LFO-synced LED on EA Trem - opposite phase?

[Mark Hammer]

Following up on a suggested mod from the runoffgroove gang (BT, et al), I replaced the 15k resistor from V+ to the LFO transistor on an EA tremolo with an 11k resistor and an LED.  Works great.  No audible tick (yet).  But unless there is something wrong with my vision, it really seems as if the LED gets brighter as the volume decreases, and darker as the volume increases.  If I was only interest in having a rough sense of the speed, that would be fine, but sometimes (especially at slower speeds) you want to sync yur playing with the effect and the LED would be more help in that regard if it got brighter at the right time.  Did I do something wrong?

[cpm]

when the LFO is up (led on) the FET is up thus eating signal into ground (decreasing volume)
so the led lights at the same rate as the LFO oscillates up and down, wich is inverse as the effected volume.

for what you want i guess it can be done using the LFO output to control the lgiht of the led in the same way it does with volume... interesting point though

[B Tremblay]

Mark, I just checked my build and sure enough, the LED pulse is opposite to the audio.  I'd never noticed that!

[R.G.]

No audible tick (yet). 

There shouldn't be. The current changes are all sine wave, or very nearly so, and cannot couple well through capacitances. The LED does not suddenly come on and go off to change the current. Hence, no ticks. The circuit is relatively insensitive because of this.

But unless there is something wrong with my vision, it really seems as if the LED gets brighter as the volume decreases, and darker as the volume increases.  If I was only interest in having a rough sense of the speed, that would be fine, but sometimes (especially at slower speeds) you want to sync yur playing with the effect and the LED would be more help in that regard if it got brighter at the right time.  Did I do something wrong?

Your vision is OK, and there is nothing wrong. This is a consequence of the way the JFET and LED work, as cpm noted. With an N-channel JFET there, the JFET is most off when its gate is most negative. Most off gives highest resistance and lowest gain to signal. This happens when the bipolar collector is lowest and hence the LED is brightest.

To fix this, you need to invert when the LED comes on with another transistor or such device, or to invert the sense of the JFET. The simple thing to do is to whip in an opamp to invert the signal from the transistor and drive the LED. A transistor to do this would probably take more parts to get right. There's probably something that could be done with diffamps and/or current mirrors, but again, more parts.

One oddity about the EA Trem I noted but never dug into much is that the JFET is tied to ground for a static bias point. The gate signal it's being fed is centered on ground nominally, but when the signal at the gate gets bigger than about 0.9V peak to peak, the gate conducts and clamps the signal to ground by bleeding charge to ground. That shifts the whole driving signal negative by charging the cap from the LFO. The JFET is fully on and the signal at highest gain for about half the LFO signal. It's only on the negative swings that there's any amplitude change on the signal. It sounds OK, but it's electronically curious. Maybe some Muntzing went on.