Just for scientific purposes

[Nasse]

I found this circuit in an old magzine, and thought to post it for some theoretical discussion. It may be illegal or someones property but anyhow. Maybe i must remove it soon.

Just an example of diode controlled phase sifter. It obviously has drawbacks but maybe benefits too. Common and cheap parts, easy matching.

[gez]

Hmmm, something strange going on here!  When I click on the link you gave Nasse, it opens up another window containing this page!

Strange...very strange... :?

[Nasse]

Yep, same problem, maybe I did something wrong when posting. But if you copy the link the pic is there...

Maybe I should study how to use this forum... :oops:  :x

[Tim Escobedo]

I've used diodes as voltage controlled resistors. They generally work OK, can introduce unwanted distorion. I'd thought about using them in a tunable state variable filter (VCF) but never in a phase shifter. Interesting idea.

BTW, I had to view the source of this page to find the url to the graphic. Doesn't load for me.  :?:

[Brett Clark]

This circuit should work pretty well, except that there will be some control voltage "feedthrough" at higher LFO frequencies. You could make the diode pairs "balanced" by applying another, out-of-phase LFO voltage through another resistor to the bottom diode (instead of just grounding it). This would fix the feedthrough problem. Look at "Mystery Circuit #2" on Geofex for an example of a balanced diode pair.

The original ring modulators used a double-balanced diode circuit. And somebody (either ARP or Korg, don't remember which) had a balanced-diode-based filter in their analog synth. So, this kind of stuff goes back a while.

How old was the magazine article that this came from?

[BillyJ]

Ummm..is the link gone or am I just needing help?

[gez]

Ummm..is the link gone or am I just needing help?

The link's gone for me too (still haven't bagged it!).

[amz-fx]

For those with problems, I'm seeing it at:

http://www.freewebs.com/nassetronics/dphase.jpg

Looks interesting.

regards, Jack

[mattv]

copy the link, but remove the "http://"

[drew]

It looks interesting and pretty simple.... anyone built something like this before?

I've seen diode-ladder VCFs before, but never a phaser (then again, I have enough phasers to last me a lifetime, so I usually don't look...)

[puretube]

some synth-DIY pages have examples (j.haible - google!)
ARP string-phaser had 14 stages with diode-array ICs.....

[R.G.]

Diodes have a variable resistance region in their characteristics. It's the part of the diode's conduction between when it's not yet forward biased enough to conduct significantly, and when it's conducting hard - that is, it's the tiny region between diode-off and diode-on.

Since we use diodes a lot for generating distortion, you might guess that you have to be very careful with using diodes as variable resistance devices to avoid distortions.

The generally-accepted figure in the engineering community is that diodes can be used with little (less than 1% or so) distortion only if the signal across them is less than 25mv. Since a single coil guitar produces a signal of around 100mv and humbuckers produce up to a volt, you have to attentuate the signal down by something like 4 to 40 to keep distortion under control with diode-based variable resistors. You then have to amplify it back up. The concept is a lot like the "Adjusticator" project at GEO.

On top of that, there is the control voltage feedthrough mentioned above. You're trying to keep signals under 25mv, but you need to change the overall voltage across the diode by about a tenth of a volt to do this. The signal gets mixed with a control voltage four times its size. You have to somehow reject the control voltage while keeping the signal.

Most good circuits for this do it with a balancing scheme where there are an even number of diodes in a string and the control voltage is pulled high on one end of the string, low on the other, and the signal is injected into the middle of the string. If this is done well, the control signal cancells out in the middle of the string.

Using multiple diodes in series is another common trick. You can put 25mv of signal across any one diode, so with two you can use 50mv, four gives 100mv, etc. It helps.

Diode modulators are fun to play with, but they very quickly consume more parts that an equivalent design with a transistor modulator like the MC1496 or an OTA. An OTA, looked at in one way, is effectively a diode modulator using the base-emitter junctions of the input transistors and the transistor action of the input devices to amplify the signals up.

[Nasse]

Thank you all for informative comments, hope I did not remove the circuit too soon. I did it because it was someones original from some old magazine and those copyright and so on things. If some one wants to see the pic I can put it available for some time, well it is so old anyway and maybe little commercial value, or perhaps better e-mail it.

Just wanted to share. I have not tested the diode phaser application but did few tests with basic attenuator for some voltage controlled audio level control and tested it as tremolo. The author claimed that with say four diodes in series and signal in the middle (and with that cap from top of diode system to ground that does good job at distortion cancelling, I used 470 nF) you can increase signal level somewhat, I think 250 mV and more was mentioned in original article with some tricks. I tried with singlecoil strat level and from some "preamplifier" like my Tube Sound Fuzz and Alesis Microverb (not at full level) and the distortion was surprisingly low with four diodes. But low signal level and noise may be what you get in phaser. But you can add more diodes to increase signal level, so there may be some use of this by some tweaking...

Maybe I do some more testing, thanks anyway.

[Paul Perry (Frostwave)]

And not forgetting the "diode switch".
http://www.tpub.com/content/neets/14179/css/14179_39.htm
that link might make it clearer (or not!)
might be an idea for people who want a 'chopper' trem.