ring modulator issue

Thomas P. · December 30, 2003, 12:31:14 PM

I found this really interesting stuff:

This is a ringmodulator based on a AD633 chip:
(watch out for other nice stuff on this site)
http://www.sowa.synth.net/m_pics/m_rm.gif

And here is a nice application with a LM13700:
http://www.ori.org/~aaronc/synth/RingMod_V1.pdf

If someone has interesting material about this issue, it would be nice to post it here and maybe we can cook a good and cheap ringmodulator.

Paul Perry (Frostwave) · December 31, 2003, 01:20:15 AM

Roman Sowa (responsible for the AD633 ckt above) is a pretty smart cookie, if he says it works, it's good. He's on the synth-diy list.

Thomas P. · January 01, 2004, 08:43:57 AM

After reading and searching about this issue I took the AD633-based schematic and modded it to my need. I added a signal generator in place of the carrier input. Therefor I took Mark Hammers advice (thanks :wink: ) and simply used the LFO from the ZombieChorus. I changed the 0.01uF cap to 0.001uF which gives a frequenzy range from I think ~50Hz to ~110Hz (my software says that). I also found out that takeing the signal from lug 2 of the speed-pot will give a squarewave, so I took this as a feature to switch between triangle- and squarewave.
I didn't use the AC/DC coupling switch since I don't think this is usefull for guitar use.

The simulation looks good so far but I can't get rid of the carrier signal while the input is quiet. Another proble is the adjustment of the signal- and carriernull since I don't really know how to tune them.

(I've drawn a schematic so far but I can't post it, sorry)

ExpAnonColin · January 01, 2004, 01:40:31 PM

I was going to say-you're going to need a function generator for the carrier input. Nice idea for useing the one from the zombie chorus, but aren't those frequencies a bit low for it? Normally, ring mods have a larger frequency range than that.

For the audible carrier, chances are you're giving the circuit a little bit more voltage that it can handle, just a tiny bit, but more probably, the frequency is jumping it's way over to your output wire because there's very little shielding (you're breadboarding this, right?)

-Colin

Thomas P. · January 01, 2004, 02:40:29 PM

Quote from: anonymousexperimentalistI was going to say-you're going to need a function generator for the carrier input. Nice idea for useing the one from the zombie chorus, but aren't those frequencies a bit low for it? Normally, ring mods have a larger frequency range than that.

For the audible carrier, chances are you're giving the circuit a little bit more voltage that it can handle, just a tiny bit, but more probably, the frequency is jumping it's way over to your output wire because there's very little shielding (you're breadboarding this, right?)

-Colin

Well, I actually simulating it on the computer.
I'm useing a 1kHz sinewave for the instrument input and the Oscillator for the carrier input. The results look good, but at this time I can't really tell if they sound good :wink: Therefor I can't say if those frequenzies are too low - later on (as I'm takeing this to the real world) I can change the range with swapping to a lower cap.
Now you might know that I'm not hearing the carrier, I'm actually seeing it when I'm turn the signal input off.
I thought the output level should be controlled by on of the input signals so if this particular signal is at zero the output becomes zero!? Maybe I have the inputs reversed.

Maybe I'll send you the schematic so you can take a look on it.

Paul Perry (Frostwave) · January 01, 2004, 07:01:26 PM

Are you using a bipolar supply in the simulation? Because, if the signal has a voltage offset, then the carrier can never go to zero.

Thomas P. · January 02, 2004, 05:18:09 AM

Quote from: Paul Perry (Frostwave)Are you using a bipolar supply in the simulation? Because, if the signal has a voltage offset, then the carrier can never go to zero.

Yes, I'm useing a +/-15V supply. Is it possible that the voltage offset comes from the IC input buffer and may it be eleminated if I decouple the AD633's input with a cap?

Paul Perry (Frostwave) · January 02, 2004, 06:24:14 AM

I've never used a simulator, but I guess you can check the voltage inputs at the AD633 pins. Note in Roman's ckt, the offset provided by P1 and P2 pots.. the signal inputs to the AD633 are differential with respect to these nulling pots.

Thomas P. · January 02, 2004, 11:00:57 AM

I checked the voltage at the signal input and had 0.9V there. Then I tuned the trimmer but I couldn't get closer to zero than 23mV. During this I had an eye for the carrier signal at the output which didn't went lower than 700mV.

Paul Perry (Frostwave) · January 03, 2004, 12:15:32 AM

Quote from: tomboyI checked the voltage at the signal input and had 0.9V there. Then I tuned the trimmer but I couldn't get closer to zero than 23mV. During this I had an eye for the carrier signal at the output which didn't went lower than 700mV.

Do you mean you were trying to set the trimmer to zero? Because, you would expect it to want ot be not far from 0.9v, because you are trying to overcome the effect of the input offset.
Also, in a practical situation, it might help to have a 100k resistor to ground, from the AD633 input pin (in case the input cap is charging up or somethig).
And be sure the trimer is from + to -, not from + to ground (yeah, anothe dumb thing I fell for once).

Thomas P. · January 03, 2004, 10:26:15 AM

Thanks Paul for constantly replying my stupid questions!!!

What I did was I turned the signal off so that only the carrier appeared in the circuit. Then I measured the DC-voltage at the AD633's signal input. On the scope I watched the curcuit output which showed the carrier at 900mV.
At this point I adjusted the carrier-null trimmer to get the DC-voltage at the AD633's signal input to become zero (with no signal appearing) because I thought then the output will become zero.
From what I've read I thought that the output amplitude follows the input signal amplitude so when I eleminate any voltage at the AD633's input I will have no carrier audible at the output.

Maybe I have a totally wrong idea of what's happening?!

I hope this is understandable to you (I know I have some strange expressions) - I'm no native speaker so you might forgive me :wink:

Paul Perry (Frostwave) · January 03, 2004, 08:16:54 PM

Quote from: tomboyWhat I did was I turned the signal off so that only the carrier appeared in the circuit. Then I measured the DC-voltage at the AD633's signal input. On the scope I watched the curcuit output which showed the carrier at 900mV.
At this point I adjusted the carrier-null trimmer to get the DC-voltage at the AD633's signal input to become zero (with no signal appearing) because I thought then the output will become zero. :

You might need to adjust the other null pot as well.
Because, even though there is no 'signal' going into the AD633, there might be a constant DC offset.
It is not very simple.. even though the signal input to the AD633 is zero DC, and there is no signal, carrier can still get through if the relavant trimmer is not set right.

Thomas P. · January 04, 2004, 06:44:06 AM

The problem is that the other trimmer doesn't seem to effect the carrier.
But I think I have to physicaly build it since I don't trust my simulation any more :?

News:

ring modulator issue