Simplified Ringmod with LFO

[Hambo]

I seem to remember a guy posting after a simple ring mod. Well, I've been wanting something like that myself, so I've been messing about on my fancy schmancy new breadboard using a bunch of ad633 chips I got free from Analog devices . So I managed to get a fairly smooth sounding ring modulation effect down to 2 IC's. The final opamp stage is ripped directly from Roman Sowa's ring mod with LFO, simply because they seem to be the right values to produce the correct output level... I actually used a 33k resistor going into the opamp, still good.

Its very clean sounding, with some VERY minor carrier signal slipping through.. The carrier itself I took from an article describing how to get a sine wave from a single opamp, and I found that using a 250K pot (not 470k as in the schem) gives a pretty good range of freq without loss of level.

So it would be 3 chips in total using a charge pump to run from 9v, and maybe 4 if using another opamp do a wet/dry mix function?

So the questions:
1) When I first rigged it up, the carrier signal was coming through quite pronounced, but I played a few notes and it just dissolved into almost nothing, and didnt come back! can this be explained, could the chip somehow balance itself given that I have no offset/null thingy apparatus in place? Should I just be thankful and forget it? I think so ha! Half the hassle with the Sowa design was having to *&%^($&^@$( about with those sodding trimpots all the time.

2)What needs to be added to this? I didnt use any caps on the input/output for simplicities sake, but I found it sounds pretty sweet anyway, so what exactly do those caps do?

3)I did not use a buffer stage (voltage follower opamp thingy) as in Roman Sowas ring mod, I did to start with, but found no difference when I ran the signal straight in so I decided to save half an opamp.. what difference will this possibly make?
Will I need a buffer opamp stage to do a blend function, or could I just split off from the input signal and run that to a pot with the output coming into the other side? That would be magical.

4)Finally, I put the input signal to both modulation and carrier sources, and got a relatively clean octave up sound (with the usual octave rigmarole.. bridge pickup, low tone, high neck notes etc) but the level is much much lower in this setup, Anyone have any insight as to how to tidy that up? It would be cool to get it on a switch.. I would mess about with it some more tomorrow, but I'm buggered with it for today .

Thanks anyhoo, heres the schematic anyway, sorry about the crapness, I know, I suck.

[Paul Perry (Frostwave)]

Why the bleedthrough varied over time, is probably due to the DC varying at the audio input to the AD633. When there is some DC, you are multiplying (DC + input signal) by the carrier, and not suprisingly, that gives you carreir in the output.

If you get serious about bleedthrough, you will be putting theose trimpots of Roman's back in, believe me

[Hambo]

Well, you may be right, but I doubt it. No disrespect, I understand you and pretty much everyone here are streets ahead of me in terms of knowledge, but well you just have to hear it. The carrier signal is so tiny, even with the volume cranked you can barely pick it up when your not playing as an extremely faint wooooo in the background, could this be due to the nature of using a sine wave rather than square or triangle (urgh)!.. sounds a ton better than when those trimpots needed readjusting again, and you dont have to worry about it.
The real problem is, I want a ring mod that is compact, and will run on my pedal board supply and not be a pain in the arse. with the Sowa ringmod, everytime the power supply changed a bit (Ie you un/plug another pedal into the supply somewhere else, sometimes I swear it would just go off from being moved around a bit!) it would sound like absolute death until you yank it off the board, unscrew the back, adjust 2 trimpots by ear until theres no noise, put it all back.. not exactly jam friendly.

So is that what the large caps are for on the input/output, I read something about caps in the signal path converting it from DC to AC or some such? I'm going to put them in and see if I can get any difference.. except now I need more batteries coz I left them in there all night like the twadge I am.

OK anyway, so no-one is much interested in ring mods eh? ha, well, could someone please explain to me what the buffer does? I mean, I can read, I knowwwwww input vs output impedance, doesnt load down the signal something loading errors or thingy. I still don't really understand. What difference does it make? Say I dont have one where there should be one, and the signal gets "loaded down" or whatever it is, what does that actually SOUND like? what am I trying to avoid?

[slacker]

Looks cool, nice and simple 
In simple terms "loading down" means that the impedance of the input is too low compared with the output impedance of the thing driving it. This normally results in a loss of treble and volume from the signal which you don't normally want. According to the AD633 datasheet it has high impedance inputs, it doesn't say how high, so you probably don't need a buffer. You might need to put a cap on the input though to block any DC that might be there, although if you're using a bipolar supply there's possibly 0 volts DC on the input so it would be fine without a cap.
Do you have a link for that sine wave article? It sounds interesting.

[Hambo]

http://sound.westhost.com/dwopa.htm#phaseshiftosc

Took some finding in the first case, I was trying to use a 555 timer I have lying around with really bad results till I found that.
Thanks for the explanation, so your saying that the opamp takes a weak signal, and makes it a strong signal basically? does it not do the same thing when you use it in the normal manner- like on the end of that circuit above - to amplify a signal then? like as a byproduct? I need to read a bit about impedance I think, to me the term is still just something to get right when plugging in speakers.

[Hambo]

Further update for anyone interested. Poly caps seem to block signal altogether going into ad633. Large and small electros seem to make no disernable difference (at least not to my cloth ears). Blend pot works bloomin perfectly, 100k pot, left lug split from in signal, other end from output opamp, middle to volume pot. For some reason the carrier signal dropped a little bit more when attached! I'm beginning to think boxed up with some liberal amounts of shielded cable and this wont be half bad!

I couldnt get a very good level on modulating the guitar with other stuff (a bass, a small FM radio) but it did surprisingly work quite well with just a cable plugged in as the second input and someone touching the end of it.. conjured up all kinds of bizarre images of weird performance art type stuff with a guitar and a jack plugged up the you know where!

It'll be a p*&% easy circuit, the only downside being if you cant get free samples from Analog devices, the AD633 is like a fiver a piece!

[Paul Perry (Frostwave)]

Well, congratulattions on making a ring modulator that you are happy with! I always say, if you want to learn electronics, build stuff, as you are doing.

And I am sure that you will learn even more, as you try to make your ring modulator even MORE perfect.

You are also correct about AD633 not being cheap. Even when you buy them by the 100s.

[Hambo]

Hey thanks man. You know I just wish it was something cool that a lot of people would be interested in, instead of just me and my ring mod obsession . I suppose someone would already have done it though if that was the case. You are right about the trimpots anyway, they can still be added back in and keep the chip count down to 2, so still keeping it a simple build, maybe I'll strip 2 up both ways and compare, maybe in this layout they would not need so much readjustment... just hoping.

Anyway.. very small (30-47pf) cap in parallel with 1m resistor in output opamp cuts down hiss whilst keeping nice bell like tones on mod.

[Hambo]

Back again.

Yeah I tried adding the trimpots and adjusting them just seems to cut down to the same signal bleedthrough you get when you just ground the pins, after all that signal is one of two integral parts of the modulation function, Its never going to go completely away, it didnt in the full Sowa design.... It seems the trimpits really just add more hassle than they solve, unless they are there to provide a way to balance uneven biploar supplies along the lines of what slacker said? So the question must be asked, how "even" will a bipolar supply from a MAX1044 be?

This is running at present on the +-12v supply I have on my board for bipolar boxes, I found that when I unplugged the supply from the other box (FSH1) the whine came back and was very pronounced... I was thinking it was because there was suddenly more power to the oscillator, it took a while to realise that it was more to do with power supply filtering in that pedal suddely being disconnected. Couple of caps later and it was tamed down again. a 20k in series with freq pot keeps the osc signal from attracting dogs from all around the neighbourhood

Still when on this circuit is still quite hissy, I suppose from the AD633 chip then being boosted by the tl072. Whats the best way to fight hiss in this kind of situation without affecting the signal too much, anyone?

Hehe I've been playing with Express too...
What about those 2 caps in parallel on the -9v output of the max1044, Ive added 2 for filtering but does that mean the one that needs to go with the max can be removed now? or should I use a bigger cap on that rail? confusing. I havent got any max1044s to test with.

[Cliff Schecht]

Hambo, maybe you are referring to the ring modulator I did a good while back? It all fits easily into a Hammond box and allows for an external carrier via the extra jack. I instead just run my Whammy's unaffected signal into the carrier so that I can mix it against itself and get some seriously crazy effects. It makes the perfect octave up and octave down fuzz as well as a bunch of other great little tricks I've figured out. I play around with my looper a lot, here's something I recorded using only ring modulation and a few other cool little tricks .

Listen to the third one (Ringmod Loop):
www.purevolume.com/cliffsjams

[Hambo]

Hi Cliff.

That looks like a classic germanium diode ring ringmod. I Did build one of those before with a single transistor boost on the output, as the passive type are very quiet, Im not sure If I screwed it up, but the output was unbalanced between the input sources.. I went onto get some of these ad633 chips to mess about with then, these circuits tend to be a little bit more stable and cleaner sounding I think. Is the opamp in your circuit a booster for the output, or an lfo generator (or both)?

Some cool sounds you got there man, That first one sound like an escapee with a kazoo hahaha! That Ringmod Loop is kind of hard to pick out but it reminded me instantly of the start of a Meshuggah song - Future Breed Machine. Cool to get that kind of sound with a guitar fella!

[Processaurus]

Hambo, have you tried the tl072 gain stage before the multiplier chip?  That would probably lessen the noise considerably.

Cliff, i like your band

[Paul Perry (Frostwave)]

Yeah, +1 on gain for the signal before the AD633, for best signal to noise, you want both the audio & the carrier to be as high as possible without overloading.

[Hambo]

Yeah, I understand that, but theres two signals going into the ad633 and only one comes out, so it would need another half a tl072 to boost both signals.. and I am in chip/space miser mode. I did notice however that the max1044 has an Oscillator? with at least 2 frequencies... hmmm wonder if that could be used somehow. As usual I dont know enough about these things and I dont have any to play around with.. and I cant easily get any to play around with because the only place I have seen a MAX1044 is that mail order place in Germany!! CRIPES!

Anyway I see theres 2 frequencies on the max by connecting two pins, wonder if theres a way to vary the frequency.. A pot between these pins, or would it screw with what the chip is there for? Anyway If there was a way to use that as the carrier... probably not

Hey I really appreciate all the good advice guys! I'll probably shift stuff around a bit accordingly later on

[Cliff Schecht]

Hi Cliff.

That looks like a classic germanium diode ring ringmod. I Did build one of those before with a single transistor boost on the output, as the passive type are very quiet, Im not sure If I screwed it up, but the output was unbalanced between the input sources.. I went onto get some of these ad633 chips to mess about with then, these circuits tend to be a little bit more stable and cleaner sounding I think. Is the opamp in your circuit a booster for the output, or an lfo generator (or both)?

Some cool sounds you got there man, That first one sound like an escapee with a kazoo hahaha! That Ringmod Loop is kind of hard to pick out but it reminded me instantly of the start of a Meshuggah song - Future Breed Machine. Cool to get that kind of sound with a guitar fella!

It's a buffer for both the input and the carrier. My ring modulator has no carrier, I use an external carrier in the form of the Digitech Whammy's wet and dry outputs. The Whammy makes it sound perfect, there is never bleed through and notes always come through clearly. I was also doing a cool thing the other day by plugging in two guitars (input and carrier) and playing harmonics over the fretted notes a friend was playing. When you hit certain intervals, there is no bleedthrough and the sounds coming out are so freaking cool, a very intense harmonics on steroids type sound. I don't really care for the LFO >> Carrier setup, I made a little external box to do this and didn't care for the sound. The Whammy deal also allows for perfect octave up or octave down fuzz magic.

Hambo, have you tried the tl072 gain stage before the multiplier chip?  That would probably lessen the noise considerably.

Cliff, i like your band

There was no band there, just a few buds I jam with occasionally. That's a short clip off of about 2 hours worth of jamming when I had whipped out my guitarimin deal I made. It was a synthstick I built into a guitar with a broken headstock, I made the "pot" by double sided sticky taping the VHS tape to a fretless fretboard and mounting two strings over the tape, isolated by the nut and bridge. I actually had it tuned like how a guitar was, the length was almost perfectly 1 Meg. There was also a kill switch and some other fun stuff. Everything else on that Purevolume is me using my looper .

[Hambo]

Well I tried with the buffers and caps from roman sowas design, and for some reason it still hisses, Could be just because its on a breadboard?

found that adding the cap  and 1meg to ground at the start causes a lot of bleedthrough and neccesitates the trimpots in Sowa design... weird..  theres no other difference!

I prodded around the oscillator, read up a bit on phase shift oscillators and came up with a slightly better sounding, louder output, I put a switched carrier jack and it can be used as a regular ringmod. I'm giving up and putting it in a box soon, I cant spend any more time pushing it around

[markusw]

could this be due to the nature of using a sine wave rather than square or triangle (urgh)!..

Looking at your schem it seems you're actually using the square wave as a carrier.... but maybe I miss something.

Thanks a lot for sharing your experience btw!!! 

Markus