LED trick

[smoguzbenjamin]

Check this out:


I tried this with a BC547 but the first LED fried (stinky :evil: ) so I popped in a 10k limiting resistor. But now it's just dim and flipping the DPDT toggle switch didn't do diddly squat. I'm using this in my ggg mini-mixer to turn the volume boost (LPB2) off/on.

What am I doing wrong?

[javacody]

According to the article, you must adjust the resistor to your tranny/led. I would start with a much lower value than 330K, and work your way up to the brightness you want. Or put a pot in there.

I don't think that putting a limiting resistor in between the emitter of the transistor and the led will work.

[Transmogrifox]

Actually, if the LED is too bright (or burning out) increase the 330k to 1 Meg.  If that isn't bright enough, continue to experiment with values in between.  I suggest using a 1 Meg pot to adjust the circuit to the desired brightness, then take the pot out, measure the resistance and replace the 330k with the value that you measured from the pot.

The Vox Valvetone (and many other true bypass pedals) use a similar circuit.  The V810 DOES use a limiting resistor in series with the LED as you tried.  The key is to pick the correct value.  If it's dim and flickering, make the limiting resistor smaller.

Another useful thing to do is to add a buffer stage to the output of your effect and set put a 1k - 5k ohm resistor in parrallel to ground AFTER the output coupling capacitor.  This decreases the popping switch problem.  That is the biggest problem I have had with this true bypass circuit trick.  

I have resorted to DPDT latching relays and a relay driver that does not drain current until the switch is set so that the bypass network can be completely independent of parasitic DC voltages.  I have found that this is much less expensive if one uses a cheap SPST or SPDT stomp switch.

DPDT or 3PDT stompswitch:    $6.00
cheap SPST + Latching relay:  $3.00

email me @  transmogrifox@yahoo.com if you want a relay schematic.

[javacody]

Doh! Of course the resistor should be larger. Jeez, I could smack myself sometimes.   :oops:

[smoguzbenjamin]

I think an impedance-lowering buffer wouldn't really help and it whould only use up space. The output impedance of my LPB2 is (I think) lower than 1k so that might not be the problem. BTW read "input jack" in the diagram as "mixer output". The mixer does have a 1M pulldown resistor to ground so that means high output impedance, I think, but it works fine driving my Boss effects and home-made EA tremolo...

I'll try changing that resistor. Thanks, I'll try that tomorrow if I get half a chance.

[Transmogrifox]

The AC output impedance is irrelevant to the "switch pop" problem.  That's why I said add a 1k in parallel to the output.  The reason for the buffer is that you can design it so that it has no problem driving the resulting 1k load.  

Usually if you have a 100k to 1 Meg pull down resistor on the output, the DC voltage resulting from the resistor divider on the LED driver charges up the output coupling capacitor, then when the switch is actuated, the coupling capacitor discharges into the signal path, causing a "pop" to occur.  If the pull down resistor is replaced with a 1k or so, the DC voltage in the "off" position is negligible.  In your case, it's a (1k)/331k ratio of the supply voltage...about 9 divided by 330, less than 3 mV which is most generally an inaudible pop in most systems.

this is all I was talking about.  If you don't have switch popping, then don't worry about it, but it's a useful thought if you notice a loud pop when you switch the effect on.

[Joe Davisson]

The circuit is a bit of a hack requiring the right transistor. It really should be a "real" 2N3904/2N3906 (marked with those numbers, not "2N3904-type" sold at RS), since the internal resistance is high enough (& can handle more power without heating than say a 2N5089). A high-brightness LED (i.e. http://www.superbrightleds.com) is preferrable also. Good Luck

-Joe

[smoguzbenjamin]

I'm using high-efficiency LEDs. Ah heck I'll just go out on tuesday and get the right transistor, can't be bothered to try resistances at will etc etc.

I'm gonna put my mixer and EA tremolo in an old amp chassis while I wait for my enclosures!

[Transmogrifox]

A Relay trick:

http://www.geocities.com/transmogrifox/relay.html

[RJ]

The Joe Davisson LED trick is really cool....I've used it successfully on a couple home-made Big Muff style pedals and it works great.  I'm having a problem getting it to work on a LM386 based circuit though, something like Aron's Smash Drive pedal.  The LED lights but won't turn off when bypassed....I've tried various resistors and transistors but I can't make the LED go out....will this LED trick not work with a LM386?

Thanks in advance for any suggestions, and thanks for posting the LED  schematic!

-Ryan

[Transmogrifox]

I've tried various resistors and transistors but I can't make the LED go out....will this LED trick not work with a LM386?

The first thing that comes to mind is that the Smashdrive does not have a pull down resistor to ground on the output.  The reason the LED turns off is because the resistance to ground on the pedal output creates a resistor divider that pulls the transistor bias too low to be on, thus turning off the LED.  Try sticking a 100k volume pot on the output AFTER the output coupling capacitor...(so if you were to put one probe of a DMM on the output and one on ground, you would measure something between 0 and 100k ohm, depending on the setting).

If you are using a device that does not have a capacitor on the output (10uF is good for most purposes), put one there, then reference the output to ground as above.

[smoguzbenjamin]

Pulldown resisors are always good. Bah, got to go to school tomorrw But then again I can get my transistors

[RJ]

Thanks Transmogrifox, a coupling cap and a resistor to ground made all the difference!

-Ryan