B&K 3011B signal generator, no output

[Johnnieboy]

I was checking a tube amp and intending to inject a signal into the grid of one tube, but attached to an incorrect pin.  Afterwards I was not getting any signal output.  I opened the generator and found no obvious issues.  Anyone familiar with this model that can offer a suggestion.  Later units of this model have a fuse installed immediately before the output.  Mine has no such protection.  Thanks!

[bdunham7]

What position was the output switch in and what voltage/current do you think it was subjected to by being connected to that pin?  I suppose you'd just have to start poking at stuff starting at the output and working your way back.  Here's a schematic of the output stage.

[Johnnieboy]

I was injecting a 1k sine wave.  The voltages could've been pretty high.  I was expecting to see something blown up, but really there was no physical evidence of anything abnormal.  I realized I was not outputting anything on all 3 waveform outputs.  One thing for sure, I'm adding a fuse in that output circuit when fixed!

[Wallace Gasiewicz]

I would check the output transistors in place first with a diode checker to see if the PN junctions have about .6 volt across them.It is likely you blew themAlso you could start by checking the output resistors for proper values. They may be blown without visible damage.

[bdunham7]

Let's say it was subject to +100V at the output.  That means that about 1 amp of current would flow through R518 + R519, Q507 and R515.  That means 0.5A through each of R518 and R519, which means 50V across them and 25W dissipation, 50X their rating.  They are small metal-film resistors and very well might blow quickly and quietly.  R515 is a 47R 2W carbon composite, that seems much less likely to blow quickly.  Q507 has a rated collector current of 0.6A, so a brief 1A pulse likely won't kill it.  Also, if Q506 or Q507 were blown, whether short or open, I don't think I'd expect "no output" but either noisy, clipped output or large DC offset.  So, I'm thinking R518 and R519 are blown and that's the first thing to check. 

[Johnnieboy]

Checked all the resistors in your statement and all are spot on value.  Thanks for your input bdunham7

[wasedadoc]

If you put +100V on the output I would expect that Q503 would cease to be a happy bunny. Or a parrot .

[Grandchuck]

Have you measured the voltage at the junction of the two 10 ohm resistors?  It should be close to 0V.  Have you verified the supply voltages?

[bdunham7]

Checked all the resistors in your statement and all are spot on value.

OK, so what do you have for other test equipment?  I would check the +/-15V and =/-22V power supplies and then start poking my way through the output amp with an oscilloscope starting with the base of Q501 all the way through to the output.  Also, you might just check with an ohmmeter from the output pin of the BNC (probing from the center of the BNC on the outside of the front panel) all the way to the junction of R516 and R517, then to the base of Q502.  You should get ~50R then ~18K.

[james_s]

Most of the time when parts fail there is no physical evidence, you need a lot of power before something craters. Test the transistors with your multimeter, and if those are all good (unlikely) then start checking resistors in the output stage.

Next time you need to inject a signal, do so through a film capacitor of sufficient voltage rating.

[Johnnieboy]

Q503 seems ok, thanks

[Johnnieboy]

+/-15 volts good.  +/-22 volts showing +/-28 volts dc?

[Johnnieboy]

Junction of r16/r17 is -25vdc.

[Grandchuck]

Q507 might be shorted.

[james_s]

Just check all of those transistors, you can use the diode test function on a multimeter, each transistor should appear as two diode junctions, one B-E and one B-C. Polarity is swapped for PNP vs NPN.

[Johnnieboy]

Replaced Q507 and all is well.  What a great site, thanks to all, every suggestion was noteworthy.  This site is fantastic!   

[james_s]

Glad to hear you got it working, thanks for following up with the outcome. Next time you kill it you'll know right where to look.

[Johnnieboy]

 
I will!  Thanks again.

[floobydust]

You don't want to go directly to the "grid" anyway, as that scuttles any bias and +ve excursions draw grid current.
Just add a blocking capacitor, a 10k series resistor etc, anything to protect the signal generator and not upset the DC bias.

[Johnnieboy]

Thanks, good advice.

[Johnnieboy]

One last question, before I add a 200ma fuse to the output.  The manual I found online said it should be a slow blow fuse...does that make any sense?  I would have said a fast blow type.

[bdunham7]

One last question, before I add a 200ma fuse to the output.  The manual I found online said it should be a slow blow fuse...does that make any sense?  I would have said a fast blow type.

I agree that a fast-blow is probably appropriate, but don't count on it to save your transistors if you have a similar event.  It will prevent a fire though. 

[Johnnieboy]

I plan on using both the fuse and a blocking capacitor on the output.

[james_s]

Even a fast blow fuse isn't going to prevent damage from excessive voltage. It takes almost no current for a high voltage to fry a semiconductor. If the fuse blows, it will be because the transistor shorted.

[Grandchuck]

Glad it is fixed!  One possibility is to place an MOV from the junction of the two ten ohm resistors to ground.
https://www.digikey.com/en/products/detail/panasonic-electronic-components/ERZ-V07D270/227552
I do not know the maximum peak to peak output swing of this signal generator, so the MOV linked might not be the best choice.