Another O-scope thread: Multiplexer?

[Mark Black]

I have come to the conclusion I would like lots more channels/traces on my scope, specifically for working on mixing consoles. Like maybe 8 or more. It seems like it would speed up my testing considerably if I could look at the outputs of all the Aux sends and masters at once, rather than making multiple passes across the console with re-plugging of outputs. Or on really big consoles at least limiting to a couple of passes.

I guess I could get two or three 4-channel scopes but that sounds a bit expensive for my budget and would occupy a lot of real estate.

Has anyone any experience with O-scope input multiplexers? Could that be a usable solution or are they not all that great?

Or what about an older logic analyzer - can they be used as a multi-channel scope?

I once saw a cute rack-mount setup with something like 5 small scope CRT modules at a surplus electronics store but didn't buy it and now of course the place is long gone and I can't find info on anything like it on the web. Might have been made by Kenwood...

Opinions/experiences welcome - thanks!

Mark

[Satamax]

Mark, i know nothing about thoses multiplexers, but i think someone in the console and desk industry has had the same idea as you and may be giving a call to a specialist might lead to a good info. Would it be repairer or maker.

Bye.

Max.

[Enzo]

I have never fooled myself into thinking I could see distortion on my scope unless it was severe. My ears however can assess the sound in short order. Once I determine the individual Aux masters are good and the returns as well, then all I have to do is mix them all into the master, and on each channel strip, go down the row twist twist twist on each knob and move to the next channel. Monitor the master - one master channel.

You can make a multiplexer, yes. Now you have eight tiny traces on your screen. My scope in front of me has 8 divisions top to bottom. I can either stick one of your channels in each division for a truly small signal display, or I can overlap them all and struggle with trying to see which is which. I am not happy with the choice.

If your idea is a sequencing multiplexer that constantly scans through the inputs, well, at that point I would take one scope channel and stick it on the master and proceed as if I were doing it by ear, since you have to cycle through the series, you might as well just go down the channel with the knobs.

Besides, we want to not only determine if the signal is OK, we also need to check the action of the pot.

[Mark Black]

Thanks for the replies guys -

I'm usually looking for frequency response anomalies by using a square wave signal, along with hash from noisy switches or pots. Additionally I look to see if the signal floats around indicating possible DC leakage within the channel or output strip. These things can be seen using the mixdown method of course.

What I'm always afraid I'll miss is crosstalk, which requires seeing all of the outputs concurrently.

I agree the traces would be pretty mashed together and that's a good point, but overlap wouldn't necessarily bother me if all the signals should be identical. I've just never touched, used or contemplated a multiplexer before and was wondering if they work well in general, or are somewhat a toy.

Twist, turn and push - yup - familiar with that. A 40-plus channel console with lots of auxes, sweep-EQ's and a matrix can make your fingers pretty sore, especially if you do it twice: once on the way in to document all the problems and again on the way out to verify the fix (and to make sure you haven't caused anything new to appear). Tons of fun!

Thanks again -

Mark

[bob p]

If you want to demux data, maybe you'd be better off getting a digital sampling board for your PC. You can find boards for this sort of thing from Keithley and similar suppliers. you can configure the number of mpx'd channels yourself. Then all that you have to do is crunch the numbers inthe data stream as they come in. A simple for loop is all that's required to perform demuxing.

Its actually quite simple to do if you have rudimentary programming skills. On the first day that I tinkered with this sort of stuff I learned basic syntax and wrote the demuxing routine. It took me another day or two to write the code to auto-size the waveforms/channels on the PC screen. I coded a complete medical monitoring system using DSP, a PC and no tools beyond DOS and C/ASM compilers. but that was a long, long time ago...

IMHO a PC is actually a better device for a lot of this DSP stuff because the display is so large. but there's a huge trade off in convenience when you have to create your virtual device from scratch.

If you have very special needs, rolling your can give you an exact solution for your needs that might not otherwise be available. It may not be time-efficient, but its inexpensive dollar-wise compared to buying a custom solution.