Nurse Quaky LEDs.... Matched how?

[AllyP]

On the Nurse Quaky scheme on runoffgroove it says that the LEDs should be matched....

Does this mean the same colour/size/model?  Or is there more to it than that?


Thanks for the help AllyP

[B Tremblay]

Same color and size will get you in the right neighborhood, but measuring their voltage drop with a DMM is the best method.

[AllyP]

Cheers for the email aswell :wink:

I take it the voltage drops need to be in the same neighbourhood, but how big is that?  +-10% ?

Cheers again,

[B Tremblay]

I chose a pair that were within a few hundredths of a volt.

[Mark Hammer]

They really don't need to be that matched actually.  LEDs of the same colour are sufficiently different from regular SI or Ge diodes, and sufficiently like each other that if the second one is somewhere in the ballpark to subtract what the first one alters you'll be fine.  Whatever difference remains is easily compensated for by the sensitivity control and the trimpot.

One of the problems the LEDs are intended to address is how much voltage swing can be provided by different op-amps.  The Dr Q circuit used a 1458, and when other op-amps are subbed for that in the stock DQ circuit the amount of sweep is often ridiculously large and this is disruptive to playing.  The compensation circuit that Jack cleverly devised provides the appropriate amount of sweep regardless of the peculiarities of the op-amp one uses.  

Please note that the dual-LED add-on is NOT intended to address things like op-amp noise, battery life, high-frequency response, etc.  On the other hand, since it frees you up to use any op-amp you want, it *allows* you to make choices with respect to those additional considerations rather than find yourself stuck in 1458-land.

[AlexK]

What would radically mismatched LEDs do to the sound?  I'm still having problems with my nurse quack with loud popping sounds slightly after the initial attack.  I've traced out the circuit 3 times and it's definitely wired correctly.  I used 2 unmatched red LEDs.

Thanks,
Alex

[Mark Hammer]

I've built 3 of them with no attention whatsoever paid to matching LEDs and never experienced anything like that.

One possibility is that the attack resistor is too small a value.  I generally wouldn't go much below 47-68R there.  A very fast attack will produce a broader and very abrupt-sounding sweep that can sound like a burp or pop.

Another possibility is the perennial favourite - the bipolar transistor is not installed correctly.  While you may have used a transistor that can *functionally* substitute for the 2N3904, it may have a different pinout.  Finally, if the LEDs are installed correctly, one of them (the one at the input to the op-amp) should always be on, and the other one should turn on briefly during peaks (it serves as a terrific envelope strength indicator).  If neither of these things happens then perhaps the LEDs are not oriented properly.

[AlexK]

Thanks for your response Mark.  Unfortunately, I've checked all that stuff.

My attack resistor is 50R + 1K pot.  Even with the pot all the way on, it doesn't get better.

I'm using a 2n3904 and I have the correct pinout for it.

The LEDs operate like you mentioned also.  What really frustrates me is, I re-built another one on a breadboard and it works awesome.  I'm trying to get around re-perfing this circuit.  I've pulled out the bipolar from the perfed circuit and tried it in the breadboard and it was fine.  I did the same thing for the opamp and the rectifier LED.  AHHHH!!!  I think this guy's ready for the sledge hammer.

Thanks,
Alex

[Mark Hammer]

Have you considered demonic possession? :twisted:  :wink:

[AlexK]

Actually, yeah.  The priest was in yesterday and he said it looked ok.  I was thinking it might be gremlins, but I never let my Mogwai get close to water or feed him after midnight.  Hmmm...  

-Alex

[gez]

Have you considered demonic possession? :twisted:  :wink:

Wouldn't that just make things worse???