De-Ticking a Shaky Jimi Univibe - help please

[52 Bill]

The resistor didn't seem to help much either way.

I'm confused here, did you try installing both the resistor and the cap together at the LFO chip or just one or the other?

[Mark Hammer]

Yep, the two need to be used together.

[Regis]

I'm confused here, did you try installing both the resistor and the cap together at the LFO chip or just one or the other?

Um, I put the resistor directly after the power jack before it feeds anything on the board. I cut the trace and bypassed it with the resistor. I put the cap on the chip. I'll move the resistor to the chip as you instruct and let y'all know what happens.

[Mark Hammer]

Basically, what yu need to do - rather, ALL you need to do - is locate pin 8 on the chip used for the LFO, neatly cut the trace that runs up to is, so that pin 8 is isolated from everything else but has enough copper arund it to solder to, bridge the cut you made with a 100R resistor, and run the cap from pin 8 to gnd like I assume you already did.

What you are essentially trying to do is give that chip it's own individual little "battery" that's good for a few milliseconds.

[J M Fahey]

after the power jack before it feeds anything on the board

Then they are all still rowing in the same boat.
Worse, becuse the now higher PSU impedance now makes everybody even more dependent on each other.
Mark Hammer's solution *separates* the LFO from everybody else and is the right one.
I'd add another 10uF to ground on the other end of the 100 ohm resistor to turn it into something like a "Pi" filter, which will slightly improve things.

[Regis]

Hi guys,

I'm a little late replying here, I had some computer problems this week, then yesterday I kept getting "Database Error" from the forum.


I followed instructions and now the thing is dead quiet. Well, the tick is gone anyway.

I had already installed the 10uf cap correctly, but apparently the resistor was in the wrong place as was pointed out to me.

To isolate the LFO chip I had to cut two traces. One to insert the 100r resistor between the LFO chip and 9v feed, and another because the 9v feed for the other chips was attached to the LFO feed. I cut through the trace and ran a wire around the resistor and LFO chip to the first +9v pin of the chain.

So now the LFO chip has it's own +9v isolated feed going through the 100R resistor as well as the capacitor between +9v and the its ground pin, and the tick is gone. Now I sound just like Robin Trower, ha ha

Thanks for everyones patience and help, as always it's been a learning experience at MEF.

I looked around the internet and found a couple of ideas:
Shielding the individual LED/LDR combo's with tape or shrinkwrap, and increasing the input cap value.

Does anyone has any opinion on either of these?

I wonder if the light from the bias LED's affects the LED/LDR combos, and would shielding help?


You guys are great as always, thanks!

[Mark Hammer]

"Houston, we have liftoff." Always a great phrase to hear.

Re: LED/LDR combos

the ideal is an epoxy-encapsulated opto-isolator with the LED and LDR inside. They're kinds pricey, though. Not unreasonable, but they do raise your production costs, so some companies avoid 'em. And in some cases,the design simply points towards use of one LED to illuminate multiple LDRs, or a particular LDR, for whatever reasons. In those latter cases, there is the task of keeping extraneous light away from the LDRs.

On the other hand, the design itself can be predicated on the LDRs having more resistance swing than is needed for pleasing musical sweep, and the resistance drop introduced by light leakage into the box does not interfere with that.

[J M Fahey]

If it's one on one, I always shrink wrap them.
If you have, say, 4 LDR around a single lamp or LED; I cover them with a small plastic cap from a spray can, painted white internally and black outside.