EM3207 (v1.1) - MN3207 based EHX Electric Mistress (9V) clone

[DrAlx]

One other thing I should mention is that my stripboard layout uses the opamps on IC1 and IC3 in different order to what is shown in Tomas' schematic and PCB.
This was to make a more compact layout.  That means the pin-outs in my build are a permutation of those in the Tomas' schematic.
In Tomas' schematic, the IC3 opamps are used in this order (IC3A, IC3B, IC3C, IC3D) but my stripboard layout uses them in reverse order (IC3D, IC3C, IC3B, IC3A).
Similarly for IC1.  

Here is my EM3207 stripboard build in a 1590BB. The range and rate pots haven't arrived yet, so I am using 2 fixed resistors for the moment (far right of pic):

[DrAlx]

And the messy solder back ...

[DrAlx]

And here's what it sounds like with those 2 resistors replacing the rate and range pots (roughly 50% on each).
First minute is with the filter matrix on (nice bell sounds from the harmonics  
Second minute with the sweep on.
It's a crazy tuning on the guitar (bottom E string tuned down to B)

https://soundcloud.com/alex-lawrow/em3207_flanger_demo

A  BIG THANK YOU to Tomas for making this cool circuit available to everyone.

[Scruffie]

Great Vero layout! Don't use it anymore my self but sure a lot of guys are going to appreciate that.

Nice build too 

[DrAlx]

Apologies. There is a mistake in image (4) of my original post with the stripboard design.
That's the image of the strip-side of the board, showing where to put the breaks and solder blobs.
In the process of taking a screenshot, I must have clicked on the program window and accidentally removed a strip break.
That's what caused the colors to mismatch, not my coloring algorithm as I first thought.

I have corrected this now and put the whole set of diagrams in a single 1600x1400 image here...

http://imageshack.com/a/img855/4543/mw48.jpg

[DrAlx]

My pots finally arrived from Banzai so I finished my build and reset the trim-pots.
I previously set RT3 to zero to give quickest clock, and it was flanging but sounding quite phaser-like.
I have now set RT3 to give a maximum delay of 13 ms in filter matrix mode (which is larger than the
7.69 ms delay measured by Thomeeque on the vintage EM in filter-matrix mode) but I like the result.
I don't have an oscilloscope so I set RT3 as follows. I hope the method is useful to other people who want to set RT3 but don't have a scope.

1) Use filter-matrix mode with Range pot at maximum value (longest delay) and Color pot to maximum (maximum feedback).
2) Start with RT3 around 20% from zero.  Set RT1 so it is just short of causing self oscillation.
3) Make a short loud slap on the guitar neck.  The pedal should produce a sound that is similar to the sound
you get when you hold a ruler on the edge of a table and "twang it".  RT3 sets the pitch of the twang.   RT1 sets the decay of the twang.   It must not decay too quickly or you will have difficulty determining the pitch. The idea is to trim RT3 until the pitch of the twang matches a reference pitch.
4) I wanted a maximum flanger delay time of 13 ms, which corresponds to a twang with pitch of 1/0.013 = 77 Hz.
   You can use an online tone generator (http://onlinetonegenerator.com/) to give you a reference pitch.
   What I did instead was take a digital delay pedal with a delay of 13 ms and lots of repeats,
   and used that to produce a reference twang sound. I then toggled between the delay pedal and the EM3207, tweaking RT3 until the twang sounds had the same pitch.
5) Once RT3 is set, move RT1 away from the self-oscillation point until clock noise is acceptably low.  I like twangs to sound metallic but too metallic.

I recorded samples of the tweaked pedal here. 
This is the guitar into the EM3207 and then straight into a BR600 recorder with only a little reverb added and nothing else.

http://soundcloud.com/alex-lawrow/em3207-flanger-after-trimming

[DrAlx]

Thomeeque, I have a question for you about component values.

I know that the BBD output needs a LPF with a cutoff frequency that is less than 1/2 the clock frequency
in order to avoid aliasing effects.  I have read that in practice the LPF cutoff frequency should be
even lower (less than 1/3 of the clock frequency).  See link to book extract at very bottom of this message.
Presumably that is because real world filters don't have a rectangular cutoff.

For the EM3207 I see there is an LPF on the BBD output formed by the parallel combination of
of (R12,R13,R14) = 2.23k and C7 = 680pF.  
I calculate the 3dB point for that RC filter to be about 100 kHz
which is 1/2 the 3207's maximum supported clock frequency of 200 kHz.
A clock frequency of 200 kHz gives the minimum supported BBD delay of 2.56 ms.

My question is this:  Shouldn't the LPF filter cutoff be determined by the minimum clock frequency
that is encountered during a sweep rather than the maximum clock frequency?
For example, let's say that at the bottom of the flange sweep the BBD delay is 10.24ms.
The clock is only running at 50kHz at that point, not 200kHz, so for that part of the sweep the BBD
requires an LPF with a much lower cutoff of 25 kHz, not 100 kHz.  Am I missing something here, or is there
some other reason why that bit of the circuit is using a 100kHz cutoff?

I increased C7 to 2.7nF to lower the LPF cutoff, and it does noticeably reduce the clock noise at the bottom of the sweep
but it also makes the flange feedback a little less strong, so RT1 needed tweaking to increase feedback level.
The strange thing is that I prefer the default C7 value of 680 pF, because there is a bit more treble in the sound
despite the louder clock noise.  So maybe aliasing effects contribute to the sound in a good way?

Here's the link to the BBD book extract:

http://books.google.co.uk/books?id=9_tkus_y8BQC&pg=PA194&lpg=PA194&dq=MN3207+clock+cancel&source=bl&ots=QyiGHC0Z7W&sig=pxqOY3-vjj8qvlzdRfBw_GV4XuE&hl=en&sa=X&ei=3Xf3UY3uFeHD7AbLvoEw&redir_esc=y#v=onepage&q=MN3207%20clock%20cancel&f=false

[Hemmel]

http://imageshack.com/a/img855/4543/mw48.jpg

Hi DrAlx,

This looks very interesting to me cuz my brother is bugging me for an Electric Mistress. I was looking at your layouts and I'm wondering what are numbers 5 through 11 on the bottom-right layout ? You've identified 1-4 and 12,13 from bottom-left to bottom-right layouts, but not the other numbers....
Could you clarify please ?
Thanks !!!

[DrAlx]

Thomeeque kindly produced a pdf document with detailed build instructions on page 1 of the thread.
Here's the link to his pdf:

http://thmq.mysteria.cz/em3207/build/EM3207_v1.1_Build_Instructions.pdf

My stripboard layout has the exact same component numbering and pad numbering as that document.
The only component value that is not up-to-date in that PDF is  C17.
Thomeeque recently posted than he now recommends 47pF instead of 22pF, so that's what I used.

(The 5,6,7,8,9) are the PAD numbers in Thomeeque's schematic.
They are to a double-pole double-throw (DPDT) switch that lets you toggle between flanger mode and filter-matrix mode.
The DPDT switch is wired like this

 6--O  O--7
 8--O  O--9
 5--O  O--6

10 and 11 are not needed.  If you want to experiment with different C17 values you can use them to place an extra cap there.
I didn't bother using them in by build.  I only marked them in my layout for sake of completeness.

[Hemmel]

Oh yeah... that's me, whenever I see a thread that's 18 pages long I'm usually lazy and don't read it all before replying...  

Thanks for clearing that up.

Now I need to find that MN3207 chip (neither Small Bear nor Tayda seem to have it)

[Scruffie]

Oh yeah... that's me, whenever I see a thread that's 18 pages long I'm usually lazy and don't read it all before replying...  

Thanks for clearing that up.

Now I need to find that MN3207 chip (neither Small Bear nor Tayda seem to have it)

Use a V3207 or BL3207, Smallbear has both, just different branded modern reproductions.

[Hemmel]

Thanks Scruffie ! I tried googling for "MN3207 equivalent" but didn't find any  

Now if I put a 22pF on C17, then use pads 10 and 11 to connect a SPDT switch to add either a 22pF or a 47pF, wouldn't this provide a choice between 22, 44 and 69pF (more or less) ? I'm guessing I'd need an ON-OFF-ON SPDT ?

Here's what I mean :



Would this be right ?

[DrAlx]

Thanks Scruffie ! I tried googling for "MN3207 equivalent" but didn't find any  

Now if I put a 22pF on C17, then use pads 10 and 11 to connect a SPDT switch to add either a 22pF or a 47pF, wouldn't this provide a choice between 22, 44 and 69pF (more or less) ? I'm guessing I'd need an ON-OFF-ON SPDT ?

Here's what I mean :



Would this be right ?

Yes. That's correct.  If you don't mind ordering from overseas, Banzai Music in Germany have both MN3207 and BL3207.  That's where I ordered all the ICs,  but their delivery costs may be too high for you.  It wasn't too much of an overhead for me because I'm in the UK and I ordered other parts at the same time (other ICs, the enclosure, 9mm alpha pots, and most of the foil caps).

[DrAlx]

Hemmel, I have just uploaded version 2 of my stripboard layout.
It's pretty much the same layout but I moved some components to make things a little less cramped.
It will make for a slightly easier build.  The improved layout is here...

http://imageshack.com/a/img35/8502/xeaj.jpg

[Hemmel]

Thanks DrAlx !

BTW, I'm trying to figure how I could setup that switch on C17 so that once boxed, instead of the switch going 69pF - 22pF - 44pF from left to right, it could go 22 - 44 - 69 (or 69 - 44 - 22).
If this setup needs another kind of switch it's ok since I haven't bought any yet.

Any ideas ?

[DrAlx]

Thanks DrAlx !

BTW, I'm trying to figure how I could setup that switch on C17 so that once boxed, instead of the switch going 69pF - 22pF - 44pF from left to right, it could go 22 - 44 - 69 (or 69 - 44 - 22).
If this setup needs another kind of switch it's ok since I haven't bought any yet.

Any ideas ?

I thought I found a way to wire the switch so that one throw would add capacitance in parallel to increase overall capacitance and the other throw would add capacitance in series to lower overall capacitance but my calculations were wrong.

To be honest I am not sure of the benefit of having the cap switchable since
you may need to retrim one of the pots when you switched caps.
As Thomeeque says, 22 pF isn't recommended because the effect won't sweep properly.
The original idea was to use that low value to overclock the delay chip, but there is a fundamental limitation to the
speed that you can clock that chip. Overclocking will cause it not shift charge effectively between buckets.
As I understand it, that would cause the signal to get weaker as it goes through the delay line, and also smear itself out
to some extent because charge it not transferred fully between buckets..

[DrAlx]

I'm working on a vero layout to try and fit a TZF flanger into a 1590BB (using the EM3207 as a starting point).  While going over my vero layout for the EM3207, I realized I got the values for RT2 and RT3 the wrong way around in my layout and build.  They can still be trimmed OK of course.  In any case, i have put a corrected layout (with switch wiring and alignment procedure) in a single image here...

http://www.imageshack.com/i/5njet6j

BTW, has anyone tried Thomas's TZF add-on for the EM3207, and if so were there any heterodyning issues ?

[Hemmel]

I get "lost image" from that link ...

[Scruffie]

I'm working on a vero layout to try and fit a TZF flanger into a 1590BB (using the EM3207 as a starting point).  While going over my vero layout for the EM3207, I realized I got the values for RT2 and RT3 the wrong way around in my layout and build.  They can still be trimmed OK of course.  In any case, i have put a corrected layout (with switch wiring and alignment procedure) in a single image here...

http://www.imageshack.com/i/5njet6j

BTW, has anyone tried Thomas's TZF add-on for the EM3207, and if so were there any heterodyning issues ?

Yes, No (although that is not necessarily gunna be the same for each build) and it didn't work that well.

I think the EM is too 'light' on the flange for it, i'd go with an MXR-117 to try it.

[DrAlx]

I get "lost image" from that link ...

Strange. Try this instead...

http://imageshack.com/a/img203/7361/jet6.jpg