Sample and Hold with Noise Source

[brr0]

First of I want ot make sure that the LF444 is an acceptable substitute for the TL084. Assuming that is true, here's my question. when I use the noise source as the input for the sample and hold, instead of a random variety of notes (when sending to an ocsillator) I get a very limited number of values. It depends a bit on the speed of the s&h but, generally I just get maybe 2 or 3 different notes. Any ideas where I should look for the problem?

Thanks.

[thapoke]

It could be the colour of noise that the noice source generates . White noise and pink noise have different frequency content so you get different 'random' values from them in a sample and hold.

[fluxmonkey]

could also be a problem with the level of the signal you're sampling... try using an oscillator as the input (tri or ramp, not square... and at a different rate than your sampling clock!) to see if that gives different results. if that's better, then try subbing in different 2n3904s in your noise circuit to find one that puts out a stronger level.

b

[fonik]

the good thing is, your circuit samples and holds -

i am with bbob: i would suggest either a problem with the input signal to be sampled (the higher the noise level, the wider the range of the voltage that could be sampled and hold ... i once had 2N3904s that refused to produce noise at all! ) or with the input or the output buffer circuitry (false resistor values?).
different noises (pink, white or whatever) will provide different weighted values, i guess.
(white noise provides all frequencies with the same level, pink noise provides all frequencies with decreasing level for the higher frequencies. there are more uncommon noise colors like brown/violet...)
matthias

[brr0]

Thanks, I think that's it. With my alligator clip test spaghetti I think I was tapping from the noise output pot, and that level is lower. I thought I also attempted to tap directly from the output pin, but perhaps I didn't since I just tried again and it works better. I suspect the noise output might be on the low side, so I'll see if another 3904 works better.

My multimeter as a transistor tester. Would testing for the highest output transistor help with this or is the noise making capabilities of a transitor not directly related to its output?

Thanks much.

[fonik]