Soft On (LED)

[Toney]

As opposed to mother-in-law/bikini/beach..

Anyway, I Have been looking at a way to slowly ramp up a LED to avoid thumping. I have a RM I built that is infected with Satan's electrons.
I have tried everything, including building a CMOS Millennium, but that sucker just WILL thump no matter what. Seem it just does not like any sudden current changes.
Now this:
It works beautifully to slow on/off the LED (sorry for the basic paint schem) but you get the idea. If provides a delay at "on" and a nice "analog" fade out when it is switched off.



  Any other ways to ramp up the LED slowly and cause the least disruption to a finicky circuit?

[Toney]

[~arph]

Nice and simple, I like it. Might try it on my SHO.

[cloudscapes]

I always star-ground the bypass led straight to where the power supply is hooked up. Either with wiring or on the PCB. I don't get the led to "borrow" one of the sensitive component's power rails, even if it makes it easier.

[Toney]

I always star-ground the bypass led straight to where the power supply is hooked up. Either with wiring or on the PCB. I don't get the led to "borrow" one of the sensitive component's power rails, even if it makes it easier.

Yes that's pretty much the issue with this.
Its just a standard positive ground Rangemaster. Actually in this case, the thump is only audible when the volume pot is in first half of rotation, worse as the output cap is "closer" to the rail. So any less than 5k resistance and she thumps. Its livable with (sorta) but I really want to defeat it completely.

[R.G.]

Yep. Two ways to stop clicks/thumps: (1) ramp them to make the transition slower and (2) provide a source of current for the changes that doesn't affect the power supply voltage - both power and ground - to the rest of the circuit to be amplified.

[merlinb]

  Any other ways to ramp up the LED slowly and cause the least disruption to a finicky circuit?

Use a regulated power supply.

[R.G.]

That helps, but poor power and ground distribution wires/traces still leaves click and tick problems by transients on the power and ground as well as capacitive pickup from the LED voltage transition and M-field radiation and pickup.

These things should not be problems, but there are places of wickedly high impedance and gain in today's pedals.

[deadastronaut]

I always star-ground the bypass led straight to where the power supply is hooked up. Either with wiring or on the PCB. I don't get the led to "borrow" one of the sensitive component's power rails, even if it makes it easier.

hmmm could you elaborate on this?....3pdt /led diagram maybe?..


rod: cool idea.  thump is not pleasant especially when loud!.. 

[runmikeyrun]

I have a RM I built that is infected with Satan's electrons.

this is the coolest thing i've read all week!!  lol

[Toney]

Yep. Two ways to stop clicks/thumps: (1) ramp them to make the transition slower and (2) provide a source of current for the changes that doesn't affect the power supply voltage - both power and ground - to the rest of the circuit to be amplified.

I read you mention using a differential amp back in an old thread but I was unclear on the details. Something like this perhaps?



Here is the layout I used. I made it to mount on a pot and added the cap switch a bit later. Perhaps one too many compromises? I can say that the input set up is not responsible for the popping, nor are there any construction errors causing it. So does it come down to layout? One too many compromises?

[R.G.]

I read you mention using a differential amp back in an old thread but I was unclear on the details. Something like this perhaps?

Yep. Done properly, the current flowing out of the top/power rail and into the bottom/ground rail is constant. There are no current changes on them, so they can't make transients. The only transients are at the bases for the drive signals; this can be only about 1x Vbe; and at the collectors. The "off" collector is at the power supply and the "on" collector is one Vcesat above the emitters. 

Here is the layout I used. I made it to mount on a pot and added the cap switch a bit later. Perhaps one too many compromises? I can say that the input set up is not responsible for the popping, nor are there any construction errors causing it. So does it come down to layout? One too many compromises?

Even for people who are trained at electronics, current-induced interference is hard to get, probably because we're trained in a voltage-centric manner.

All wires are actually resistors. Pour current into a wire from two places and the wire makes a *voltage* equal to the instant sum of the two currents times the wire resistance. It can be milli- or micro-volts. But if your voltage amplifying stage uses that wire for a reference (that is, it has zero common mode rejection as we say in the biz), then any currents flowing in the wire are mixed with the input signal and get amplified.

The premium solution to this is to make sure that you **know** what currents flow in what wires, and the wires that contain sudden steps and extraneous signals are not the same wires that have your input signal reference. I know I've mentioned this several times, but having the input jack of an effect switch the negative side of the power on amounts to forcing all currents, both linear and switching, to flow in the reference wire for the input to an effect. That this works at all is a miracle, and one of the many things we just get away with.

[Toney]

All wires are actually resistors. Pour current into a wire from two places and the wire makes a *voltage* equal to the instant sum of the two currents times the wire resistance. It can be milli- or micro-volts. But if your voltage amplifying stage uses that wire for a reference (that is, it has zero common mode rejection as we say in the biz), then any currents flowing in the wire are mixed with the input signal and get amplified.

The premium solution to this is to make sure that you **know** what currents flow in what wires, and the wires that contain sudden steps and extraneous signals are not the same wires that have your input signal reference. I know I've mentioned this several times, but having the input jack of an effect switch the negative side of the power on amounts to forcing all currents, both linear and switching, to flow in the reference wire for the input to an effect. That this works at all is a miracle, and one of the many things we just get away with.

Ok thanks.
I'll try two ways -  by redoing the layout in a tidier fashion with all of the -9v tied together in a small space - no 'wandering' across the board and tying all the grounds to the output ring.
The fact that this pedal only thumps with the pot in the lower half of rotation the resistance to power - around 2k and below says it 'wants' to be separate by that amount to reject the change as the LED comes on. I guess could install a 2k resistor in series with my handy revised taper 8KA pot and be done with it.   
Layout redone below in case its of use to anyone:

[Satchmoedie]

I love this! So simple and cheap! I used or tried using LEDs as a version for an effect with popping. How fast can this cycle? I also have used CMOS switching with a diode to try to keep the popping down. This is getting filed away for sure! This is too simple! Yet, it should work great.


By the way, Does anyone etch boards on this forum? Strip boards are really cool, and great, but etching you can engineers keeping certain things away from each other, heat sinks, ground planes, etc.


Also who sells cheap strip boards? Yeah I use them too, but eventually I just enjoy either designing a board manually with my feeble brain power, pad and paper, or Dip Trace, or some other free program. These are all usually a rip and replace methods that yield various results from great to nightmarishly problematic. I can usually get a good 1 sided board designed on the 2nd attempt for most analog stomp boxes. This is on average. Simple stuff takes 1 try, and I have artwork. The Fender Fuzzwah clone (1970s) took 2 tries.

[.Mike]

Are there any advantages to using the method in the first post of this topic over using the methods in the AMZ LED Popping article?

I'm going to breadboard and test 'em both, but figured I would ask.

Mike

[Toney]

This way the transistor base is held open as the cap drains, so a slow "off" fade.
I tried the AMZ approach but it still popped at shut off.

@Satchmoedie
The caps size is all you need to set the time. Small cap = fast/big cap slow.

[Toney]

 A note:

I finally got around to building this on a micro-dot of Vero but I ended up using a vastly smaller cap - 180nF.
It seems my breadboard was adding some kind of conductance of its own. I build it originally as it was working on the breadboard and was puzzled as to why the LED faded over about five minutes 

So if anyone uses this, please start with a cap say 220nF and increase or decrease as required - works well.

[artifus]

Nice and simple, I like it. Might try it on my SHO.

ba-dum-tsh! 

this looks like it could be useful in an envelope detector circuit in an auto filter or whatever too.

[earthtonesaudio]

That circuit should not use a polarized cap.  A (smaller) non-polar cap together with a larger resistor in place of the 10k will retain the same time constant.

[Toney]

earthtonesaudio, yep a valid point, please see the post above yours.

Have updated the images as built. If anyone does this,
Just start with 22onF and increase/decrease as required to set delay.