Need LFOs that will drive a rate LED *brighter* as the depth is turned down

[midwayfair]

This is related to my harmonic tremolo thread, but it seems like it's general enough to put in a different topic.

A basic phase shift oscillator can drive an LED in series from the 9V to collector, and the LED gets dimmer as you turn down the depth control. Like in the Magnavibe:
http://4.bp.blogspot.com/-Ty3MQgwVUrA/TxQZTr8oQYI/AAAAAAAAAkA/o9f18Xyru1k/s1600/Bigfoot_FX_MagnaVibe.jpg
(It's a big picture.)

I use that series LED trick to add rate indicators to lots of pedals, like the EA Tremolo.

Is it possible to reverse that, so that the LED gets BRIGHTER spends more time on when the depth is turned down with a basic phase-shift oscillator?

Alternatively, what are some other LFO designs that could do what I'm after? I know that a TL072-based oscillator like in the Tremulus Lune (and various similar LFOs like in Rob Henry's Lightwah, CultureJam's Shoot the Moon, etc.) does it, as does the TAPLFO chip. Is there anything else I'm missing?

[samhay]

Short answer is probably yes, but I guess I'm confused - the depth control is only there (at least in the schematic you linked) to determine how bright your LED gets right? So, what else does the LFO have to do at the same time?

[midwayfair]

Short answer is probably yes, but I guess I'm confused - the depth control is only there (at least in the schematic you linked) to determine how bright your LED gets right? So, what else does the LFO have to do at the same time?

When you turn that depth control down, the LED spends less time on, eventually spending all its time off if you use a big enough depth pot. At the point just before it stops oscillating, at minimum resistance (i.e., max depth), it's about 50/50 on/off. This is fine for some applications; in the Magnavibe it means that you get no signal from the emitter when the LED's off and half and half at max.

I need the LED to do exactly the opposite -- to spend all or most of its time "on" at one extreme and spend less and less time on as the depth control is turned (until it's 50/50 like above). As an example, the Tremulus Lune's LFO does this.

I'm beginning to suspect it's not possible with a simple phase shift oscillator like that. I feel like I've plugged the LED into every hole possible on the breadboard.

EDIT: Also tried to clarify in the OP.

[Jdansti]

I haven't had a chance to research your question yet, but my first thought is to use an add-on circuit that provides the inverse response. Would a second LED/LDR pair with the LED connected to the first LDR do the trick?  Sorry I don't have time to draw it up but can do so later if necessary.

[midwayfair]

I haven't had a chance to research your question yet, but my first thought is to use an add-on circuit that provides the inverse response. Would a second LED/LDR pair with the LED connected to the first LDR do the trick?  Sorry I don't have time to draw it up but can do so later if necessary.

The resistance on LDRs gets higher as you decrease the light on them. There's no way to hook up an LED so it gets brighter as the LDR's resistance increases. (This is ignoring the fact that I'd really like to avoid a third LDR).

I'd like to believe there's something basic with a second transistor that I'm missing, I just can't think of what it is.

Um, is my brain really this fried? Could it be this simple: tapping the collector, putting a cap and 1M pot to the gate of a FET, connect the drain to 9v, and the LEDs in series with the source to ground? I.e., the EA Tremolo's depth control, with the FET between 9v and ground instead of on the gain control?:
http://www.generalguitargadgets.com/diagrams/eatrem_sc.gif?phpMyAdmin=78482479fd7e7fc3768044a841b3e85a

Will that work? As the depth is turned on way, the resistance between the gate and drain of the FET becomes less, lighting up the LEDs more. As it's turned the other way, it would increase the resistance and the LED would get dimmer. That should reverse it. Am I thinking right?

Crud, stuck and work and now I want to play with my breadboard ....

[Lurco]

R.G. would probably do it with a PIC!

[slacker]

Yeah I think that should work, might need a resistor in series with the LEDs as a safety measure.

This might be a dumb question but in the magnavibe example couldn't you just wire the depth pot backwards?

EDIT: ignore me of course you can't cos the LED is also driving the LDR.

[midwayfair]

Yeah I think that should work, might need a resistor in series with the LEDs as a safety measure.

Excellent! I'll try it out tonight on the BB.

[Gurner]

R.G. would probably do it with a PIC!

I'ts certainly the way I'd approach it  ...IMHO LEDs are most flexibly controlled via PWM    ... the plethora of "My LED doesn't turn off" or my "LED doesn't respond how I want it to" type posts seen round these parts, are often related to LED biasing/voltage incompatibilities for the circuit they've been speculatively dropped into (not all LEDs are created equal) - also there's a big component count 'win' controlling an LED with PWM (PIC).

So IMHO a $1 PIC is certainly the smallest (& controlable) hammer to crack this particular nut!

[midwayfair]

Sorry, what's a PWM or PIC? Can someone link to some reading material?

[Gurner]

Sorry, what's a PWM or PIC? Can someone link to some reading material?

PWM is pulse width modulation , which is a bit of a mouthful....hence abbreviating it.The easiest way to grasp what PWM is all about is to imagine an LED that you control with a switch.... flick the switch on, the LED will be at full brightness, flick the switch off, the LED will be at zero brightness, switch it on & off equal amounts of time (rapidly), the LED will appear at half brightness (not actually technically correct as our eyes aren't linear) & so on...in other words, the apparent brightness is controlled by switching it on/off very rapidly at different ratios between 0% (fully off) & 100% (fully on). LEDs can be controlled more predictably via PWM, because the driving current through the LED remains the same throughout the whole PWM sweep (0% thru 100%), whereas trying to control an LED by changing the supply voltage via a series limiting resistor is a more hit & miss affair.

PICs are essentially a computer on a chip (CPU, RAM, IO)....they are cheap (some are less than a dollar) & nearly all come with a PWM module which makes driving LEDs easy - they also normally come with an AtoD module, which makes driving an LED directly proportional to an incoming analogue voltage (eg aswinging LFO voltage) a breeze.     (or any other variant of - inversely proportional etc)

[midwayfair]

Turned out to be much simpler than I though to do it with the simple phase oscillator. I just made the depth pot a voltage divider and increased it to 25K:

Lug 1 to ground
Lug 2 to the cathode of the rate/Vactrol LDR
Lug 3 to collector of the oscillator.

Works great in the Harmonic Tremolo with a simpler, sine wave LFO. I need to test it in the Blue Warbler now.

Edit: Here's a schematic. I'll post it in the other thread, too.

[Jdansti]

Glad you found what you were looking for!  I'm posting this just to spur some additional thoughts (for example, "only a nut would propose this! ).

This is what I was thinking of, but I'm having a hard time obtaining a large voltage difference at the new LED as the LFO's LED cycles on and off. Maybe the resistors can be tweaked further or some other components thrown in to make it work better.

On the other hand, this might not be feasible at all.

[newperson]

Neat.

Turned out to be much simpler than I though to do it with the simple phase oscillator. I just made the depth pot a voltage divider and increased it to 25K:

Lug 1 to ground
Lug 2 to the cathode of the rate/Vactrol LDR
Lug 3 to collector of the oscillator.

Works great in the Harmonic Tremolo with a simpler, sine wave LFO. I need to test it in the Blue Warbler now.

Edit: Here's a schematic. I'll post it in the other thread, too.

[samhay]

Cool - that's a very elegant solution Jon.
This thread has got me re-evaluating the LFO that will end up driving my optical phaser - although I suspect I will end up with an op-amp implementation. I guess I have tended to discount this type of PSO in the past, but it looks like it is more than good enough for the job.