Build Your Own Clone Message Board

I wanted to enhance the difference between the normal and differential mode because to my ears it was not very perceptible. Also, I did not like the ever present level of "feedback" when using the special BBD with double the delay in position B.

The “differential” mode mixing point of the two delay IC's in the flanger circuit uses a 22k for BBDb vs. a 10k (for the single BBDa) apparently to lower the signal level of the 2nd BBD at the circuit mix point and make the transition between normal and differential less apparent. However, this would appear to lessen the impact of the differential mode altogether. It is worth noting that other circuits (Electric Mistress and ADA) do not create a level difference between the two BBD outputs.

When I changed R17 to 10k then the differential mode became noticeably louder than the normal mode because apparently now there is two times the delayed signal being summed together. This actually made it difficult to do an A/B comparison of the two modes and was pretty annoying when using the differential mode especially with distortion.

So a change was needed to make the mixing point equal for both signal halves as well as make this combined “differential” output the same level as the single BBD normal output. A change to the use of the existing SPDT switch was schematically drawn out and a PCB mapping was determined to allow the use of a new 10k resistor to lower the combined signal level when in differential mode and to bypass the same resistor in normal mode.

I'll attempt to describe what was done here. I cut through the two traces of the PCB layout for the SPDT switch so that I could then rewire it directly at the mixing point (the intersection of R17, R18 and R19). Then I cut the trace connecting R17 and R18 on the non-component side of the PCB. And finally made a 4th cut on the non-component side between R17 and C13 (which is still connected to R19). This left the mixing point junction free for wires to connect these as needed.

Next I added wires to reconnect the circuits. First, I added a wire on the non-component side from BBDb pin 8 to R13 (this is where the switch used to connect). Then I added 3 wires to the component side of the PCB as follows:
1 - connect the previously unused contact from the switch to R19.
2 - connect the common of the switch to R18.
3 - connect the 2nd contact of switch to R17.
And then finally add a 10k resistor between the common and the previously unused contact of the switch.

This was implemented and tested and the results were surprisingly good between the levels of normal and differential and resulted in almost no perceptible difference in volume between the two. Once significant bonus with the change is that there is no longer a loud pop when switching between the two modes.

Now I could better A/B the two modes and determined that a change in the value of C10 from 0.033uF to 0.1uF made the differential mode warmer for both flanging and chorus and kept the normal mode basically stock (which I always liked). I had tried making C10 smaller but this made the differential mode noticeably thinner which I didn't like.

Interesting. Could you post a clip with non- and differential?

Awesome!

_________________
==||==
Function f(x)
TightCamera - Gear Reviews

Here's a schematic of the switch in the modified circuit (it was re-purposed from the original). I'll add that this change enabled me to better tell the difference between the normal and differential modes. However, the difference was very subtle and I changed caps to make it more so. If I'm not mistaken there should be some phase cancellation in differential mode but it is again very subtle. I will say that to get a good EVH tone you need to be in normal mode now and the differential mode is great for that Electric Mistress tone (Gilmour, Police, Lifeson) and chorus.

Perhaps this should be a new thread, but because I mentioned at the start of this item that "I did not like the ever present level of "feedback" when using the special BBD with double the delay in position B" and the results of correcting this have now affected the new resistor value added in the revised switch position - it seems appropriate here.

To start with, I was still having problems with the feedback circuit (it seemed like it was always on when I had distortion before the pedal even with the Regen pot at fully CCW) and was reviewing the sticky post on the original schematic and noticed that there were some "corrections" from the original schematic noted and a revised schematic was included. In the correction it was mentioned that the feedback resistor, R23, was actually connected to pin 2 of IC1. I assumed my recently purchased pedal was already connected that way but when I checked the circuit it was not. This led me to wonder what was going on with the mix point of the feedback. I looked at a couple of Flanger schematics and found some (e.g, Boss Flanger) where the feedback was connected directly to the op amp pin. I found the Pearl Flanger schematic and noticed that Dry/Wet mix was wired directly to the Op Amp (like a typical mixer which prevents signal bleed through). I checked and rechecked the feedback path for some shorts or coupling but found none and even disconnect R23 and C17 and still had feedback - so I came to the realization that the original delayed signal must be feeding back to the IC1a input via the "Dry" signal path. I tried shorting R27 but this did not help. Then I took a stab at connecting R23 directly to pin 2 of IC1 (changing the resistor to 47k as in the orignal) and no longer had the constant feedback with the Regen pot at fully CCW. The feedback was much more pronounced and controllable. However, oscillation started at around 2:00 on the Regen knob. So I increased the value of R23 from 47k to 100k and simply wired it (untrimmed) directly from the R23/R24 hole all the way to the IC1 pin 2 connection at R8. Now I have zero oscillation and a very well controlled level of feedback or even no feedback with the Regen control at fully CCW.

What surprised me next was that my prior circuit to create a transparent signal level between Normal and Differential now had a much higher level in the normal position and was attenuated in the differential position. I can only guess that the additional feedback that was always in the background was just enough to create this difference and now with it gone was noticeably lower when both BBD's were used. So I tested some values for the new differential resistor and ended up with the 10k being lowered to 5.2k. Even using the 3208 BBD (with double delay) in BBDb position no longer has an ever present flange tone. However, I still prefer to use two 3207's to get the Electric Mistress tone I'm after in the differential mode.

At this point I'm very happy with the pedal. There is still no pop when using the mode switch and the balance between the two modes is perfect. Alas, the difference between the two modes is still very subtle but you can actually tell that the differential mode is fuller (not louder).

Bob

BYOC Flanger - with modified mix circuits (3 total):
1. Moved R23 flanger feedback to input of IC1a pin 2
2. Changed BBDa/b mix circuit for equal balance and volume (as shown above)
3. Moved wet/dry mix to input of IC5b pin 2 (no separate R27)

First, a reference article for the last post explaining the difference between the
passive and active mixing circuits and why I changed the mixing point to be directly
at the op amp IC5b input pin 3 instead of first passing through R27.

http://sound.westhost.com/articles/audio-mixing.htm

After multiple tests and configurations I could not hear any difference between
the BYOC Normal/Differential modes. With the change to the BBDb cap to 0.2uF the
tone in the Differential mode is slightly warmer but not really significant. TODO:

There were some suggestions from Mark Hammer that a useful improvement for increased
feedback would be from changing the cap after the feedback pot from 0.1uF (as on
BYOC) to something smaller. Both the MXR (15nF) and EM (4.7nF) use smaller caps in
this position so this should be changed. This may improve the deep flanging
typically used with the MXR. NOTE: The MXR uses the neg input (mix point) of the op
amp where the clean and feedback sum. The EM uses the pos input of the op amp where
the signal then feeds/passes the dry signal to be mixed later. The BYOC is really
like neither but could be made to be more like the MXR (as was done previously).

For me, the original goal of building the BYOC Flanger was to get a combination of
both the MXR Flanger and the EH Electric Mistress.
The MXR Flanger was always obtainable but was vastly improved by two things:
1. The movement of the feedback to the input of the Opamp at IC1a pin 2 (I used the 10k resistor R8 as my connection point).
2. The movement of the wet/dry mix point to a “proper” active mix circuit at the
input of the Opamp at IC5b pin 2.

The Electric Mistress was more elusive. While it was certainly close in the original
configuration. The differential mode really did nothing to enhance or replicate the
EM tone. After testing it appears that the most significant difference created by
the EM circuit is that the feedback path enters the circuit BEFORE the dry signal is
separated. This means that when feedback is used even a little there is some delayed
signal injected into the “dry” signal and then mixed with the delayed (and feedback
enhanced delayed) signal with likely more than just a normal dry and delayed signal.
When the BYOC had the differential mode using two different delays (1024 and 2048)
the tone was very much like the EM in clean settings with much feedback. However,
with distortion it is less controllable.

I was not successful in moving the feedback signal before the dry signal (using
either the + or - input of IC1b did not appear to have any effect or properly mix the
feedback). When I tried to move the dry signal to the output of IC1a it was clear
that this “worked” but actually created “pos” feedback (nasal tone) which is not
desired because the feedback needs to be at the “-” input of IC1a.

However, using a properly mixed BBDa and BBDb with two different delays and an
improved wet/dry mix circuit I was able to achieve a very good EM-like flanger while
maintaining a very good MXR style standard flanger.

I posted a YouTube video at the following so you could hear the improvement in the
Differential mode replicating the EM.

http://www.youtube.com/watch?v=MV-7wzD3XOg

[some thread necromancy here...]
Bob, I like the way this mod sounds. Thanks for the demo clip! I'd been considering buying an old Electric Mistress, but I think this mod sounds just about as good (to my ears anyway) and has more flexibility.

Unfortunately, I'm an electronics n00b (haven't studied it except in college physics) and don't feel confident about interpreting your posts to do the mod. Would you (or anyone else who understands this) be willing to write up some step-by-step instructions? Or failing that, to critique some instructions I try to write, until I get them right?

Thanks!

Bob thank you so much for posting this, is there any way you could provide some pics of all the mods, I'm on the limits of my knowledge here and would love to make the changes you suggest. Could you summarise your 'final' build/mods (as I got a bit lost above as to what should/should not be changed!) to get the sounds you achieved in your youtube video please? Did you stay with the 3208 or revert to the 3207? Thanks!

Sorry to resurrect this thread, but I was curious if anyone has actually done these mods on their pedal? I'm about to start work on it (finally after buying the kit years ago!)

If you've done the mod, could you post what you did and pics if you have them?

_________________
Take a break from Building Pedals and
Read our webcomic ~ nemu*nemu at:
http://nemu-nemu.com /