reasonable test signal amplitude

[pula58]

I am wondering what a good sized signal (in amplitude) to input to my amp build to test-out various things about it. If I strum my guitar (low-output single coils, 6K impedance or so) I can get a pretty big p-p as seen on the o'scope - on the order of many volts p-p. Yet, I am pretty sure that if I put in a 1kHz sine wave with the same ampltiude that I'll be driving the heck out of the preamp.

I am trying to work with a reverb circuit in the amp (it is an AB763 build)..and if I put in too-large an input signal the reverb circuit p-p output (as seen at anodes of the 12AT7 reverb driver) it is several hundreds of volts -p-p. Seems kinda large to me.

Any ideas/thoughts/opinions are welcome!

[husky]

50mv~100mv rms 1Khz is normal test level to the input but you might need to go lower, try and adjust for a clean signal.

[Cliff Schecht]

Depends on the pickups. I have a Hot Rails that will shoot out a volt or so when you really dig in and IIRC EMG's put out a few volts. With that said, unless I am looking for distortion issues in an amp (cathodyne PI clipping, power tube bias shifts, blocking distortion, etc..) I'll run about a 50-100mV P-P 1 kHz sine wave into the amps input. When I am tracing out an amp looking for the point where a signal is dying, it doesn't necessarily help to have the entire amp clipping hard, I want to see a nice pretty sine wave coming out of each stage (if that's what I expect).

Also triangle waves are a fun way to look for distortion issues in any circuit (tube, discrete or integrated). Since what you are expecting is perfectly straight lines ramping up and down, it's much easier to see any distortion than on a sine wave. Of course this is assuming your function gen doesn't suck and can put out a clean triangle wave on it's own. It's hard for the untrained eye to see slight distortion on a sine wave but anybody can see distortion on a triangle.

[TUBEDUDE]

A 400 Hz square is handy also for seeing the bandpass of stages, adjusting feedback, and finding potnntial instability. As it contains all of those nice harmonics you can see high filtering. (leading edge rounding), high freq over emphasis (ringing on the leading edge) or low freq issues (trailing edge abnormalties). Square waves have helped me find instability issues I could fix before I found them with the speaker connected. ( everyone in my neighborhood appreciates that).

[jjman]

I use around 300-400mv p-p. That's not going to clip the 1st stage. Then I use the volume control on the amp (which is before the 2nd stage) if I want a strong signal moving down the line. I want to be able to see the clipping that a real guitar-strength signal will achieve. Of course the guitar signal won't be a nice sine wave. All my pickups are HB or pretty loud single coils.

[katopan]

I usually use 100-300mV depending on what I'm testing for but at 333Hz (3mSec period) Gives a good view of the waveform on 1mSec grid setting and is more useful sort of freq for guitar amp measurement.

[pula58]

thanks for all the great tips!

I think what I am trying to get at is to put in an appropriate (in terms of amplitude) test signal that will show the signal swings through-out the amplifier to try to 1) learn what is happening 2) use the signal swing info to help me design a new amp 3) try to understand some dynamic range issues ...but, I want to ascertain these things with the appropriate amplitude signal that is equivalent to a guitar (single coil) output. The thing is, it is hard to get a good measurement when I strum my guitar into the oscilloscope!

[jjman]

If you have compressor, put it between the guitar and amp/scope. This should make the signal's strength more "stable" and readable.

[Masco]

A little off subject of the OP, but does anyone here sweep the frequency response of their amps? Maybe by using a sweep generator and spectrum analyzer w/ peak hold function?
I imagine once everything goes through the speaker/load the response is changed, but still would be neat to compare preamp freq. response/signature.

[katopan]

I often use a pink noise wav file put through a neutral SS amp into the speaker being tested, then use a spectrum analyser plugin to see what's coming back in through a mic in front of the speaker to get a rough frequency plot of the speaker response.

I have also used that technique (but feeding the pink noise signal straight into the amp being tested, in other words without the SS amp) to get a frequency plot of a couple of amps. Much quicker with it's instant plot than a manual or automated sweep. Not as accurate as lab test gear, but it gives a rough plot and responds as you'd expect to tone control changes, etc.