ideal value for low ripple Envelope follower??

[loss1234]

well i have been trying out a lot of env follower ideas and they all work well enough but the RIPPLE seems to be a problem

So...

1. Can i get rid of ripple with a simple RC filter BEFORE the rectifier circuit? and if so, what RC values work well (and what is the cutoff? am i trying to cut everything above 100 hz? 5k? etc)

2. does having the input sens HIGH make the ripple worse? or better?

(as it sounds bad as it decays mainly)

does anyone have an opinion on which pedals got this right?

thanks so much

[loss1234]

one more question on this

in the past i have always used SYNTH style followers like this:

http://electro-music.com/forum/phpbb-files/paia_envfoll_129.jpg
but have never know how to SIMPLY change the attack or decay without adding more diodes.

is it possible?

(maybe by making r7 a pot?)

thanks so much

[Mark Hammer]

Ripple is largely a problem when one aims to have lag built into the response; most particularly a long-ish decay time.  Ripple is always most audible during the decay phase of a picked note or strummed chord, so if the decay is fast, then you don't hear the ripple.  The PAiA follower you link to has no real attempt to introduce lag.  There is low pass filtering via R8/C6 which rolls off ripple at 6db/oct starting around 14hz.  As near as I can tell, that sort of filtering does not result in lag, since there is always a feedback path via R8.  Of course I say that without any experience with this design.

[CynicalMan]

This thread talks about attack and decay for simple envelope filters:
www.diystompboxes.com/smfforum/index.php?topic=79074

I think putting an RC filter before the circuit would just reduce the circuit's sensitivity to high frequencies.

[aziltz]

http://www.diystompboxes.com/smfforum/index.php?topic=79185.0

in this thread I started regarding the Orange Squeezer's envelope detector, Mark and some others discussed using Hi-Pass Filters before the Detector, since the ripples come from the lowest of lows.  Beats and stuff like that.

[loss1234]

thanks guys!


i will try a high pass filter
i figure i can use one of those RC filter calculators if know the range i am trying to cut

oddly i have seen low pass filters after envelope followers before...but those WERE in synth circuits so it might be a very different aim.

[StephenGiles]

Thinking about this, I'm not sure why it's referred to as an Envelope follower because all it usually follows is the initial peak voltage of the input signal, which charges up a capacitor. This voltage is allowed to drain away to 0v or whatever at a rate set by a resistor/pot in the circuit. If the input note decay is followed, the result will be a mess.

Now when the unit is at rest, suppose that the capacitor is held high, and when a note is played, gate circuitry immediately discharges the capacitor to a low state. Very soon after, the capacitor is allowed to charge back up to a high state at a rate set by a resistor/pot, producing a control voltage  for further processing as required, to drive the voltage controlled element (s).

[Cliff Schecht]

one more question on this

in the past i have always used SYNTH style followers like this:

http://electro-music.com/forum/phpbb-files/paia_envfoll_129.jpg
but have never know how to SIMPLY change the attack or decay without adding more diodes.

is it possible?

(maybe by making r7 a pot?)

thanks so much

The problem with modifying the heart of this circuit, the full-wave rectifier composed of IC2 and surrounding components, is that you will also change the gain of the system and will throw off how the circuit operates. Changing R7 will change the gain of the output filter and will also move the cutoff frequency. I do like the idea of throwing a high-pass filter in front of this to remove excess low-frequency ripple, but I've found that for the guitars range this circuit really has minimal ripple.

To be honest, this circuit as is isn't well suited for what you are trying to do as is. What you need to do is add another stage that can add in some finite amount of attack-release time to the envelope output.

[StephenGiles]

well i have been trying out a lot of env follower ideas and they all work well enough but the RIPPLE seems to be a problem

So...

Can i get rid of ripple with a simple RC filter BEFORE the rectifier circuit? and if so, what RC values work well (and what is the cutoff? am i trying to cut everything above 100 hz? 5k? etc)


does anyone have an opinion on which pedals got this right?

thanks so much

Nothing simple will get rid of ripple, this has been asked time and time again. Look at Larry Bissell's design:
http://www.edn.com/article/CA265499.html

I had no success in getting it to work but I may have missed something.

Once again, EH got it right in the Space Drum - not for guitar of course but the peak follower could be adapted so that the peak start frequency point follows the pluck strength - but I've burbled on about that before and nothing has been forthcoming. Perhaps Cliff Schecht could help out here?

[StephenGiles]

<a href="http://www.4shared.com/file/31630821/a3d91443/HSR_Gate_sch.html" target=_blank><img src="http://dc127.4shared.com/img/31630821/a3d91443/HSR_Gate_sch.jpg" border="0"></a>
<a href="http://www.4shared.com/file/31630869/c96e5975/HSR_Gate_text1.html" target=_blank>H&amp;SR Gate text1.doc</a>
<a href="http://www.4shared.com/file/31630873/30a0812a/HSR_Gate_text2.html" target=_blank>H&amp;SR Gate text2.doc</a>
<a href="http://www.4shared.com/file/31630785/55077fed/HSR_Gate_text3.html" target=_blank>H&amp;SR Gate text3.doc</a>

Some going home reading for those 6hrs behind GMT, yes I know it's a gate circuit but there is some clever technique in there which may be of use in a low ripple envelope generator.

Download for full size.

[StephenGiles]

Loss - Have you learned anything from the Gate article that may help in your quest for low ripple?

[StephenGiles]

Calling Loss1234..........................are you out there?

[loss1234]

sorry man...was out of it for 2 days working on other stuff....i will read the articles right now and get back to you

I APPRECIATE EVERYONES HELP!!!

(stephen...i have actually been learning from reading your old AUTO WAH article and looking at your royal filters,etc)

[StephenGiles]

sorry man...was out of it for 2 days working on other stuff....i will read the articles right now and get back to you

I APPRECIATE EVERYONES HELP!!!

(stephen...i have actually been learning from reading your old AUTO WAH article and looking at your royal filters,etc)

That old Auto Wah article is probably 80% mine. The idiot editor of Hobby Electronics would not believe my original circuit worked, despite providing a woking unit!

[StephenGiles]

"Working circuit" I should say, Woking is a town to the west of London!