Fairfield Circuitry - Randy's Revenge

soulsonic · Post by **soulsonic** » 21 Jul 2009, 05:25

If I were building a Ring Mod with a 633 (which I currently am, actually), I would power it with an AC wall wart to get a proper +/-15v supply. When I first breadboarded mine, I tried powering it with a 555 voltage multiplier, and the high current draw of the circuit was totally killing it... granted, this circuit had the ICL8038 in it too, so there is really alot of current being drawn, but it definitely showed me that it wasn't worth messing with faking the voltage.
I guess if the rest of the circuit has very low current demands, an ICL7660 might do okay, but I'm much more satisfied with the results of using an AC wall wart that can give plenty of juice.

Whoismarykelly · Post by **Whoismarykelly** » 21 Jul 2009, 21:44

soulsonic wrote:If I were building a Ring Mod with a 633 (which I currently am, actually), I would power it with an AC wall wart to get a proper +/-15v supply. When I first breadboarded mine, I tried powering it with a 555 voltage multiplier, and the high current draw of the circuit was totally killing it... granted, this circuit had the ICL8038 in it too, so there is really alot of current being drawn, but it definitely showed me that it wasn't worth messing with faking the voltage.
I guess if the rest of the circuit has very low current demands, an ICL7660 might do okay, but I'm much more satisfied with the results of using an AC wall wart that can give plenty of juice.

Thats one of the problems with a clean ring mod design. Invariably they take a lot of power which makes them a hassle on pedalboards with usual brick supplies.

fairfield · Post by **fairfield** » 22 Jul 2009, 00:07

I agree that having +-15V is ideal for any circuit using opamps and other such devices. That said, the AD633 operates just fine with +- 9V. Apart from having less room to play with, everything else sounds exactly the same. Even if you lose 4V at each rail, there is still room for a 10Vpp signal.

The ICL7660S has the same boost function as the MAX1044, the switching frequency goes up to 30kHz or something like that. Apart from that it also accepts up to 12V instead of 10. Not sure what the other differences are besides the price. When testing this setup at first, I was able to pull up to 30mA on the negative rail without any significant loss in voltage or squeal. I was quite happy that I would have a pedal that would be operating from a 9V center negative instead of some incompatible 15VAC adapter, and that without any artifacts.

The key to smooth ring mod is indeed in the carrier soggybag. The purest your sine wave, the purest the bell tones. I used a phase-shift oscillator, built with 2 LM13700s (it's in the datasheet). Very cool circuit, pretty easy to add voltage control and the output is clean as a whistle.

Anyone ever tried the Gilbert cell mixer? I was thinking about ordering some matched pairs for this. I bet it would sound real nice when properly tuned.

soggybag · Post by **soggybag** » 22 Jul 2009, 01:22

fairfield wrote:I agree that having +-15V is ideal for any circuit using opamps and other such devices. That said, the AD633 operates just fine with +- 9V. Apart from having less room to play with, everything else sounds exactly the same. Even if you lose 4V at each rail, there is still room for a 10Vpp signal.

The ICL7660S has the same boost function as the MAX1044, the switching frequency goes up to 30kHz or something like that. Apart from that it also accepts up to 12V instead of 10. Not sure what the other differences are besides the price. When testing this setup at first, I was able to pull up to 30mA on the negative rail without any significant loss in voltage or squeal. I was quite happy that I would have a pedal that would be operating from a 9V center negative instead of some incompatible 15VAC adapter, and that without any artifacts.

The key to smooth ring mod is indeed in the carrier soggybag. The purest your sine wave, the purest the bell tones. I used a phase-shift oscillator, built with 2 LM13700s (it's in the datasheet). Very cool circuit, pretty easy to add voltage control and the output is clean as a whistle.

Anyone ever tried the Gilbert cell mixer? I was thinking about ordering some matched pairs for this. I bet it would sound real nice when properly tuned.

I think it's a smart idea to stick with atypical 9V adapter compatible power supply. No one wants to have a different brick for every pedal.

I agree about the sine wave. As an alternative, I was getting pretty clear ring mod tone by using a low pass filter on a triangle wave. This was an idea from the Stompboxology Ring Modu-maic. This sounded much more musical than the triangle. Though the triangle or square can sound good to if you're looking for a harsher tone. I figured the low pass rounds off the triangle to an approximated sine. I like this as it gave the option to adjust the waveform.

soggybag · Post by **soggybag** » 22 Jul 2009, 02:16

Thanks for the info Fairfield. If you don't mind me asking, is the sine wave oscillator from the app note the "FIGURE 17. Sinusoidal VCO"? (The one I'm describing uses 4 OTAs).

fairfield · Post by **fairfield** » 22 Jul 2009, 12:41

Yea, that's the one. The LM13700 contains 2 OTAs, so indeed the whole circuit uses 4 OTAs.

earthtonesaudio · Post by **earthtonesaudio** » 22 Jul 2009, 20:08

fairfield wrote: Anyone ever tried the Gilbert cell mixer? I was thinking about ordering some matched pairs for this. I bet it would sound real nice when properly tuned.

I just salvaged a bunch of CA3026 ICs, which look like good candidates for experimenting with Gilbert cells.

DWBH · Post by **DWBH** » 24 Jul 2009, 23:20

For anyone's interested, the oscillator that soggybag and fairfield were talking about is this one here:

What's Vo and Vc?
Voltage out (?) and Voltage control?

I looked up the AD633 datasheet, but I was like: waa?

soulsonic · Post by **soulsonic** » 25 Jul 2009, 06:12

yeah, VO is the output signal, and VC is the control for the frequency of the oscillator. You would connect a potentiometer in series between the VC connection and the positive voltage supply.

DWBH · Post by **DWBH** » 25 Jul 2009, 11:57

Ooh, ok. Thanks.

Is the square wave also obtained from this oscillator, or is it a different oscillator? It would seem reasonable to have both waves from the same oscillator, because different oscillators with common freq pot would mean more trouble and cost, I think.

DWBH · Post by **DWBH** » 25 Jul 2009, 19:15

Will this oscillator work as well, or will it not yield good results?
http://www.national.com/ms/LB/LB-16.pdf (first schem on that pdf)

Is the oscillator hooked up to the AD663 4th pin, like in this schem below?

DWBH · Post by **DWBH** » 07 Aug 2009, 21:34

bump

anyone seen a gutshot?

noelgrassy · Post by **noelgrassy** » 08 Aug 2009, 14:15

Anyone ever tried the Gilbert cell mixer?
I found a few things on it;

The. Gilbert. Multiplier Cell includes a source-coupled transistor pair, also referred to as the lower transistors or driver transistors, which is coupled to a pair of cross-coupled source-coupled transistor pairs, also referred to as the uppertransistors, switch transistors or active mixer transistors. For example, RF inputs may be coupled to the source-coupled transistor pair and a local oscillator is coupled to the pair of cross-coupled source-coupled transistor pairs to produce ademodulated output. The Gilbert Multiplier Cell is extensively described and analyzed in Section 10.3 of the textbook "Analysis and Design of Analog Integrated Circuits" by Paul Gray and Robert Meyer, John Wiley and Sons, N.Y., 1993, pp. 670 675.

fairfield · Post by **fairfield** » 08 Aug 2009, 15:39

HFA3101

This chip would take care of the dirty work. Too bad it's not available in DIP.

muki · Post by **muki** » 10 Aug 2009, 03:06

I'm getting +/- 16v out of a 9v using this circuit

https://i652.photobucket.com/albums/uu2 ... lier-1.png

works pretty well! the IC is the max1044, diodes are 4001 and caps are all 10uf. just make sure their voltage is high enough.

DWBH · Post by **DWBH** » 20 Jan 2010, 15:33

mictester · Post by **mictester** » 20 Jan 2010, 18:19

earthtonesaudio wrote:I want somebody to try this:
http://www.audioscientific.com/28_Simpl ... lters_.jpg

I used exactly that approach to generate a very pure 19 kHz pilot tone for a broadcast stereo coder, though I used eight resistors. The second harmonic was -65 dB below the carrier.

mictester · Post by **mictester** » 20 Jan 2010, 18:24

noelgrassy wrote:Anyone ever tried the Gilbert cell mixer?

.

I've used Gilbert Cell mixers in RF work. There is a small range of ICs based on Gilbert Cells made by Philips (or their successors) starting with the NE602. They include an oscillator which feeds one port of the mixer. I've used them a lot in RF work, but don't think they'd work at audio frequencies.

The aged TDA120 can be used at audio frequencies. It's designed as a TV IF IC but works well down to audio frequencies.

~arph · Post by **~arph** » 21 Jan 2010, 08:57

I've been using this as an approximate sine LFO for the past year

http://www.friedair.com/images/sine-lfo.png

You can easily speed it up by lowering the three .22uF caps to something lower in value (eg. 22nF)
Also you have two taps. The output of IC1a and IC1b are opposite in phase and one is slighty lower in amplitude, but both sine like.

Ignore the LED's at the output those are to drive LDR's

One thing to note is that for VR I use a simple voltage divider, but no electrolyte to filter the Vr.

seele · Post by **seele** » 28 Sep 2010, 14:22

*Bump*
Any news?