so i built this project i found on a french site...

[PRR]

> if i cover the circuit with my hand, i can hear the volume go up

Ultimately you need to make the LED/LDR area totally light-TIGHT. However it is handy to have it open for testing.

Yes the LED must flash, very obviously. Get that to happen and I think you are fine.

The transistor is NOT critical. I found a 1963 paper showing that hFE must be over 18, and >50 was more reliable, though >>50 becomes distorted. Since you can hardly buy a part with hFE<50 today, all you need is a "not-dead" transistor of correct polarity and known pinout. And your observed DC voltages suggest you have ample hFE (thus acceptable pinout).

220K:18K looks wrong (barely alive). 100K:18K (100K:22K is more exact) seems fine.

There are other ways to bias it, but your DC voltages look entirely suitable. Transistor is alive and collector is mid-zone of the power supply.

An oscillator is an amplifier and a feedback network. Your amplifier is fine. Look again at the feedback net, the three 470n caps and associated ~~18K resistors.

C4 must be "large". 100uFd is borderline, try 470uFd.

Usual follies: mis-wired, wrong-value part, bad solder, solder bridges. I prefer to confuse red and orange, but we all have our own favorite mistakes.

You can replace the R3 P1 string with an 18K, the ideal value. It won't vary frequency but right now you just want to see it WORK.

The R1 P2 string "should" also be 18K. I doubt 13.9K hurts; however it is really 9.9K here with loading from the buffer stage. You could change R1+P2 to 37K(39K).

BUT this circuit should NOT be this fussy. I'd still bet on a build mistake.

> I never got one to work in the sim.

It can be tough. Linear oscillators are not sure-to-start in an ideal world. In real life there is always hisss which will get it started, but it can take 1000+ cycles for oscillation to build. As it gets to decent level non-linearities slow the simulation.

Getting two damped cycles of ring does suggest insufficient gain; yet millions of these things work in a real world. Many are very crude.

> The spike of current from C5

Yes, it may be kick-start. In real life, if it is gonna oscillate, it does not need a kick, you just wait. For a 1MHz radio oscillator, as long as 0.001 seconds. For a 10Hz LFO the wait can be over a minute. Some guitar trems over-gain the oscillator for quicker start with distorted wave. Others use kick-start for near-instant start.

[PRR]

> Wouldn't bias from the collector effectively bypass the phase-shift network?

It actually works quite well. Note that the bias resistor will be on the order of hFE times Rc, or about 100*18K, or a couple of Megs. By itself, "large" compared to 18K network values.

But if we assume the amp's voltage gain is similar to hFe, by Miller Effect the bias resistor acts-like around 18K, and serves as a final R in the C-R-C-R... network.

See Q3 in http://www.generalguitargadgets.com/pdf/ggg_eat_sc_improved.pdf

You may steal this plan verbatim, splicing to your LED buffer at A-GGG's C4. 470nFd or 1uFd is a minor difference, use the 470nFd in your hand for now.

[pinkjimiphoton]

wow, thanks guys for all the info!!
gonna go play with this in a little bit, and try to implement some changes to get it to flash. if i can get it to flash, i know it will work, and it should sound pretty good i think.

while you guys were helping me, i was researching some stuff too.. including a simple two led flasher circuit i found..

http://www.hobby-circuits.com/circuits/led-and-light/light-flasher/770/two-flashing-leds-circuit-schematic



i was thinking maybe if i used two different led/ldr's and two different colored led's it may be interesting to have them flash at slightly different rates...

ideally maybe get 'em to be slightly staggered to make it almost a phaser kinda wobble, if that makes sense.

time to get nuts... again...

this reminds me of d'astro's circuit a bit.

[Keppy]

Thanks for the info Paul!

[petey twofinger]

i used that circuit in my pulse wah .

[pinkjimiphoton]

still not working.

swapped out the 220k/18 k for 100k and 22 k as paul suggested, also swapped out the 100u cap worked my way up to 1000u.
same thing.

presently have mpsa18's in there, so it's definitely got gain. but the light stays on. i've checked the vero over and over, and it's consistent with the schematic.

i'm freekin' stumped!!

gonna tear it down i guess, and start over. looked like a really fun, simple little circuit. wtf!!

[~arph]

You should really invest in a breadboard  ... It's like vero! But temporarily

[pinkjimiphoton]

i have one. it's lonlier than a maytag repairman usually tho. can never seem to get ANYTHING  to work on it!! lol...

[pinkjimiphoton]

success!! on the breadboard, anyways...

went with a 500k linear speed pot, 1000u cap, and a plain green led which seemed to work the best.
i believe it comes down to operator error.. the pots as shown in the schematic are backwards...or at least backwards to what i was expecting. so turning the intensity pot (10k) up the led was...predictably...always lit. 

i'm gonna try one more time to get the vero i'd laid out working. it should work, i've looked at it til my eyes were dripping down my face (thanks to soundman rob for the psilocybes, lol) and i can't see any freaking reason whiy the layout i did shouldn't work.

that said, really nice design!! not getting any ticking at all, quiet clean and very sweet tremolo. worth a build if ya wanna super simple trem with a nice sound to it that doesn't suck all your tone away or tick.

lovin' it...

thanks everybody!! especially paul, for explaining the circuit so well even a newbe like me could understand!

[PRR]

> a simple two led flasher circuit i found..

The phase-shift oscillator makes a sine wave. Very smooth. Like linking a slide-pot to a bicycle crank (or engine piston).

The flasher is square-wave. HARD on/off switching.

Go ahead and build the flasher. It's fun.

But a dozen designers have sold a million trem-amps using semi-Sine wave modulation, even when it was not easy.

Yes, sine has "been done" and an on/off stutter may be a useful new tool.

And you can build them into toys, control panels, stuffed bears, fake security cameras.

Back when you could get LM3909, I liked to use them as burglar deterrents on classroom hi-fis and for back-of-the-building office windows. One D-cell would flash an LED for 2 or 3 years non-stop. An office and a classroom-hifi that had been repeatedly burgled, and a new hifi cart that begged for trouble, never got attacked once I blinkered them. The cart's LED had a sticker HIGH VOLTAGE, and nobody ever even mis-plugged a plug.

> have them flash at slightly different rates...

That 2-transistor flasher, both LEDs must flash at the SAME rate. (One on, one off.)

The dual pot is a mild annoyance. You can vary just one but the on/off won't stay 50:50. There's ways to be symmetrical with one rate pot but I don't have those notes here.

You sure can build multiple 2-transistor flashers and tune them different.

However if they are very close in rate (which may be where it gets interesting, at least to spacers), you may find they fall into sync. Hefty supply cap is a first fix. Separate R-C power filters can be necessary.

[~arph]

success!! on the breadboard, anyways... newbe like me could understand!

See... breadboard. 

[pinkjimiphoton]

> a simple two led flasher circuit i found..

The phase-shift oscillator makes a sine wave. Very smooth. Like linking a slide-pot to a bicycle crank (or engine piston).

The flasher is square-wave. HARD on/off switching.

Go ahead and build the flasher. It's fun.

i was thinking, if i can get the offset between the flashers right, it may sound a little wobbly... may be no more than twice the speed of a single trem tho.

But a dozen designers have sold a million trem-amps using semi-Sine wave modulation, even when it was not easy.

Yes, sine has "been done" and an on/off stutter may be a useful new tool.

And you can build them into toys, control panels, stuffed bears, fake security cameras.

that's kinda what i was thinking... if i could set the pots to flash opposingly, but slightly off time from each other, it may make a pretty cool sound... a hiccup, or stuttering kind of trem sound?

Back when you could get LM3909, I liked to use them as burglar deterrents on classroom hi-fis and for back-of-the-building office windows. One D-cell would flash an LED for 2 or 3 years non-stop. An office and a classroom-hifi that had been repeatedly burgled, and a new hifi cart that begged for trouble, never got attacked once I blinkered them. The cart's LED had a sticker HIGH VOLTAGE, and nobody ever even mis-plugged a plug.

love it!

> have them flash at slightly different rates...

That 2-transistor flasher, both LEDs must flash at the SAME rate. (One on, one off.)

pretty sure the one i'm looking at (may be different from what i posted) has a 10k pot to independently control each flasher...a long time ago, i used to use two tremolos and make 'em duke it out, and you could get some VERY psychedelic sounds out of it!! the key would be i imagine to have each trem flash at different rates, and then make a master oscillator to control the overall speed of the two of them, with whatever offset you come up with simultaneously.
is there a relatively easy way to do that do you think? seems like it could be pretty cool.
what i really liked about using the two trems (one was regular amp tremolo, the other was a little out board spring reverb unit, i think made by "marlboro"...it had reverb, trem, and "echo"...with the echo being tremolo'd reverb. when you got the two trems going the right speed, it made it delightfully easy to mask your note attacks and start sounding very frippish, even kinda backwards. it was trippy as hell!

The dual pot is a mild annoyance. You can vary just one but the on/off won't stay 50:50. There's ways to be symmetrical with one rate pot but I don't have those notes here.

You sure can build multiple 2-transistor flashers and tune them different.

i'm thinking 2 of them with 4 stages total would be the tremolo version of a phaser. may be worth exploring!!

However if they are very close in rate (which may be where it gets interesting, at least to spacers), you may find they fall into sync. Hefty supply cap is a first fix. Separate R-C power filters can be necessary.

thanks for the advice, paul!
one more thing to think about. tremolos are cool, i wonder how rg is making out with the kustom idea?

[pinkjimiphoton]

success!! on the breadboard, anyways... newbe like me could understand!

See... breadboard. 

lol...rub it in, arph!! lol

[Keppy]

that's kinda what i was thinking... it may make a pretty cool sound... a hiccup, or stuttering kind of trem sound?

Funny, that's one of the flaws I had to get rid of in the last trem I built.

[pinkjimiphoton]

WELL, there's an ass for every seat, they say!!

[duck_arse]

point a second ldr at the tremolo led, and use it to variable rate a second trem oscillator. dunno how you'd do it, but it sounds as though you'll try .....

[pinkjimiphoton]

needs a completely second, independent circuit to do that.

[pinkjimiphoton]

wanted to say, thanks for the support guys, especially paul.
i got the damned thing working. it works really good!! i was pleasantly surprised, no ticking, at least none i can hear thru my ruby.
i took paul's advice, and made the 100u electro 1,000u, and the 50k rate pot 500k..also made the 220k/18k votage divider 100k/18k, worked perfect... gives me a real nice range of speed, from a slow, almost "vibe" kinda pulse to fast enough for miserlou. i may add a switch to go between the 1,000u and maybe a 47u, so i can go from super slow to mosquito speeds.
when all was said and done..
i found the problem...after breadboarding, and building a completely different circuit..(i mean, the same circuit, different layout without the power supply filtering)..

it wasn't the wiring, it wasn't the components, it wasn't even a gosh-dang solder bridge.

THE MOTHER @#$%ING BATTERY WAS DEAD!!!!!!!!!!!!!!!!!!!!!
AY YI YI YI YI!!!
so...the DUMBEST mistake of all.
when in doubt, make sure ya gotta decent battery or power supply!!


this circuit is cool! it's tiny, works great, if you're bored, check it out!
it's buffered, and even has a slight boost when engaged... very nice!!

[duck_arse]

the 2 hardest parts of any build/building:

1. is it turned on?
2. is it plugged in?

I've fallen fowl to both this week.

[pinkjimiphoton]

lol... it looks like there was *just* enough voltage to light the led.

ay yi yi...

this thing sounds freakin' great!!

iit's a lot like d'astro's tiny trem, but without the issues.

i went for it cuz it looked EASY. and it was. once i got a good power supply!!
lol