Mark Hammer's BF-2 mod report

[PenPen]

I know this info is all available if you search and peice stuff together, but I thought I'd put it all in one post for new members to find.
Quoting Mark Hammer:
There are many easy mods you can do to change the sound of the BF-2. Your ears will tell you if these are "improvements" or not.  For example:

a) C27 and C28 make up a non-polarized 16.5uf capacitor which sets the speed range.  If the capacitance value is reduced, the speed range goes higher.  Faster speed can produce an effect that sounds very much like a ring modulator.  I would suggest a pair of 4.7uf caps, back to back, as a suitable replacement.

b) Maximum flanging effect needs a 50/50 balance between delay and clean signal, but sometimes you want just a taste of flanging.  If the value of R29 is increased, the balance is shifted in favour of clean.  If you replaced R29 with a 22k fixed resistor, and a 100k pot, wired as a variable resistor (2 lugs, not all 3), you would be able to get everything from a *perfect* 50/50 balance to a much softer flanging effect, even with very high regen/res.

c) R11, D3, and D4 form a soft limiter, to keep high regeneration/resonance signals from overloading the delay chip.  If you shunt/short R11, the clipping of D3/D4 becomes "harder" and introduces more harmonic content.  If you want to experiment with more interesting high regeneration signals, solder in a wire bridge across R11 on the copper side of the PCB, turn up VR5 a bit higher (VR5 is normally set so that turning up the RES control all the way puts you just below the point of oscillation).

d) C7 sets the lower bandwidth point (bass rolloff) of the regeneration signal.  If you make it smaller in value, this will make high regen/res settings sound less metallic.  A value of .01uf might be a good place to start.

e)  The depth contol (VR2) adjusts the balance of voltages coming from the LFO and the Manual control to drive the clock   (IC4).  VR7 is used to adjust or fine tune the delay time by adding (or maybe subtracting) a small DC voltage on top of whatever comes through VR2/R48.  C30 sets the delay range of the clock circuit.  Values larger than 47pf will set the minimum delay time much longer.  Smaller values will reduce the minimum delay time.  You can change the sound of the pedal by changing the range of delays it can produce.  Try a cap of 39pf and see if you like it.  If there are problems with the sweep, you may be able to fix them with VR7.  Alternatively, if you are able to find an MN3209 chip somewhere, you can take the MN3207 out and replace it with the MN3209 for shorter (X/4) delay times and more of a "jet plane" sound.

f) Vibrato is created when you use ONLY the delay signal, and leave out the clean signal.  If you lift one end of R28, that will prevent the clean signal from being mixed in with the delay signal, and give vibrato.  Unfortunately, because of how bypass is done in this pedal, this means that when you try to bypass the pedal, there will be NO sound coming out.  There is a solution to this problem shown here: http://diy.erikhansen.net/ce2.htm

Of these, I have done D this morning, replacing C7. This cap is stock valued 0.047uF (47nF). I first tried 10n per Mark's suggestion. This made the flange way more natural sounding at the highest RES setting. I liked it, but 'rich, lush flange' aren't terms I'd describe it as. Dry, natural are terms I'd use. I then tried 4.7n, half of the previous. At this point, the flange gets more treblely, higher toned and punchy. I didn't like it. It wasn't metalic, at least, but it was peircing and high. Perhaps this would be good for a mod for solo's or something.

I really wanted to try 33n and 22n, but unfortunately I didn't have those values. I suspect there is a point that the metal can sound disappears but the thickness stays, and I suspect it is at one of those two values. I need to get some more caps. However, the 10n cap was way nicer than the stock, I just wanted a bit more thickness to the sound. As my control group, I replaced every cap I played a bit with it, and went back to stock at the end, then asked my wife to tell me which one she liked most, without me telling her which one was in there. She liked the 10n sound the best, and even though she isn't a musician, she does have good ears and is a good unbiased source of opinion I usually ask (or don't ask )

[StephenGiles]

Very interesting - wives always seem to be "busy" when such favours are asked!!
Stephen

[vanhansen]

PenPen, the link in f) has moved.  The new location is http://diy.erikhansen.net/ce2.htm

[Jaicen_solo]

What i'd like is a mod that will get rid of that jet plane swoosh that seems to overlay the sounds I play, and give a more natural comb filter sound. I find that the standard sweeping flanger just sounds fake, like it's overlaying it's own sound. I guess that's why I prefer phasers.

[PenPen]

PenPen, the link in f) has moved.  The new location is http://diy.erikhansen.net/ce2.htm

Modified the posting to reflect this. Thanks for letting us know.

What i'd like is a mod that will get rid of that jet plane swoosh that seems to overlay the sounds I play, and give a more natural comb filter sound. I find that the standard sweeping flanger just sounds fake, like it's overlaying it's own sound. I guess that's why I prefer phasers.

You mean like a fixed notch effect instead of a sweeping notch? That would be nice, though I can get something close by setting the RATE all the way down and properly setting the MANUAL and DEPTH controls. There is a sweet spot where that weird filtered sound comes through w/o the sweep.

[spudulike]

Is there a diy version of the BF-2 (ie without the crap switching) and a pcb layout ? Im going to start one myself after I finish the Blues Driver (BD2) clone but an existing layout would save time ...

[Mark Hammer]

What i'd like is a mod that will get rid of that jet plane swoosh that seems to overlay the sounds I play, and give a more natural comb filter sound. I find that the standard sweeping flanger just sounds fake, like it's overlaying it's own sound. I guess that's why I prefer phasers.

It's not immediately clear what it is you are objecting too, unless it is the "metallic" sound that cmes with too much resonance and too much low end in the regeneration signal.  The Mod discussed in this thread cuts out much of the low end in the regen signal that tends to produce the annoying metallic quality.  The alternative, I suppose is to completely kill the regen path for the most "natural comb filter tone".

Just out of curiosity, where do you set the regen control on your phaser, or do you prefer to play without any?  The answer to that may yield some clues as to what you are seeking.

Addendum for Spud:  Just took a peek at the CE-2 project on Tonepad, and the BF-2 schem. Unfortunately, there are enough differences between them that it would not be a simple matter of building the CE-2 and changing a few component values. Perhaps even more unfortunately, there are enough similarities to make you say "Aw, crap!  Soooooo, close!"

[Jaicen_solo]

I know i'm not too clear, I can't really get my head round what i'm trying to describe myself.
What I object to is the swoosh, like it's sweeping white noise. You know when the flange starts up in 'Alright Now' ??
Sometimes you can hear it when you're not even playing. If I could host a sample it'd be so easy to illustrate.
I want to hear detailed high end, but not the hissing swoosh!

I like this: http://www.ehx.com/ehx2/Catalog/14_Modulation/19_Deluxe_Electric_Mistress/01_Elliott_Randall/04_E_Randall_Smooth_Filter.mp3

And this is not bad:http://www.ehx.com/ehx2/Catalog/14_Modulation/19_Deluxe_Electric_Mistress/01_Elliott_Randall/10_E_Randall_Mistress_Q-Tron+.mp3

I don't like this, exactly what I'm talking about:
I want highly resonant flange, without this sort of audible artefact. Am I demanding???
http://www.ehx.com/ehx2/Catalog/14_Modulation/19_Deluxe_Electric_Mistress/01_Elliott_Randall/09_E_Randall_Sweep_Chug.mp3

[Mark Hammer]

Okay, now we're talking.  The 2nd example is what you get when you crank up the resonance; especially if that regen signal has much bottom to it.  What it seems like you want is an effect which about the ENTIRE input signal.  The heavy regen sample has a clear emphasis on one part of the spectrum, and the listener is forced to follow it up and down.

Note that when you add a regen signal to the input signal, you end up with a combined/mixed signal that is higher in amplitude than the original input signal alone.  The possibilities for clipping and 'misrepresentation" of the signal in the BBD can be substantial.  The ideal is to use companding to keep a lid on signal level going into the BBD, no matter how hot the input is, so that the BBD is allowed to show its good side.  What the BF-2 and many other flangers (e.g., the A/DA) do is to use a crude limiting/soft-clipping circuit instead.  If you peek at the BF-2 schem, you will see a pair of clipping diodes in the feedback path of the last op-amp before the BBD.  They have a resistor in series to soften the clipping enough that it resists being a fuzzbox, but it is still a diode clipper at the end of the day, and unlikely to be as clean-sounding as a compander.

When the regen signal is kept to a minimum in such a pedal, several things occur.  First, the signal going to the BBD does not consist of a disproportionately lows-to-mids spectrum.  Keep in mind there is generally more filtering AFTER the BBD than before it, and the regen signal is tapped from after the post-BBD filtering.  What this means is that when you add up the input and regen, what is getting fed to the BBD tends to be weighted in favour of lower end. 

The sample you tell us you like strikes me a having very little regen in it.

[Jaicen_solo]

The sample you tell us you like strikes me a having very little regen in it.

That's the point, although I do like it, Ideally i'd like to have more depth and resonance to the sound, just without all the extra crap that comes with that.
I think you're probably right that it's not possible with such simple designs. I see where you're going with the low/high bias, but are you saying that filtering the low frequencies in the regeneration path will help?  Mod  D above says that this will reduce the metallic qualities of the filter, is this what you're talking about?? I'm confusing myself now

[Mark Hammer]

There are a number of misconceptions about flangers.  One of them is that what you hear on record is comparable to what you can do by plugging one instrument into a stompbox.

The key thing about flangers is how noticeable the notches are, and especially how many there are at time T (the more noticeable the entire piopulation of generated notches is, the easier it is to notice their plentifulness or absence).  The general rule of flangers is that the greater your coverage of the audible spectrum, the better they sound.  Feeding an undistorted Les Paul into a flanger sounds...okay...but not inspiring.  Feeding an acoustic guitar into a flanger sounds pretty ggod.  Feeding a mono mixdown of an entire band (heavy on the cymbals) into a flanger sounds damn good.  Feeding white noise into a flanger sounds fabulous.  Most of the most impressive examples of flanging we tend to be aware of are post-production efects superimposed over multi-source mixes.  By virtue of having multiple sources combined, the listener notices all those notches all over the place, and especially when they go away and come back.  When the signal source is confined in bandwidth, that contrast isn't nearly so noticeable.

The way to overcome the "limpness" of using flangers on single sources is to increase the bandwidth of that single source (i.e., fuzz it up, baby) and increase the resonance setting.  Of course, if you don't like the boxiness that increased resonance imparts, and if you wanted to play something cleanly, you're kind of stuck.  By virtue of its greater focus on a few specific notches, phasers are are not nearly so dependent on the bandwidth of the signal source, and can sound impressive with a solitary clean instrument.

By trimming back the low end in the regen path, you get greater relative emphasis of the notches in the upper frequencies when regen is cranked.  If the instrument does not have wide bandwidth, what this can do is make the notches in the upper ranges of the instrument more audible/noticeable.  So, at point X in the sweep cycle, nstead of having a quintet of VERY noticeable notches and a couple of others that you can easily miss, you may end up with 7 or 8 equally noticeable notches.

[PenPen]

Mark, it is great to have you join this discussion. There's been a lot of discussion this past week about chorus/flangers, and I've had some questions to specifically ask you.

First, that is a good point about the bandwidth. from my experimenting with caps, I can't help but feel that there really ISN'T a point that I'm going to be happy with in this pedal. 10n is about the best I can hear, though its still a bit dry like I mentioned. So the range of 'good' sounds are very limited, 47n is the highest, 10n is the lowest before it is too high toned. And I still can't find a point I'm truely happy with. I think I'm going to have to end up putting together something to get what I truely want.

Which leads me to my next point. There was a thread a bit ago discussing the MN3207 as a replacement for the MN3007. Can this be easily done w/o affecting the sound of the pedal, as long as the power supply issues are addressed, of course. What exactly would need to be done for, say, the CE-2?

Lastly, I think what Jacen is saying, is the color the flange adds when the res is turned up is nice, but it would be cool if the sweeping could be stopped. Is this possible?

[Jaicen_solo]

I don't want to stop the sweep! That can be done by freezing the LFO, or something. Doesn't setting the speed to 0 do that on the BF-2?? I'm sure that's what the manual pot is for.
Anyway, that's not really what I had in mind, though now you mention it, I might look into designing a comb filter with a couple of MN3207's i've got that i've never used.
I think i'm just jaded by the sound of flangers nowadays. Can't say the flanger hoax excites me at all.

[nuttmeg1]

Hey Mark,

Is there a way to add a "delay" knob to the CE-2 circuit?


Thanks,
Brian

[PenPen]

I don't want to stop the sweep! That can be done by freezing the LFO, or something. Doesn't setting the speed to 0 do that on the BF-2?? I'm sure that's what the manual pot is for.
Anyway, that's not really what I had in mind, though now you mention it, I might look into designing a comb filter with a couple of MN3207's i've got that i've never used.
I think i'm just jaded by the sound of flangers nowadays. Can't say the flanger hoax excites me at all.

Ah. I apologize, I thought thats what you were describing when you said you wanted the tone without the annoying swoosh.

No, setting the RATE on the BF-2 doesn't fully stop the sweep. I can still hear it. The MANUAL knob seems to change the pitch or something (I don't know what word I'm looking for here) of the flange, and is active at any RATE setting.

[Mark Hammer]

Lotta questions to catch up on.

1) The Depth cotnrol on the BF-2, like many flangers, is essentially a mix/blend control.  What is it blending/mixing?  Two sources of control voltages that will make the clock chip do its thing.  One of those sources is a voltage that goes up and down (the LFO), and the other source is a DC voltage that is adjustable but steady.  If a person wants to "freeze" the unit and tune it, like the "filter matrix" setting on the EM, you have to turn the depth control all the way down.  This will effectively render the LFO silent (it is still going but very little LFO signal/voltage will find its way to the clock chip), and give priority to whatever control voltage is coming from the Manual control.  So, if you want to use the Manual control to adjust the freeze tone, turn Depth to min.  If the Depth control is turned up a bit, you can combine the Manual set voltage to redefine where the sweep starts from. 

At present, the Depth control is a 50k pot.  If a person wanted to be absolutely picky about cancelling the LFO, you could just wire up a 220k-470k resistor in series between the output of the LFO and where it connects to the Depth pot.  A SPST toggle would shunt the added resistor when you wanted LFO and add the resistor when left open circuit.

2) Can you make the starting point of the CE-2 variable?  Certainly.  One example is given in the BF-2.  That would be the portion of the circuit from the Manual control pot up to the Depth pot.  Rather than having the ground lug of the Depth pot go to ground, it would go to that added circuit fragment.

What I am concerned about is the range of control voltages provided by the Manual control.  That circuit may provide what is appropriate for the BF-2, but I have no idea if it provides something that is useful or suitable for the CE-2.  I can't see it blowing up anything, but there is no point in retrofitting the tunable chorussing if it's frustrating to set or yields insignificant improvement to the unit.  On a DOD floor unit I have, there is a 3007-based chorus module that has a tunable range function like I've described.  I would NOT describe it as any sort of "chameleon" knob that delivers radically dfifferent tones, but it changes the character of the chorus enough to warrant including it in the design.

3) "Tuning" the low-frequency rolloff in the regen path.  All we've explored so far is varying a single capacitor, so I would not expect this to make everyone as happy as it has made some.  The problem with single caps is that such "filtering" has a very shallow slope, that is far from the "brick wall" you would need to keep THOSE frequencies out of the regen path while keeping THESE ones in.  A cap change may *reduce* the bass, but it doesn't shut it out.  Moreover, it isn't exactly tunable.  A better arrangement might be to have a simple tunable passive 2-pole highpass filter.

To do this, you would keep the stock .047uf cap in the regen path, and add a second one is series.  You would then add a fixed resistor and one section of a dual-ganged pot to ground from the junction of the two caps and from the junction of the added cap and the 22k trimpot (which will undoubtedly have to be adjusted to a lower resistance setting).  What would work for you is not something I can say, but let's start with the following values.  First, score a dual 10k pot.  Taper doesn't matter.  The series resistors can be 15k or 18k.  As the resistance of each pot section is reduced from 10k to 0k, the rolloff will change from that lowest point (271hz with 15k; 242hz with 18k) to  its highest point (451hz with 15k; 376hz with 18k).  That may or may not be enough range for you, but it will provide some usable variation as well as cut out more of the bass in the regen.  Of course, whether you can stuff that all in the box is another matter.

[PenPen]

Ok, update. I just found a 22nF cap in an old cordless phone I was parting out, so I tried it in the BF-2. This is the one that I was looking for. Still a bit thick and a tad metallic, but clearer and more open. This is the value I like.

However, I still plan to get a DPDT mini-switch with center ON, and I'm intending to run 10n and 22n caps on either side, because 10n does still have a nice quality I'd like to use sometimes. Then, put it in the middle position, and you get a 32n cap as well, for going 'more metallic' sounding.

I can't give a conclusion yet until I test out 33n in there, but in all I think 22n is the sweet spot I was looking for, and 10n is also a great natural toned flange.