leds in loops - up and down?

[duck_arse]

I have a question about driving the led in the opamp feedback loop method:



in circuit A the led goes brighter as the input V rises. so if I want the reverse, and the led dimmers as the Vin rises, do I reverse the led and the supply to it?

further question without notice - if "yes" above, can I then apply a buffered V/2 to the supply end of both resistors, and drive both Vin's with a rising V to get one led dimmer, then one led brighter? and a dead spot in the middle?

[anotherjim]

I think it would work. You may only need one op-amp with LEDs anti-parallel? To get a dead spot in the middle, because the op-amp is hiding the LED Vf, you might need anti-parallel diodes (type Vf according to width of the dead bit) towards your Vcc/2.

[electrosonic]

In the first circuit I think the current through the LED will be Vin/R. I like putting LEDs in a op amp feed back loop because you can pretty much ignore the forward voltage of the LED. i.e. the current through the LED is just proportional to in the input voltage. You would need to use an op amp like a LM358 that can operate down to zero volts I guess.

In the second example the current through the LED will be (Vcc-Vin)/R. You would need an op amp that operate with its inputs at Vcc.

I think you could have anti parallel LEDs in the op amp loop and have the end of the resistor referenced to Vcc/2. The impedance of the bias point would have to be low enough that the current supplied by the op amp does not shift the bias point though.

Andrew.

[Blitz Krieg]

like Z2-3 in this?

[anotherjim]

You mean Z2-4 Blitz? Z2-3 is the envelope rectifier. The way Z2-4 works is neat with the polarity switch, but it looks like Stephen wants 2 Vactrol. An up alternating with down from the same control without any switching?

[duck_arse]

very perceptive, jim (another). I had been working with this:


for the colourless version

but, that was a few days ago. I was thinking of doing something different on the right hand side, reference volts wise. meaning, more opamps.

[I notice the mutroxx has two "Z2-4", by the way. the first one did confuse me. the second one, too.]

[anotherjim]

On that scheme, If IC1b was a unity inverter, working off IC1a output, it would work the LED's push-pull. As pin 1 rises, pin 7 falls & vice-versa.

[duck_arse]

perhaps so. wouldn't it still need a buffered RV1?

opamps aren't my strong suit^*, I'm afraid, I had trouble working out the 1+ gain in a non-invert on the w/end.


^*spades is.

[anotherjim]

You'd still need a bias on pin 5, but it could probably be fixed. IC1B is still acting as supply, but will move to double the swing on the LED's in between . Say 10k pin 6 to pin 1 and 10k pin 6 to pin 7 makes IC1b unity inverting. Then dead zone is still going to be about 3v between a total swing of about 7.5v (if the envelope can reach 7.5). If pin 1 reaches 7.5v, pin 7 will be 0v & top led conducts. At around 3.7v (reference on pin 5), both outputs are the same -  no LED lit. If pin 1 = 0v then pin 7 = 7.5v & bottom LED lit.

If you want to include LED in the feedback with an amp wired as an inverter, you can't just put the LED anywhere in its feedback loop, but you can put it immediately on the output pin before the feedback resistor and take output from that junction - like the envelope rectifier in that Mutroxx synth above.

[duck_arse]

an update. I can answer my original kwestion with "yes". the wire it backwards approach works as expected.

but I didn't attempt Q2.

[duck_arse]

another update - original qwestion two answers to "yes". and here, for anotherjim is some circuit. some problems solved, some caused.



notes on request.

[anotherjim]

I dunno...
..erm, what problem caused? In the l/h version, is the inverting amp spending to much time in the ground?

[duck_arse]

ah-hah! it may be too long at ground, but the other end is still swinging with the envelope. I think the left works better after some tests. I was using it with different ldr's and diff leds, so R103 become two resistors to tailor. also, a reversing switch can be added between the 2 opamp outputs.

on the rhs, the opamps go squarewave silly when the led is trying to minimum brights. and is more parts. but the LED203 re-reference was a fun addition, I thought.

[anotherjim]

Staying on the left... I think I'd give the first amp some gain instead of the 2nd inverting one (which should just be the other end of the see-saw). Non-inverting, it would need 2 extra resistors with the input one DC coupled to reference. Don't forget, non-inverting, there is already always x1 gain, so gain resistors equal gives you x2. Feedback R could be adjustable, can go to zero for unity gain, to let you tweak max swing from the envelope. Inverting amp would be unity gain & same reference on input R.

...so... adjust the envelope swing with the gain of a first non-inverting amp. If you need to trim the maximum brightness, make R103 variable with a small fixed minimum R. To tweak LED's individually, have a variable in series with each LED. To tweak balance of the LED's, use the variable like a blend control between the LED's.

As to the r/h scheme oscillating, may only be breadboard noise - there's no feedback loop I can see to make it oscillate. Possible it was just jitter due to noise/hum. Again that could go away when hardwired & boxed, or maybe needs some C across R205?

[Kipper4]

Heres a quick Scheme I came up with for opposing leds. As 1 goed dark the other lights up

It will need some tweeking I'm sure