Turning audio into DC.

[Steve Jones]

I am putting together an audio to trigger circuit to convert audio pulses (such as the output of a drum machine) to 5V gate pulses to clock a sequencer. I have breadboarded up a one-shot pulse generator using a 4013, it fires with a positive voltage applied to the input.

Next I need to design a circuit for the input stage which will take an audio signal on an unbalanced input and condition it to provide a positive DC output. Can anyone suggest the best way to do this? I am guessing that I need to rectify the audio and clip it with an op-amp, or perhaps use a Schmidt trigger, but I would appreciate any ideas that anyone may have to save me re-inventing the mag wheel.

 

[amorris]

cant you just use diodes?

 

[bcarso]

[quote author="Steve Jones"]INext I need to design a circuit for the input stage which will take an audio signal on an unbalanced input and condition it to provide a positive DC output. Can anyone suggest the best way to do this? I am guessing that I need to rectify the audio and clip it with an op-amp, or perhaps use a Schmidt trigger, but I would appreciate any ideas that anyone may have to save me re-inventing the mag wheel.[/quote]

We would need to know how sensitive you need to be before going much further. As it's a drum machine I guess one can characterize the envelope of the brief damped a.c. waveform pulses fairly well. If they are always the same initial polarity it would make it easier, and I would suspect they probably are.

It's actually not that trivial a problem. You want the response to be quick, but if too much so you will get multiple triggers. If too slow you will get noticeable latency.

The best way would be to pull the clocking signal out of the innards of the drum machine!

 

[Steve Jones]

[quote author="bcarso"][quote author="Steve Jones"]INext I need to design a circuit for the input stage which will take an audio signal on an unbalanced input and condition it to provide a positive DC output. Can anyone suggest the best way to do this? I am guessing that I need to rectify the audio and clip it with an op-amp, or perhaps use a Schmidt trigger, but I would appreciate any ideas that anyone may have to save me re-inventing the mag wheel.[/quote]

We would need to know how sensitive you need to be before going much further. As it's a drum machine I guess one can characterize the envelope of the brief damped a.c. waveform pulses fairly well. If they are always the same initial polarity it would make it easier, and I would suspect they probably are.

It's actually not that trivial a problem. You want the response to be quick, but if too much so you will get multiple triggers. If too slow you will get noticeable latency.

The best way would be to pull the clocking signal out of the innards of the drum machine![/quote]

Well, it has to be able to trigger off pretty much any short input signal, much like the key input of a gate. I figured that if I firstly put the audio through an op-amp with lots of gain it would clip the audio to the supply rail level, and then I could put it through a diode as amorris suggests, and then presumably I would have a pretty much DC voltage to use.

The one shot circuit that I have already built will only clock with each + input so it shouldn't double trigger, that's why I thought of massively over gaining the input op-amp, so that even a small pulse will just stay high. The problem with this of course is having the trigger too sensitive, so maybe some kind of AND gate with one leg tied high would be a better idea, or a comparator?

 

[tony666]

Hi, you need a differentiator that goes to a comparator and use a diode after it. If you want I can send you a schematic?

 

[JohnRoberts]

You want to Full wave rectify the audio first.

Then establish a threshold, to ignore false triggers. Too much filtering will add delay that could be annoying.

Then feed a one shot to get a usable trigger pulse.

Not trivial so you may need to experiment some.

JR

 

[StephenGiles]

THAT 2252 perhaps?

 

[buschfsu]

FRED's

 

[Steve Jones]

[quote author="tony666"]Hi, you need a differentiator that goes to a comparator and use a diode after it. If you want I can send you a schematic?[/quote]

I am beginning to see that this is indeed not so simple. I thought also of putting a 555 in there as a monostable. Do you have a schematic of the configuration that you suggest?

Once I get this working it would be a pretty useful device for all the synth heads.

 

[Steve Jones]

[quote author="StephenGiles"]THAT 2252 perhaps?[/quote]

THAT looks promising too... :green:

 

[gyraf]

Use a noisegate - tap led action indicator voltage (possibly via opto-isolator if needed). That sorta thing is adjustable to many signal types.

Jakob E.

 

[clintrubber]

I didn't check, but wasn't there some Oakley-module doing an alike thing ?

 

[Sleeper]

Clint-pretty sure you are right.
my oakley moog filter has audio triggered envelope and a CV out.
So does my novation bass station rack.
Almost any drum module that takes triggers will also convert convert those impulses to midi and I push unfiltered audio into my alesis DM4 all the time.
there's a sensitivity control so it works quite painlessly.

maybe this is a better solution, + then you get a synth or a filter or a drum module out of the deal.
kelly

 

[clintrubber]

Hi Kelly,

Your drum-triggering comments triggered the though of some PAiA circuits...

Dunno Steve, maybe suited, but probably too simple/coarse ? ... you'll have seen these...


http://www.paia.com/ProdArticles/drumsens.htm

Regards,

Peter

BTW, re-reading the thread title, I'd sure love to turn certain audio into DC :twisted: :wink:

 

[Sleeper]

Steve, now that Clintrubber got me thinking... I usually just plug the drum machine straight into the gate input and adjust sounds and levels until it triggers correctly. write a pattern of quantised 8th or 16th notes. my old dr-660 has a metal stick sound that de-tuned down as far as it can go has happily synchronized a bunch of old trig input din sync stuff. roland 606 808 etc.

maybe there is something more complex going on that you need to trigger, maybe it's so easy you should give it a try.
Kelly

 

[Steve Jones]

Thanks for that, I will have a look at the circuit to see if I can adapt it to what I need. I guess between something like the PAIA circuit, or the Korg MS-20 external input section, or maybe even the key input of a Drawmer gate I may be able to come up with something.

I must say that the THAT 2252 RMS level detector that Stephen Giles mentioned looks promising too.

 

[mickey76]

If I understand correctly you need an audio to trigger converter. You should check out some synth diy sites.

http://www.musicfromouterspace.com/analogsynth/signaltogate.html
http://www.digisound80.co.uk/digisound/modules/80-13/80-13.htm

 

[Steve Jones]

[quote author="mickey76"]If I understand correctly you need an audio to trigger converter. You should check out some synth diy sites.

http://www.musicfromouterspace.com/analogsynth/signaltogate.html
http://www.digisound80.co.uk/digisound/modules/80-13/80-13.htm[/quote]

Thanks for that, I will take a look. The MFOS circuit looks interesting, perhaps with a mod for pulse length at the output.

 

[mickey76]

To modify the pulse length you could make the 1 Meg resistor in the pulse stretcher (partially) variable. There's a bit of info on this type of pulse stretchers on that same site:
http://www.musicfromouterspace.com/analogsynth/mmlogic.html

Or you could replace the pulse stretcher with a 555 in monostable configuration. That gives you more options to play with like easy control over threshold, retriggering and pulse length. I don't think you really need all that for your application though.

 

[Steve Jones]

[quote author="mickey76"]To modify the pulse length you could make the 1 Meg resistor in the pulse stretcher (partially) variable. There's a bit of info on this type of pulse stretchers on that same site:
http://www.musicfromouterspace.com/analogsynth/mmlogic.html

Or you could replace the pulse stretcher with a 555 in monostable configuration. That gives you more options to play with like easy control over threshold, retriggering and pulse length. I don't think you really need all that for your application though.[/quote]

Great link, thanks for that. I Already have breadboarded up a one-shot pulse circuit using a 4013, so I think I will hang it off the end of the signal to gate diagram you linked earlier.