ELectro Harmonics Talking Pedal schematic

[John Lyons]

Anyone have a link or place to find the EH talking Pedal schematic?
Thanks in advance.

John

[Marcos - Munky]

I have the schematic, don't know where I got. It uses a strange dual pot. One half is 11K and the other is 7K.

[joegagan]

my guess is that it is a 10k/10k that has drifted or was made poorly

[Rodgre]

If I recall from other discussions, it is a specific pot that E-H used, and that's why it's nearly impossible to DIY. If I understand how those types of formant filters work, that dual-ganged frequency pot is controlling two filters, tuned differently, and that's why the different values are necessary.

Roger

[Marcos - Munky]

Yeah, it's a kind of custom pot that EH used. But I don't think it will be too hard to DIY. Maybe we can mod a dual pot to have 7K on one half and 11K on the other using resistors. Or we can use LDRs and this.

[R.G.]

I have it somewhere.  It uses a special made pot.

The pedal is two bandpass filters, one which goes monotically up/down with pedal motion. The other is high, goes to low, then back up.

The idea is that the filters hit vowel formant regions - see http://geofex.com/Article_Folders/sing-wah/sing-wah.htm for info on the vowel formants, and "The Technology of Wah Pedals" at GEO for the multiple feedback filter.

It is probably possible to modify one wafer of a dual pot  by painting conductive paint on the wafer and bringing that out to a fourth contact to make a tapped pot. See "The Secret Life of Pots" for ideas about how to do that.

When I did the Sing-wah, I gave up on redoing the dual pot.

It is possible to do the same thing with LDRs and opamp drive-waveform shapers.

[StephenGiles]

Apparently, a job lot of these dual pots were bought way back when and MM commanded his troops to design something to use them - that's what I was told at EH UK.

[doug deeper]

ive been wondering if you took the circuit from the phozer (the newer version) and ditched the lfo and hooked up the to filter to a duel ganged
pot in opposite directions would it talk?
i think it may!

[Rodgre]

ive been wondering if you took the circuit from the phozer (the newer version) and ditched the lfo and hooked up the to filter to a duel ganged
pot in opposite directions would it talk?
i think it may!

I don't think it would necessarily talk. Read R.G.'s article on formant filters. It's not exactly as easy as sweeping filters. Formant frequencies (that simulate "A E I O U" (and not very often, Y ) are pretty specific. You have to think of it in terms of the two filters have to be at two different specific frequencies with a specific bandwidth in order to give you what you're looking for.

In 2007, I would think that this is the sort of thing that would be perfectly suited for a digital circuit, but would still be pretty prohibitive to do analog.

Roger

[R.G.]

ive been wondering if you took the circuit from the phozer (the newer version) and ditched the lfo and hooked up the to filter to a duel ganged
pot in opposite directions would it talk?
i think it may!

Back in the 1890's there was a lot of hot discovery on voice imitation. There was a mechanical speaking machine at the Chicago World's Fair, as I remember, that a young lady manipulated to change resonances as air blew through it. She could produce passable speech with it.

There was a discovery kit from Bell Labs (back when there was a Bell Telephone) that made a simple voice imitation kit.

Finally, there is a widget called the "Chatterbox" which is an all electronic voder. You press/turn controls to move formants and introduce siblants. Supposedly , it's relatively easy to learn to "talk" with it.

The basics of vowel production are two resonant filters of modest Q, that you can move around in the range 150 - 3kHz. Given sufficent coordination, you can make it talk.

The EH talking pedal subs in that tapped pot for coordination.

[birt]

google "Voder"

[R.G.]

If you want to play with real voice synths, check out:

http://yusynth.net/archives/WirelessWorld/Chatterbox-1976.pdf

[John Lyons]

Ok, got the schematic sent to me.
So it's a mechanical issue. The 11K side of the dual pot is tapped but I wonder what the tapped resistance would be?
Setting this up as an LDR/LED controlled device would be a nice way to go it seems (good idea marcos!). But I wonder how the tapped resistance fits into that setup? Ideas?

Thanks

John

[Marcos - Munky]

Maybe using a trimpot to set the LDR "maximum resistance value" and tune it by ear...

[John Lyons]

Marcos
Here is the pot section of the schematic.
The 11K section of the pot is tapped (4 connections).
I'm not clear on which section is swept or are they both connected to the wiper and swept?



John

[R.G.]

I actually have a "talking pedal" on the workbench, although it's not exactly the EH one.

I use a PIC controller to read a pot and provide a control voltage. The voltage is read by the A-D, and the software takes certain regions of the voltage as instructions to do a certain vowel sound. The actual filtering is done using some dual digital pots from Microchips technology. The pots are set to certain values to make the filters go to the formant frequencies for the chosen vowel.

I do a little computation to pick the number of steps between present vowel and next vowel and step the pots to there over a transition time. That helps with the sudden transition between frequencies and sounds better than a sudden step, although not perfect.

"Perfect" might be had by using an LED/Dual LDR like the VT5C3/2 and servoing one LDR to a certain resistance value, leaving the LDR's slow response to move the filter frequency slowly between steps. More natural than sudden stepping.

The other things I have not yet tried is using a switched resistor or switched capacitor to modulate the filter frequency by PWM'ing the R/C and using a PWM'd voltage output with an analog slew rate limiter to slow the slew between one setting and another. This needs a voltage controlled filter, but that's not all that hard.

The nice thing about the controller approach is that you can pick and choose what transitions to make, and you don't have to settle for the natural progression of vowels as you sweep a filter with a pot. It can all be table driven.

[John Lyons]

It looks like the 11k section of the pot sweeps between something like 8K and 0.
Then once it gets past the 0 ohm tap to ground sweeps between 0 and 3K.
I know these aren't the actual resistances but do I have the theory of the tapped pot correct?

John

[Gus]

Make me think of volume pots with taps for the loudness function on home stereos.

[puretube]

Y`re makin`it all too complicated... 

[Gus]

The dashed line and CW marks look like it means both pots mechanical connected together and the left goes to little less than 11K full CW(tapped part to end) and the right pot goes to closer to  0 ohms.  Full CCW is different. The left is at the tapped to left end of the pot resistance and the right closer to 7K

Now I don't know the rotation range the pots are swept or the starting points.

I would guess the triangle at the pots are going to a buffered 1/2 power supply voltage( because of the use of the 1K resistor as the left pot wiper passes the tap causing loading)

The left pot goes down to 0 ohms and back to two different higher values as it is swept .  The right changes value lin or log what ever type it is

Things needed
  resistance of the pots at both ends of the PETAL sweep
measure from the triangle to the wiper and maybe a few points in between 1/2 1/3 1/4 travel etc.(might need to dissconnect one end of the 1K to get a true resistance reading).  The resistance readings should give a good idea of the type pots
Take that value and add it to the 1K and calculate the filter response