Anderton Tremolo: clicking?

[Bucksears]

I know it's probably just the nature of all photocell trems to do this (my SRRI does), but does anyone know how to remove the ticking/clicking sound in a tremolo?
I finished up with the Anderton Tremolo and it sounds great, but I'd like to remove the noise if possible.

Thanks,
- Buck

[Mark Hammer]

That 4049 does a triangle wave thing, but it also does a square wave thing.  Every time the output swings rapidly in the other direction to the other extreme, it craves current.  Where does it get that current? From the same damn battery as the dual op-amp.  What happens when it grabs that current?  The op-amp hiccups.  This has nothing to do with the optoisolator (well, in a sense it does because the LED wants current too).  How do you overcome that?  You give the op-amp its own little "uninterruptible power supply" to tide it over for a couple of milliseconds.  How do you do this?  By having a small value resistor and separate storage capacitor dedicated to that op-amp, so that whatever happens elsewhere in the circuit, the op-amp has a reservoir of stored current it can draw on.  This is generally referred to as decoupling.

In the schematic ( http://www.generalguitargadgets.com/diagrams/catrmsc.gif ) you can see a 1k resistor and 100uf cap to the 4049 (U2).  Run a separate 33R resistor and 100uf cap to ground to the V+ pin of the op-amp (pin 8 of U1) and that should, in theory, fix it.  Under other circumstances, I would suggest the same thing for both the + and - supplies, but since the 4049 is only using the + side, it is primarily the sudden draws on the positive side that are creating the clicking.

Decoupling is something that can be targeted or general.  Clearly U2 has been allocated its own little current reservoir, but making life easier for U2 does not always imply life is dramatically easier for U1.  Conversely, if U1 had the suggested RC network but U2 didn't, you might still get the ticking.  Giving everyone their own big glass of current to drink from means that no one gets thirsty.

make sense?

[markphaser]

I think this is a Current Draw issue because its robbing the current from other components like op-amps or transitors which draws the most current because they are active components. So the LFO draws more current from the LED and lamp the op-amps get lower current. Since the op-amps get lower current and the LFO gets more current you hear the ticking because the LFO has greater  potential current than the op-amps or the op-amps ampilify the LED or lamp turning on and off sound/noise. That turning off and on sound/noise can run or bleed through the power supply lines which power up the same op-amps for the circuit so thats how it ampifiys.

I'm taking a guess don't quote me on this

[A.S.P.]

phototrems can be the least clicking trems due to their technique...

all tri/sqr LFO`s according to the trigger/integrator principle (like here),
got sharp rising signals at the trigger output (U2A in this case),
that can be transmitted into the audiocircuit by "antennas" like any
length of wire coming from that point;
(in this case: the wiring towards and including R9, R11, R12,
the waveformswitch and the diode-side of the optocoupler - the latter only in square-mode...).
Rearranging the wiring can be a huge help!

[markphaser]

really

[KMS]

The current demand/supply?   Maybe this could be overcome with a nice oversize power supply (not a battery).  You could test this out by hooking two "fresh" batteries up in parallel located at each point on the circuit where one battery is shown. I'm guessing that would be 4 batteries total (have not seen the circuit but +/- power supplies usually work best with at least 2 batteries).

if that works then maybe build a big power supply unit.

[R.G.]

Dont' forget ground-induced noise. The sudden pulse as the opamp hiccups (good one Mark!) causes a sudden pulse of current in the ground line, and if that's used as a power line by the stereo jack trick, you get a pulse on the input of the effect.

Tick!

[CS Jones]

Not that all of this isn't good info but none of it worked for me on my CA tremolo.
I use a bi-polor PS fused at 250mA. I even gave the opamp it's own separate +9v supply.
Non-switch input jack.
Built the oscillator onto a completely different breadboard to separate it from the audio side.
Tweaked the wires in and out of the 4049 for hours to see if antenna effects were causing trouble.
Nothing worked.
The squared side worked well enough with just a minor bleedthrough. The triangle side was unusable because of the tick.
What I ended up doing was stuffing up the supply side at pin 1 of the 4049. 10k resistor in place of the 1k and then a series 100k trimmer wired as a variable resistor into pin 1. It changed the nature of the effect of course but with mutual tweaks on the 100k trimmer combined with dialing up the 50k speed pot differently I was able to at least control the triangle side ticking and still get some usuable sounds out of it.

I use a gained-up tuna melt now FWIW. The CA is long gone.

[gez]

One thing you can do with CMOS LFOs is follow the advice in the Art of Electronics and run the chips from a 3V supply.  This reduces current spikes to miniscule levels and leads to tick-free operation, though obviously you still need to pay attention to layout etc.

Amplitude can be recovered by using an op-amp to amplify the integrator's output.  Cumbersome, but very quiet.

The 3V can be done using a transistor follower (circuit connects to source/emitter, which acts as V+)

[The Tone God]

I made a "tickless" single supply CA trem using a few techniques.

1. I decoupled the 4049's power supply mentioned earlier.

2. I have analog (opamps) and digital (4049) grounds which are connected at one point only, the jack ground, which helps reduce ground spikes.

3. In the PCB layout I did I put the 4049 circuitry away from the rest of the circuit and put a ground plan around it. There are only two tracks that are travelling from this confined area. V+ and ground. Note no speed control.

4. A great deal of clock noise comes through the speed control cross talking with the analog usually through the speed and depth controls which most people put in the vicinity of the each other. I choose to keep the 4049 speed control in the confined area on the board. I used the LDR of an opto to control the 4049 speed and the speed knob controls the LED, as discussed in the Rock N' Control article, thus the clock is confined.

These above steps help remove all the clock noise I had.

Andrew

[CS Jones]

Thank you guys for taking the time to write up fix info.
I'll give it a shot again with this circuit asap using your suggestions.