Modulator schematic

[brett]

I mentioned this in another thread.  Thought some people might be interested.  This has been done from memory, so there might be a few dodgy bits.  Basically, the circuit boosts the signal, then inverts it, then a bunch of switches (controlled by a 555) select the signal or the inverted signal in rapid succession.  Finally, there's a mixer and some filtering before the volume control.  

*Haven't built it myself*, it's based on my memory of the circuit in a kid's toy I have at home.  Looks like it will work, though.

***This schematic now deleted - see below***

[Peter Snowberg]

Very cool! Thanks for posting it!

Take care,
-Peter

[Marcos - Munky]

Thanks for the schematic. I like to know if this worked with guitar when you finish it.

[brett]

I've deleted the schematic from my site because there's a couple of problems with it.  But I've built a draft circuit and it works, but lets thru lots of modulation noise.  So I'm going to try and improve that today.

I'll post a new schematic on Thursday or Friday.

[brett]

This is the modified toy circuit:


It works well, except for the amount of background modulation noise (I suspect it's only -20 or -30dB with respect to the signal - definately too noisey for most situations).  Maybe someone will find a way to quieten this circuit?

[Paul Perry (Frostwave)]

Is the background modulation noise there when there is no input signal?
If the input signal is a sine wave, does the noise change radically in pitch as the frequency of the sine changes?

[Marcos - Munky]

Thanks again for the schematic. Could you record some soudclips?

[Peter Snowberg]

One thing you might want to try would be to add a 4013 wired as a divide by 2 to the 555 output. That way you get a 50% duty cycle and probably a less complex harmonic content to any bleed through.

Take care,
-Peter

[brett]

The waveform of the noise is kinda like a camel's hump with a long tail.  It's not high frequency hash, but a distinct hum at the fequency that the TLC555 is running.  On RG's advice, I switched from a 555 to the cmos version.  Perhaps an op-amp-based square wave might do the trick.

Unfortunately, I don't have a setup for posting soundwaves, but it's very similar to other good modulators.  Because I've been running the 555 at lowish frequencies, the sound ranges from a very fast tremolo to some nice, clear bell-like tones.

Next step is to replace the 555.  Any suggestions?

[Peter Snowberg]

Next step is to replace the 555.  Any suggestions?

How 'bout the CD4046?

-Peter

[Paul Perry (Frostwave)]

If changing to a CMOS 555 didn't help the noise, and the noise is at the 555 frequency, then it must be something integral to the circuit.
If it does it in the absence of input signal, then I suspect that the bias level of the inputs to the switches is unequal, so you are effectively switching a DC offset on and off.

[brett]

Thanks everyone for the suggestions.  Paul P - yeah I've wondered whether it's a DC difference between the signal and inverted signal.  I'll re-check the DC on them.  Funny thing is that I get something close to a sinewave out (but that might be the final low-pass filter smoothing things out, too - I'll check it pre-filter and see if the noise is square).  I also might use a simple CD40106 flip-flop for the square wave - I've used it before on something else and think I have the parts at home available for this project (how cheap is that - scavenging a $1 part!).  Anyway, I'll let everyone know if I get the clock noise out of this beast.