Boss flangers?

[kevn]

I'm thinking of getting a flanger and noticed a steady stream of boss bf-2's on ebay. They usually sell at about $50 and theres either made in Japan or Taiwan. The MIJ seem to be the more desireable. Does anyone know what the difference is between the MIJ & Taiwan models? For someone lwho ikes to tinker with pedals is one more desireable than the other? Thanks in advance - Kevin

[Hal]

I once played one of these a long time ago, and really liked it, but I havent really heard good reviews about them in general...

[nelson]

I used to own one, had it for around 6 months, modified it heavily and still didnt like it. Ended up on ebay.

My advise, wait for francisco at tonepad to finish the layout for the 9v elec mistress and build that.

[kevn]

Thanks for advise nelson, I might just give that a shot. My first build from scratch, the addiction is starting to take hold....

I'll do a search for tonepad and see what I can find.

[vanhansen]

The bf-2b bass flanger is really nice.  It's the same as the bf-2 with a few minor differences, mainly caps, that affect the sound.  To me the bf-2b is the closest I've been able to get to an MXR without getting an MXR (and I used to have an MXR that I sold... ).  A bf-2 can easily be modified to bf-2b specs.  It's the same circuit.

[George Giblet]

I think the Taiwan/Japan thing is just more mythical junk - they use the same circuit.

The BF-2 is an OK flanger, can be modded to sound better.

> It's the same as the bf-2 with a few minor differences

Yes!  The BF-2B uses a smaller delay chip (half the number stages).  So you have to change the chip or tinker with the VCO.  The other differences are very minor.

[Steben]

Is it possible to just swap the BBD chips maybe rebiasing the trimpots?
Is it a MN3209 that replaces the MN3207? mmm decreasing delay time...
I remember Mark Hammer suggested something like that.
Maybe it gives less noise because of the shorter delay line (common analog knowledge)?

[Toney]

If yr not familiar with it, check out the Boss HF2.....
Hi band octave flange....Fun fun....

[stm]

I owned a BF-2 for about a year. As it was said here, it is an "OK flanger". To my taste, more like a Chorus on steroids, if you know what I mean, but not really good for jet sounds. If you are aiming for radical jet flanging, then go with a design that uses an SAD1024 (like the electric mistress).

Looking at the schematics of the BF-2 and the HF-2 (Hi Frequency Band Flanger), the only relevant difference is that the latter uses 512  stages in the BBD (MN3204) instead of the 1024.  If you can get one, it will make instant conversion from BF-2 to HF-2.  If you can get a 256-stage MN3209, don't be afraid to try it; it should get you closer to those jet phasing sounds.

[Mark Hammer]

Yup, don't quote me on it, but it would seem that the fundamental difference between the HF-2 and the BF-2 is the use of a 3209 256-stage BBD, instead of a 1024-stager.

This will NOT score you an A/DA-sounding machine for peanuts, though.  Note that the most dramatic slow sweeps are achieved by having a wide sweep, from very short to longer delays, and by having the right sort of sweep waveform, such that the sweep decelerates as it approaches longest delay time and accelerates again as delay time starts to shorten.  Swapping the resident MN3207 for an MN3209 or MN3204 will shorten the minimum delaytime, but it wil do so by essentially shifting the delay range over.  So, if (hypothetically) your stock delay range was 1-10msec with an MN3207, swapping for a 3209 and retaining everything else about the circuit would shift things to a .25-2.5msec delay range.  So, you would get close to the through zero point (never truly reaching it), but wouldn't have the sweep width on the other end, at longer delays.

The upshot is that the flanging sound will be pleasingly subtle...and appropriate for bass and possibly acoustic guitar.  So, no $5 miracles, but a useful and pleasing change.  I don't know that HF-2 pedals are any sort of collector's item, but I tend not to see them around.  If purchase of a 3209 gives you an extra arsenal of tones from the same pedal, that may be a worthwhile expenditure.  Don't knock yourself out about it, but give it a shot if possible.  I would also recommend modding the regen path to take some of the hollow-metallic quality out of high regen sounds.  That may be what a lot of people actually want from their BF-2.

[Steben]

Another smack of questions :

* Will the BF-2 still sound like a chorus on the right settings if you put a 256-stager in it?

* Can the BF-2 by your experience handle more input than direct guitar? I'm asking because I thought this could raise the signal/noise ratio, you see?

* What if you put a 256-stager and another one (a 512 or 1024) after one another on different clocks but driven by the same oscillator? wouldn't that deepen the sweep? In theory not...

* Another thing: I noticed in the schematic that the mix resistors (dry-wet to opamp) are not equal in the end, where other flangers do feature this: loss of jet sound or is the circuit different and made to compensate?

I have a friends' Ibanez FL301 here also because a trimpot broke. I'm keen on comparing them soon...

[stm]

Hi, let me clarify that Boss HF-2 uses an MN3204, a 512 stage BBD (and NOT a 256 stage unit like the MN3209!).  So, it has HALF the number of stages in there.

* Will the BF-2 still sound like a chorus on the right settings if you put a 256-stager in it?

With a 256 stage BBD probably not. The max delay for the flanger is already set for a shorter value than the chorus. IIRC, Boss Chorus specs stated a delay between 4 and 20 ms, while the Boss Flanger states 1 to 16 ms; both units use the same 1024 stage BBD, so the clock frequency ranges are indeed different.

* Can the BF-2 by your experience handle more input than direct guitar? I'm asking because I thought this could raise the signal/noise ratio, you see?

Probably the input level adjustment is already optimized for maximum S/N ratio. In fact, I've always noticed with analog delays, chorus and flangers that if you use a hot pickup and pick with anger you may hear some clipping. There is a tradeoff between how much clipping you are willing to tolerate and the noise floor you'll get in the end.

* What if you put a 256-stager and another one (a 512 or 1024) after one another on different clocks but driven by the same oscillator? wouldn't that deepen the sweep? In theory not...

Unfortunately it would only affect the amount of delay, but the ratio or quotient between MAX and MIN delays will remain the same. In the Boss Flanger it is around 16:1.  As Mark explained before, the larger this ratio, the more radical tha delay. I think between 25:1 and 30:1 you are set for good jet phasing, provided minimum delay goes as low as 0.5 msec or less--not in the ballpark of the stock BF-2.

* Another thing: I noticed in the schematic that the mix resistors (dry-wet to opamp) are not equal in the end, where other flangers do feature this: loss of jet sound or is the circuit different and made to compensate?

This has to do with the pre-emphasis and de-empasis applied to the high frequencies in the delayed path. In the end you have unity gain--to verify this yourself you need to follow this path all the way from the input to the output, taking into account the input frequency as well as the resistor and capacitor values in the feedback loops of the OpAmps.

Suggestion for improvement of BF-2 flanger sounds:

1) Get and place an MN3204 BBD instead of the MN3207 (save this for a chorus!). This will change the stock delay range from 0.5 msec to 8 msec.

2) Tweak the LFO section so the minimum clock frequency gets near to 20 kHz, thus producing 12.5 msec delay out of the MN3204 BBD, without altering the maximum clock frequency.

3) Changes 1) and 2) should give now a 25:1 sweep ratio and lower minimum delay, which should get you closer to jet flanging sounds.  I have the belief this is doable. Maybe you might ask the commercial MOD gurus to see if they have such a mod for the BF-2.  I don't have have a BF-2 unit to experiment myself though.

Good luck.

[Steben]

Well, thanks stm. Excellent reply.
I need to do those mods together with the anti-metal regen mod.
Only the LFO mod seems a harder nut to crack. Best ways to start this? A lot of trims, but widening the sweep...?

I actually used the BF-2 rather a lot as a weird chorus. Well, since jets are not coming...  On the other hand a chorus is used more often than a jet flanger (I mean its less a one trick pony), so if I mod the BF-2, I need to build a chorus. Hollis' Zombie chorus seems a quick build?

[stm]

Well, thanks stm. Excellent reply.
I need to do those mods together with the anti-metal regen mod.
Only the LFO mod seems a harder nut to crack. Best ways to start this? A lot of trims, but widening the sweep...?

Well, hope it helps. The trick for the LFO change is to widen the range in ONE DIRECTION only, which is the longer delay. I'll take a look at the schematic and come back later with something to start with. Please keep us informed on your progress (either success or failure), since this is really interesting. Are you planning on getting an MN3204 ?

I actually used the BF-2 rather a lot as a weird chorus. Well, since jets are not coming...   On the other hand a chorus is used more often than a jet flanger (I mean its less a one trick pony), so if I mod the BF-2, I need to build a chorus. Hollis' Zombie chorus seems a quick build?

You will still get those standard/milder flanger sounds as well, however as you mention the chorus sounds are not going to be so close to a Chorus now.  If you are planning on building a chorus, I would suggest a CE-2 from Tonepad's layout. A classic! It uses a MN3007 instead of a MN3207 though, however it is better than the Zombie Chorus due to the additional circuit complexity.

[Steben]

Well, a CE-2 is a BF-2 without feedback and higher delay setting... 
Maybe I should mod to a CE-2 instead ?

By the way: better due to complexity? Do you mean filtering etc...? Better modulation wave shape perhaps. Zombie chorus uses MN3007 as well.

[stm]

Well, a CE-2 is a BF-2 without feedback and higher delay setting... 
Maybe I should mod to a CE-2 instead ?

It depends on what you want: a good chorus or a good flanger. If it were for myself, I would build a CE-2 and try to improve the BF-2 as a flanger. It is easier to build a CE-2 than a BF-2 clone.

By the way: better due to complexity? Do you mean filtering etc...? Better modulation wave shape perhaps. Zombie chorus uses MN3007 as well.

The CE-2 has better filtering than the zombie, and I have the feeling the modulation is more even across the sweep also. The zombie sounds more to me like ....eeeE!eeeeeE!eeeeeE!eee... at the turning points of the LFO.

In case you are still interested in attemping the mod on the BF-2, I studied the schematic available from the Free Information Society and came to the following changes in order to achieve the increased sweep after replacing the BBD for a 512 stage MN3204:

1) IC5 needs to be replaced with a full rail-to-rail output opamp like the TLC2262, TLC2272 or MAX492, to name some.

2) Then, resistors R36 (180k) and R37 (220k) need to be swapped with each other to increase the output voltage range of the LFO.

The use of a rail-to-rail output opamp is necessary, otherwise the LFO won't work with the extended output voltage range.  Eventually, but not necessarily, VR7 might need a slight adjustment in order to center the new range of the sweep.

In summary, the required changes are pretty simple, except for the difficulty in getting the MN3204 and TLC2262 ICs.

Take care.

P.D. Maybe Brian (wampcat1) could verify/add something to this discussion.

[Steben]

Isn't a 256-stage MN3009 or MN3209 (latter better with MN3102 inside) better for the flanger job? (although not that easy to find too).
According to datasheets it has "minimum of maximum" 0.64msec delay and maximum delay of 12.8msec (this is 20:1). You cannot get more than 20:1 with MN-series. So if the pedals are stock 18:1 there is not that much modding left. I wonder whether a rail-to-rail opamp is worth the cost.

And why is a SAD1024 so much better in this? It gets a minimum of 0.5msec and a max of 200msec. Is it another technology or what?

[stm]

This is one of the few cases where the datasheet (DS) can be disregarded to some extent.

Let's take the MN3207 for instance: The DS states a clock range from 10 kHz to 200 kHz, producing a delay range from 51.2msec to 2.56msec.  How is it possible then that Boss' specs for the BF-2 indicate 1 msec then?  Simple, the MN3207 is clocked at 500kHz instead of the 200 kHz maximum indicated in the DS.  A first thought might lead you to think "oh boy this IC's gonna get fried!".  Fortunately it is not the case. In fact some MN3007's (which are rated for 100kHz max. clock) have been reported to operate as high as 1.5 Mhz.  The key here lies in the "strength" of the driving circuitry. The high frequency limitation is more related to the high input capacitance of the clock lines that need to be driven in order to obtain a decent frequency response.  So, a MN3204 (512-stage) driven at 500 kHz might leave you with 0.5 msec.

On the other hand, the minimum USEFUL lower frequency clock is around 20 kHz (10 kHz is too low!) in order to have at least a adequate bandwidth so the flanging effect is produced. The filters on the CE-2 and BF-2 are around 7 kHz. (BTW, Nyquist states you would need only twice the sampling frequency or 14 kHz in this case, however that's only true if you have "brickwall" filters that produce infinite attenuation above 7 kHz which is not the case. You need this 3:1 ratio between clock frequency and highest frequency of interest for practical reasons.)  So, at a 20 kHz clock using a MN3204 you will have basically 12.5 msec of delay.

Finally, 12.5 msec / 0.5 msec = 25:1 ratio, which is certainly better than the 16:1 BF-2 ratio. And that together with the reduction of the minimum delay from 1msec to 0.5msec is what should get the overall effect closer to jet flanging.

Whether you consider taking the chance of purchasing the ICs for performing the mod or not is only up to you! As in any mod there is some amount of (hopefully controlled) risk, so you and only you can decide if it's worth it.  ("I've only opened the door. Now's your turn to decide if you enter or not." or something like that.)

Best regards.

EDIT: I didn't mention that the suggested LFO mod would also lower the minimum delay, so probably the final sweep range might be better than the 25:1 calculated above.

[Mark Hammer]

I had to be somewhere else, so thanks to STM/Sebastian for picking this one up.  Thanks for noting that the HF-2 uses 512 stages not 256 as I incorrectly remembered (serves me right for guessing without having my schem collection on hand).

"Isn't a 256-stage MN3009 or MN3209 (latter better with MN3102 inside) better for the flanger job? (although not that easy to find too).
According to datasheets it has "minimum of maximum" 0.64msec delay and maximum delay of 12.8msec (this is 20:1). You cannot get more than 20:1 with MN-series. So if the pedals are stock 18:1 there is not that much modding left. I wonder whether a rail-to-rail opamp is worth the cost.

And why is a SAD1024 so much better in this? It gets a minimum of 0.5msec and a max of 200msec. Is it another technology or what?"

There is easy and there is better.  Having only 256 stages means that some fairly pedestrian clock frequencies can be used to produce fairly short delays.  Of course, when frequencies that low can be used, that also means there is greater risk for clock whine becoming audible (though less so in this particular instance).  In some respects, it can be a better choice to use the same higher-capacity BBD, and use an improved clock circuit to step it through its paces even faster.  That will put the clock pulse way up in the inaudible range, which reduces a lot of the filtering needs on the delay audio path.

The MN3xxx chips HAVE been successfully clocked as high as 2mhz.  The thing is that the datasheets show what you can do with a Matsushita MN3101/3102 clock chip, and do not discuss what is feasible under other arrangements.  I'm not as well informed as some regarding what the actual clock pin input capacitance is for the Reticon vs Matsushita chips, but the Reticon is sdescribed in datasheets as being "normally" clockable up to 2mhz.  On the other hand, the clocks shown are almost always CMOS clock circuits.

[BDuguay]

Kevin.
I have a MIJ Boss Flanger that's for sale.
PM if you're interested.
Thanks,
B.