check around the area so I'll measure R98, R99, R102, R103 and see if they are still 5% as it's going to affect the transistors.

Plus check the line voltage selection to address the 20% high voltage issue. That would mean 30V on those 35V caps and they may be running out of spec. Line voltage selection plugged into the 100V instead of the 120V perhaps?

[injb]

Fixed! Thanks for the help everyone. I got some replacement transistors and I also removed the 2 big resistors - they were still in spec but they looked a bit rough so I replaced them too. I never got to the bottom of the voltage being off what it says in the schematic, but with the new transistors in place it's working nicely, so I'm just going to leave it as it is.

[Shock]

Good to know it's operational.

If you are keeping it, as I said the voltage line selection being set incorrectly could explain the ~20% over voltage, it's in the manual. It wouldn't be the first time someone has sold something with the line voltage set wrong internally and it caused a failure.

Hopefully you replaced C48, as well right? High voltage and high ripple will make that voltage regulator go next. You could be in for another $1-6 plus shipping and end up soldering a new U13 in.

[injb]

Good to know it's operational.

If you are keeping it, as I said the voltage line selection being set incorrectly could explain the ~20% over voltage, it's in the manual. It wouldn't be the first time someone has sold something with the line voltage set wrong internally and it caused a failure.

Hopefully you replaced C48, as well right? High voltage and high ripple will make that voltage regulator go next. You could be in for another $1-6 plus shipping and end up soldering a new U13 in.

Oh yeah, I forgot to mention that i checked the line selection on the transformer, and it was correct. I didn't replace C48...I'll try to remember it next time I'm ordering parts. Hopefully it won't be too late!