Which ring modulator to build

[Cliff Schecht]

I'm gonna try 2 more places, but I've already ordered those Mouser parts. Bantam Electronics and/or Howard Electronics in Austin might have what I'm looking for.

[markusw]

racedriver...

for a String Ringer? PITA, isn't it?

I had to buy from a UK source. $20 for a pair of lousy trannies (and yes, they are lousy), all because of shipping... 

I'm pretty sure the 42TM006 will be fine for the String Ringer. Also the 42TM019 should work IMHO.

Would be interesting to see how a 1:1 works in contrast to a 20:1.

Didn't ckeck it's dimensions though.

Regards,

Markus

[Cliff Schecht]

Alright, I got the transformers in. I breadboarded up a quicky modulator without matching the diodes or doing any other prepwork. Well, I must say I was quite surprised with a pretty nasty ring modulation sound, but it basically worked.

I've got a few questions I think some of the guys around here could help answer and when I'm done, I plan on writing something of a manual on how to build one of these things.

1) How much output do I need from my guitar and oscillator? Does 1.25-1.75V sound like enough?
2) If the 74HCT14 is running off of 3V, what would the actual output be under load? I'm not gonna have my scope running until my dad finds the da,m probes, so I can't really do a lot of testing until I get that running.
3) I want to keep this thing passive, so the 600ohm:600ohm will give me a louder signal than the 10k:10k transformer, right?
4) Does anybody else have experience building one of these things? Any advice or warnings would be greatly appreciated

[Cliff Schecht]

Alright, aftersome perfboarding a version with the 600:600 xfrmr, I got nothing. I tinkered with it for a good while but I couldn't get a damn sound out of it. Scrap pile.

I then rebreadboarded the ring mod with the 10k:10k xfrmr and fairly closely matched 1n34a's, I got some interesting results. When I plug a guitar into just the input or carrier jack, I get a decent sounding effect, but quite weak. Obviously this isn't going to be passive for long. The problem is when I plug a signal into both the carrier and the input I get no effect whatsoever, it's just both of the signals completely uneffected. I'm not sure what the problem is or why it won't mix the signals properly... Anybody know what's wrong?

[markusw]

Alright, aftersome perfboarding a version with the 600:600 xfrmr, I got nothing. I tinkered with it for a good while but I couldn't get a damn sound out of it. Scrap pile.

I then rebreadboarded the ring mod with the 10k:10k xfrmr and fairly closely matched 1n34a's, I got some interesting results. When I plug a guitar into just the input or carrier jack, I get a decent sounding effect, but quite weak. Obviously this isn't going to be passive for long. The problem is when I plug a signal into both the carrier and the input I get no effect whatsoever, it's just both of the signals completely uneffected. I'm not sure what the problem is or why it won't mix the signals properly... Anybody know what's wrong?

With a signal plugged into just one input you just should hear some bleed-through. With a sine signal run into both unputs you should get some octavish sound.
However, just guitar will be too weak. You'll need 2-4 V p/p so you might stick a booster/TS/whatever in front of the input. Also the carrier signal should be around 2-4V which the 74HCT14 should deliver (maybe run it at slightly higher voltage).

Sorry, don't really know why it does not work. Also don't know whether the 20:1 in the String Ringer has any advantage over the 1:1 xfmr you tried. Maybe someone else could give us an explanation ?

[Paul Perry (Frostwave)]

1) How much output do I need from my guitar and oscillator? Does 1.25-1.75V sound like enough?
....I wouldn't think so. That's barely enough to turn on the diode!!
I'd be up for at least 5 volts, more if possible.

3) I want to keep this thing passive, so the 600ohm:600ohm will give me a louder signal than the 10k:10k transformer, right?
....I doubt it. Depending on the impedances of the signal sources. I suspect it would be much the same.
4) Does anybody else have experience building one of these things? Any advice or warnings would be greatly appreciated
....if you are ever seriously trying to match diodes, do it in a still atmosphere, and do not touch the diodes with your fingers, because that will give a variable amount of heating and throw the reading right off.

[Cliff Schecht]

Thanks for the replies!

I also didn't realize that the diodes were so touchy, but I've got a meter that can measure the voltage drop and probes that can grab the diode and hold on to them without having to ever touch them. Most of my 1n34's measure around .35V, but I found a good few as low at .3V. I'm socketing these so I don't burn them out with an iron.

Now to find a suitable chip to boost those voltages...

[Cliff Schecht]

Update: I'm very excited now... I used a MC4558 with a gain of 5 at first and was running straight guitar (EMG's@~1.5V) and I got some great sounds. I then switched out a resistor for a gain of 50(Saturation!!!) and again was very pleasantly surprised. The best was yet to come. At first, I was using my Guitarimin as a tone generator, but I noticed it bled through a lot (I probably messed it up by throwing it on concrete a lot ), so I decided to try something different. I just bought a Digitech Whammy and the thing has a wet and dry output, so I figured I should make use of both of them. I set this thing up so that I had the wet output of the Whammy going to the ring modulators new and improved buffered input section and ran the dry output of the Whammy through my pedalboard and into the carrier. Well, in short, I've gotten pretty close to what I want out of this thing...weird noises. On the pedalboard, I just clicked on my modded Metalzone and presto! The ring mod was subtle when the Whammy was in its resting position and as I rocked the pedal forward, I was rewarded with a smoothly blended ring modulation sound that gave some very cool sounding overtones as well as some very spacey sounds, but at the same time, one could still distinguish the notes being played. All in all I'm really happy I took the time to make this thing right and If anyone is looking for a cool circuit to tinker with, I would recommend this in a second. This has been the first time I really got to design anything with opamps which really is an added bonus to the entire thing, especially considering my design worked the first time I tried it. I'm getting ideas faster than I can even think to write them down... I LOVE IT!

[markusw]

Cool you got it working! Congrats!

What should sound cool too is running the Whammy in 2 octaves up mode then through a comparator and then using this signal as a carrier.

[Paul Perry (Frostwave)]

Thanks for the report, Cliff! I'm sure it will inspire!

[Mark Hammer]

A couple of summers back, JC Maillet paid me a visit and we spent the afternoon trying out pedals.  One of the things we ended up talking about, which he hadn't really paid too much attention to before, was the way in which octave-up pedals can be coaxed into producing sideband products when strings are bent.  That is one of the reasons why people can so easily confuse octave-up fuzzes and ring modulators: the fuzzes can produce sums and differences by accident, and the ring modulators always produce sums and diferences by design.  In the hands of a guy like Jeff Beck, octave boxes can be used to mimic ring modulators.

JC worked on the math aspect for a bit and he may well have it posted on his viva analog site.  That lad does love his math!

Anything that is modulated by an LFO can be coaxed into producing ring-modulator like sounds if the modulation rate can go fast enough.  My old blue rackmount MXR Digital Delay will do a credible ring modulation effect because it lets you modulate the delay time by as much as 1000hz.  The John Hollis Frobnicator applies this principle to a tremolo by extending the LFO rate into the audio range.  While you are waiting for your parts to arrive, you can easily mimic a RM with any chorus, phaser, or tremolo you happen to have hanging around by simply reducing the value of the range-setting cap in the LFO by a factor of around 4 or 5.  So if the cap value .1uf, dropping it down to .022uf would nail you some RM tones.

[object88]

I set this thing up so that I had the wet output of the Whammy going to the ring modulators new and improved buffered input section and ran the dry output of the Whammy through my pedalboard and into the carrier.

Wow, that's a really great idea!  If you get a chance to record that, I'd love to hear some audio samples...

[Cliff Schecht]

A couple of summers back, JC Maillet paid me a visit and we spent the afternoon trying out pedals.  One of the things we ended up talking about, which he hadn't really paid too much attention to before, was the way in which octave-up pedals can be coaxed into producing sideband products when strings are bent.  That is one of the reasons why people can so easily confuse octave-up fuzzes and ring modulators: the fuzzes can produce sums and differences by accident, and the ring modulators always produce sums and diferences by design.  In the hands of a guy like Jeff Beck, octave boxes can be used to mimic ring modulators.

JC worked on the math aspect for a bit and he may well have it posted on his viva analog site.  That lad does love his math!

Anything that is modulated by an LFO can be coaxed into producing ring-modulator like sounds if the modulation rate can go fast enough.  My old blue rackmount MXR Digital Delay will do a credible ring modulation effect because it lets you modulate the delay time by as much as 1000hz.  The John Hollis Frobnicator applies this principle to a tremolo by extending the LFO rate into the audio range.  While you are waiting for your parts to arrive, you can easily mimic a RM with any chorus, phaser, or tremolo you happen to have hanging around by simply reducing the value of the range-setting cap in the LFO by a factor of around 4 or 5.  So if the cap value .1uf, dropping it down to .022uf would nail you some RM tones.

That's funny you say that about the MXR digital delay... I found one at Goodwill!  I know exactly what you are talking about too. Using just straight feedback, it's really easy to coax out a noise that sounds exactly like a police siren

Also, I plan on jamming within the next few days after I box up this ring modulator and my Mr. EQ. I'm probably going to add a buffer to the carrier also and just have pots going from signal to ground for both the input and the carrier so that I can use higher signals and not oversaturate the 4558.

And I will be posting soundclips when I finish.

[Cliff Schecht]

Damnit, this thing keeps getting more and more complex... I'm thinking about adding a charge pump to the PSU so that I can get 18v and run the output of the 4558's at the voltage required to get the germaniums to turn on and do their magic. How do I find out exactly how much voltage is required to turn on the 1n34a's?

[RaceDriver205]

Turn-on voltage for 1n34s should be about 0.3V
EDIT: Or less

[Cliff Schecht]

Yeah forget all of the 18v charge pump bs, upon further research I find I should be fine with just a 4558 to add a slight gain increase to the carrier and input. I even had it clipping a little bit and noticed no adverse side-effects. I finally got a scope up and running after over 2 years of building... go figure.

[Cliff Schecht]

Alright, it's perfboarded up on a little Radioshack square and should easily fit a 1590B. It even worked on the firstt try. Would anybody be interested in the schematic and a perfboard layout?

[Threefish]

I sure would. Been watching your postings with interest.

[John Lyons]

Yes, please! Thanks for your work.

John

[Cliff Schecht]

Alright, here's a perfboard layout done on Radiosmack Part Number 276-148 (I think 278 has one less row). This is the first perfboard layout that I've done on the computer so please excuse any slopiness. As far as I know, everything is correct, but I did just do this off of the top of my head so I also appologize for any errors there.



Notes:
-I'm not sure if this thing needed pulldown resistors on the input, carrier or both and I personally omitted them, but there is room.
-The long thin cap's are all 10uf tantalum and the smaller circles are all 1uF electrolytic.
-All diodes are 1n34a and for size reasons, are all mounted vertically. Be careful with orientation and socket these, as well as the opamp
-Transformers can be any 10k:10k (maybe others?) but need center taps
-The center tap on the input and output side of the transformers are connected to nothing.
-The gain on the opamps is set to about 4.7 (47k/10k), but I left these 2 resistors (The one going from pin 1 to 2 and from 6 to 7) socketed so that I can go back and change these at a later date if needed.

I'll get the schematic up in a little bit.