about using a function generator to test circuits

[tone4days]

hi folks

i recently purchased a very inexpensive function generator to use when bench testing circuits under development
i figure a 400 Hz sine wave will serve reasonably well as a source signal and can vary the freq to roughly gauge higher or lower freq response of the circuits

i have an inexpensive oscilloscope that i will use to measure the signal at various points in the circuit and at the final output as well

my question is about the amplitude of the signal

i was figuring sine wave centered at 0 volts with an amplitude of from -1 V to 1 V would be a good approximation

is this a good choice?
should amplitude be smaller? larger? should it be centered around 0Volts or should it be 0 V to 2V instead (DC offset)

all advice welcome
thanks
t4d

[blackieNYC]

Same setup here. Cheap.
Your sine wave will be "centered at zero". If the generator has a DC offset, your input cap will take care of that.
Level:  +/- 100mv (p-p) is more typical for guitar, though 1ormaybe 2v p-p can happen with the pick attack and high output HBs. Look for clipping with that kind of generator in op amp circuits, otherwise it's hotter than you need.
Don't just use 400hz.  Use several. How is your bass response? Use 83hz, or 100. Do you have a high output at 40 cycles? It's too low, consider filtering it out. How is your 8-10khz? Maybe you are losing highs, or have too much up there. Compare your fuzz pedals to one another with a freq sweep.
Your generator might not be a reasonable output impedance.  Put the signal thru a unity gain buffer first ( and measure the buffer to see if it is flat). And for fuzz pedals that need to see the guitar, build the AMZ pickup simulator and put that after the buffer.  See if your generator is flat.  Mine isn't. So I just remember that I'm down 4 dB at 10khz. I don't need another generator.

[PRR]

> should amplitude be

Think. Speech/music audio goes from dead silent to Very Loud.

For full exploration of a toy you want to cover the range.

In practice there is no "dead silent" (always some universal hisss), and in most circuits nothing real exciting happens for medium signals on down; just gets smaller.

"Real Loud" must be in context. Some audio goes dozens of Volts or more. H-P's classic audio generator made a point of going to 25V output (and there was a High Output option too). OTOH we here are mostly about guitar-cord signals. 1V is Very Loud to a guitar amp input. More than most guitars ever approach. With many many pedals, 1V input puts the internal stages into heavy clipping.

100mV (peak or peak-peak?) is mentioned above. I would set 200mV rms as a first try. But It Depends. Most "fuzzes" will start to beat-up the signal with not much over 20mV input. That suggests a test at say 10mV to see that small signals (decays) pass without unexpected damage, and tests at 50 and 200mV to see that signal damage increases as expected with higher levels.

Your source should have a knob on it so you can dial a wide range of levels. If no knob in the kit, hang a 5K-10K *Audio* pot on the output. Set the generator to maybe 1V output. Now "5" on a 0-10 knob is around 100mV-200mV, "nominal guitar level", you can dial down to very small and up to very large. Watching the 'scope as level changes tells you a lot more about many guitar effects than any fixed level could.

[PRR]

DC offset is rarely useful INSIDE a circuit. I think I have done it once in decades. I think they put it on there just to raise the knob-count (and perceived price-point). At your point in the path, I would nail it to Zero.

Sine wave is the "simplest smooth wave". It avoids much theoretical messyness. There's hardly any Sines in the real world, and a $2 chip's "sine" will be visibly and audibly "shrill", with a glitch over the top and high overtones on the ear. That's fine, as long as you know what it looks and sounds like.

To read the *exact* clipping point, a Triangle wave is better. The sharp tips cut-off more obviously than the rounded ends of a Sine.

A Square exercises "all" frequencies (a very wide band) at once. An output that still looks square is a very smooth "flat" circuit. However tone shaping and controls will warp a square into very strange shapes with not a lot of effect on what the ear hears. In theory you can work-out the complete frequency response by how a square comes out, but that is very hard work. At the bench, rounded corners mean high-cut, slated tops mean low-cut, but unless severe the cuts may be very far from the basic frequency. (And many pedals are all about cutting extremes.)

[johngreene]

DC offset is rarely useful INSIDE a circuit. I think I have done it once in decades. I think they put it on there just to raise the knob-count (and perceived price-point). At your point in the path, I would nail it to Zero.

Unless you are using it to generate a square wave to provide a clock into a logic circuit. In that case I use the DC offset a lot. Especially when working with BBD circuits.