Tracing Signal Blew My Small Amp

[mwelch55]

I built an audio probe for tracing audio signals.  I was using a small one watt solid state amp (Danelectro Honeytone) to listen to the signal.  I was tracing through a tube preamp and it blew my HoneyTone amp.  The signal was too large for the small solid state amp to handle.  I built a DIY one watt solid state amp to replace it, but I need to protect the input from signals that are too large.  I have read about using zeners to protect the input of the amp.

What is the best way to protect a very small amp from very large signals?

[PRR]

#1 is: when poking hundreds of Volts inside a tube amp, you must use a low-uFd high voltage cap before the listening amp. Even as small as 1nFd 600V.

[Fender3D]

#2 Thank god for the high voltage blew the amp and not YOU

[mwelch55]

I did have a 600v capacitor in series with it.  That would prevent any DC from getting through.  I think that because the signal was too large it blew the op-amp in my small amp.  I need to limit the input to prevent very large AC signals from getting into my test amp.

[pinkjimiphoton]

yeah, the honeytone is probably capable of about 3v p2p, and in a tube amp in some places i'm sure you'd be way beyond what it could handle.

usually, you'll wanna add a good sized resistor to the circuit. yeah, it may change the tone, but who cares, all ya need is to hear the signal is going thru.

use a biggish resistor, i mean physically big... like 1-2 watt maybe. i'd imagine anything from around 1-10k should do the trick.

but wait til someone who knows what they're talking about confirms this, please.

please be careful messing with the gutz in tube amps... if ya can smoke your honeytone, you could smoke yourself, too.

[GibsonGM]

Lots to consider.   Understanding the circuit path (vs. power path - tube plate etc) can be helpful, LOL.

As mentioned...HIGH voltage cap....perhaps a series R of some nice wattage (1W or better gives you 'time', it will get warm...).   

Still could find yourself inputting a signal maybe 10's of volts AC into a small device like an opamp, so....how about a R and a ZENER diode to gnd?  IE, a regulator with appropriate Rs.   Maybe a 1W zener.   Just a thought, haven't heard of one used that way before. Might be too much 'spread' from 60V down to 5.7V or what have you (dissipation).

If I am not sure what's going on, I'll input a 1V, 1kHz sine wave into an amp as a test signal.   Then, use my DMM to see what kind of levels we're talking about, in the places I want to probe (grids, plate before/after coupling cap will be high. No need to probe other side of plate resistor etc!)...I know it's going to be more than transitors or opamps like, so I use my tube amp to listen, but at times want to probe w/O-scope...in that case, it's a 10X probe.

Maybe you could just make a 10X audio probe with HV cap?  Easy enough.  I've made them for cheapo O-scopes, it's just 2 resistors.

-----------------update---------

EX:  You have your DC-blocking 600V cap first, so no DC worries.   If you then used 1K in series and 18k to ground, the ratio works out to about .05 * input voltage.  So, a 60V AC signal (pretty typical between stages) would come out of that network at about 3V.    Now, the dissipation could be up there a bit, but using 2W resistors should allow you to 'hear' briefly for 'is it working' use, and allow it to cool between...and sine waves are not always 'peak', so the real dissipation is less than for a DC calculation (RMS says <2W dissipation).   If you can get 5W R's, use those, even better, you should be all set with them - less to worry about.

Make sure ALL connections are well-insulated!  And remember - a lot of wire is NOT rated for about 300V or so!  You probably want to be sure about that before you put it all together.   Go slow and be sure you're not melting anything    Esp. yourself. 

[GibsonGM]

I felt I had to come back to this just to let MWelch know that...you CAN scale those resistor values, if you make a voltage divider.  That will mean the R's draw less current, meaning less dissipation (heat!), so lower ratings would be ok.    I should have talked about that I think!

You could use a 2.2k for "R1" and 38k for "R2", for instance, and still maintain that dropping ration (actually a 'gain', just negative), but HALVE the power they need to dissipate!  And so on.   I was thinking more of real-world consequences, where higher R's = more noise, and/or loss of treble....and here it doesn't matter

[mwelch55]

Thank you for the replies.  Gives me something to think about.  The voltage divider is a good idea.  The voltage divider would not be needed for pedals.  I was thinking about making the divider switchable or make two probes.