Tremulus Lune - is this how it's supposed to work? Weird depth

[blackieNYC]

Tremulus Lune seems to be a good answer to my tremolo woes thus far, but-
The Depth Control - it's wacked!  Update - see fix below !
The depth is full strength thru most of the pot, then the signal gets quiet with the depth unchanged, right down to silence, then at the end of the pots travel toward the 9v, the signal burststhru at full volume with no trem.  A fine place to land but a rough way to get there. Is this how the depth is supposed to be?  Four 1k pots and two diodes.  I keep rebuilding that little circuit and get the same results. Checked my resistors, swapped a vactrol clone in place of my homemade one, same thing.  Whazzup with that?  Who's got a TL with a smooth depth adjustment throughout the range?
Sticking to this schematic exactly except the Smoothness is a 1m pot with a 1m resistor across 1&3, and I keep it fully turned this way it that, so it's 500 k
http://fuzzcentral.ssguitar.com/tremulus.php

[GibsonGM]

Are you sure the depth pot is a 1K linear taper pot?  And that your resistor values are correct after that? (the 1k's and 330...).    Sounds weird, man - mine had good control over depth, not erratic like you're describing...

[blackieNYC]

Yup - 330 to +9, 1k in parallel, 1k to ground.  I can remove my vactrol and move the depth control over to the indicator LED.  It behaves the same way in the scope - starts off at about 3.5Vpp, the pot turns it down to about 2v (the volume is getting quieter overall-the vactrol LED is dimmer when it's on so the ON resistance is greater -back to the vactrol now  - thus a drop in signal level as it passes from the first op amp to the 2nd? I guess?) and when the LFO gets around 2Vpp (there's almost no output now, a very quiet signal with tremolo -the led flashing is very dim) then at about the 8o'clock position on the depth, the 9vdc supply seems to take over, thru the 330r, and Bam- the volume is full up, just as when the depth is maxed - but with no modulation at this point.  It's as if the handoff between the LFO and the 9v supply isn't handled smoothly. Damn.  I feel like I should put a pot in place of the 330 ohm and see if.... I don't want to reinvent the bloody wheel. (Usually I'm down for that, but not with this)

Do you get a little volume drop as the depth gets shallower?

Is this the nature of the vactrol/led/LDR? The diode shuts off below a certain voltage?

[blackieNYC]

Can someone explain that LFO modulation - if the depth causes the LED to flash less brightly, the On resistance is higher, and the thing is quieter, no?  To keep the same overall  unity gain (on-gain, with the square wave LFO let's say), the 9v would have to be keeping this LED on, and the LFO would have to be dimming the LED, which is not this design.  My depth is a weird volume control.

(I left out the spacing control and used two 470k resistors.)

[GibsonGM]

To some degree, yes - I MAY have noticed a slight decrease in perceived volume at lower depth, but it wasn't anything extreme, IIRC.  I didn't 'keep' mine - never made it to the box, since I needed it for something else.   

I think that slight drop in volume is just part of the way these circuits work, Blackie....typically I run them damn near wide open, so it's never been a problem....maybe you could bring out the 50k trim pot to bring it back up manually, in the feedback loop of the second half of the TL072?     Don't see too many other ways to 'fix' that.

[midwayfair]

The LFO depth pans between the LED receiving a steady DC of ~9V* and the LED receiving an oscillating voltage centered around 4.5V. It is normal for 0 depth to be louder. If you want the 0 depth to be the average of the oscillating volume, you'll need to be able to zero the depth to a resting voltage of 4.5V. This might be more than you need, but I made an LFO that's based on the Lune but buffers the depth control from the LEDs and uses a reference voltage. See the first schematic here: http://www.diystompboxes.com/smfforum/index.php?topic=110779.msg1017347#msg1017347

*Okay, it's not really 9V. The anode is at a voltage divider formed by 330R and 1K (in parallel with the LED) + 1K (in series with the LED). But still, higher than 4.5V.

But it sounds like something else is not working with your build. You should have fairly mild but noticeable depth starting around 10:00, with a gradual change until it's at full depth, at which point it will be quite choppy. Things you might want to check: (a) What type of LED or vactrol are you using, and what's its current draw? A blue or white LED will have different depth resistor requirements than a red one. (b) Triple check your connections for the pot and the resistors. (c) What are the characteristics of your light dependent resistor? Are you using the NSL-32 the Lune was designed around, or did you roll your own? What's its dark resistance, rise time, fall time, and on resistance? Where is your gain set compared to the on resistance? (d) If you included all the mods, make sure your duty cycle and ramp controls are at dead center.

(Also, apropos to your ticking, if you're working on a breadboard, it's a good idea to isolate the power for your LFO and further filter it. Again, see the schematic I linked to above, or the Shoot the Moon version of the Tremulus Lune.)

[blackieNYC]

Will try the buffer tonight. Am I right that I should have my vactrol as your D2a?  Not D1a.

And I'll measure my vactrol specs tonight.  One is hand rolled, and i did try a vactrol clone from small bear. I think he carries an equivalent to the vactrol "3" and the "5". Not sure what that one is. I hoped maybe the LDR choice would be functional  at least.  Can I compensate somewhat by changing that resistor then? The 2k2 in yours, or the 330r in the Lune? I should buy the silonex anyway.

[midwayfair]

Will try the buffer tonight. Am I right that I should have my vactrol as your D2a?  Not D1a.

And I'll measure my vactrol specs tonight.  One is hand rolled, and i did try a vactrol clone from small bear. I think he carries an equivalent to the vactrol "3" and the "5". Not sure what that one is. I hoped maybe the LDR choice would be functional  at least.  Can I compensate somewhat by changing that resistor then? The 2k2 in yours, or the 330r in the Lune? I should buy the silonex anyway.

It doesn't really matter which LED you use from my schematic - the D2 LEDs (a and b) will use the op amp as their low voltage on the cathode, and the D1 (a and b) LEDs will use the op amp as their high voltage on their anodes. Regardless of which one you use, just use the other LED in series with it as the indicator so the indicator blinks off when the amplitude drops. You probably need to start with the same path to the supply rails, though, to balance things right (use both 2k2s). This was designed to have two out of phase LEDs, so I didn't do much experimentation to see about just omitting one of the pairs. It just works well for having the volume get both louder and quieter and making the volume sound more constant.

The Smallbear equivalents of the 5C1 and 5C3 have higher Fv drops than the PE parts, but they're very close when measured to the LEDs in the Silonex parts. They do have really low "on" resistance, so just watch where your gain is set; you could be clipping the dry path I guess. So you should be fine with what you have, though I don't know what's in your hand rolled.

[blackieNYC]

The cardinal V2 tremolo:
YEEEEAAAAAHHH!!!That's it!  And it makes perfect sense - your explanation.  It seems perfect! The max tremolo, you see your indicator LED pumping away, and as you turn down the depth, it doesn't get dimmer - the "off" cycles get brighter, until eventually it's on fully, and you have full volume.  There may be some very small drop in volume with depth, barely worth turning the gain knob for.  Middie - it's brilliant!

Vactrol specs - at the risk of sending the wrong message to the young folk, they don't matter. So it appears to me. To be clear, I'm talking about the Lune with just the buffers and LFO/LDR interfacing of the Cardinal.
On resistance  - doesn't matter, the gain knob makes up for it if the on resistance is a little high.  Off resistance - the LDRs I'm sticking in here seem to give plenty of depth when depth is at maximum. Some inky drawer might turn out to be shallower, but I'll bet they'll all work.
Rise time/ fall time - the speed range of the Lune is easily accommodated by whatever LDR I've got in there. If I were trying to get ring modulator sounds - audio band LFO speeds- it would matter. High speeds could peter out the oscillator.
LED current draw - it's not ticking, it hasn't blown up. Good enough.  In this particular case.  Important disclaimers regarding my laziness in slapping any old LDR / vactrol in there.

The Lune is highly esteemed, and the modulation's poor performance is probably my fault somewhere. That said, you must try the cardinal v2 LFO to vactrol/LED configuration. Looks like it could be added to any dual op amp LFO design.

[blackieNYC]

Update - the cardinal's lfo is working fine. A great improvement over the Lune, regarding the volume changes with depth changes. But there still is some volume drop. It's not excessive, I think due to the fact that a tremolo may give, perhaps, a perceived volume loss? Being half on, half off? - I didn't notice it in a quick listening, but there is a measurable difference between the audio level with the depth at zero vs the audio level of the ON stage of a slow square wave modulation.  About 8dB.  This doesn't seem quite right. The LED is brighter in the ON stage than when depth is zero. But again, in the lune I found the depth can almost turn the thing off.

[midwayfair]

Update - the cardinal's lfo is working fine. A great improvement over the Lune, regarding the volume changes with depth changes. But there still is some volume drop. It's not excessive, I think due to the fact that a tremolo may give, perhaps, a perceived volume loss? Being half on, half off? - I didn't notice it in a quick listening, but there is a measurable difference between the audio level with the depth at zero vs the audio level of the ON stage of a slow square wave modulation.  About 8dB.  This doesn't seem quite right. The LED is brighter in the ON stage than when depth is zero. But again, in the lune I found the depth can almost turn the thing off.

You should be able to adjust the gain of the make-up stage; you could also just increase the gain of the input stage. The exact resistors needed in the make-up gain stage depend on the characteristics of your photocell. Some cells don't have a very low on resistance, so they need a bigger "volume" pot. Some have a super low on resistance and anything above a 5K volume pot might be too much. If you're definitely above 1x gain with the photocell fully on, though, I'm going to say it's just the perceived loss inherent in almost any modulation effect. It helps to have some boost.

[blackieNYC]

The input stage - this is the Lune audio path circuit and not the cardinal - needed to be knocked down a bit - one of my guitars could clip it. Added 20k in series with the input 47k.  The gain is easily made up for in the output stage.  No problem there.
But regardless of variable gain, or the resistance of the photo resistor, if the depth is at minimum you want the LED at full brightness (the brightness you get in the ON cycles when the depth is at maximum).  Otherwise, you'll have this volume discrepancy.  I'd like to make it work that way.
I'm going instead to make the volume trim an external knob.  This will allow one to compensate for the gain difference with various depth settings.  Maybe this is why the Lune has this "it's a booster, too" feature.
Again, the Cardinal 's application of the LFO is superior, but can you take a look at yours?  You can audibly tell the volume change between minimum "zero" depth and the ON cycles of a low freq modulation at full depth, but a sine wave and a voltmeter will reveal the extent of the difference.
I think I can close up the pedal at this point, having added the gain control