about flanger effect delay time

[kugua]

i have a problem about the flanger delay time,someone said flanger delay time must be less 10ms??

[kugua]

or tell me some web about design flanger effect.thanks!

[Mark Hammer]

The delay range of a flanger and a chorus overlap.  What makes something sound like a flanger is the minimum delay produced.  So, if something sweeps from a 1msec delay to 10ms (and back again), it will sound like a flanger.  If it sweeps from 6msec to 12msec (and back), it will sound like a chorus, even though there is much overlap in the delay time used.

The major difference, in my view, is that once the minimum delay time is long enough, there are enough notches introduced into the signal that you stop paying attention to the notches, and pay more attention to the pitch shift, which becomes more noticeable as delay time increases.  In flangers, it's the opposite.  Very fews audible* notches are produced at the minimum delay time, so we notice as the signal gets filled up with notches.  And, of course, at very short delay times (and especially slower sweep speeds), any pitch change produced is pretty much inaudible to the listener (or at least not noticed spontaneously).

The most dramatic flanging sounds are produced when the ratio of minimum to maximum delay time is at least 20:1, the wet/dry balance is exactly 50%/50% and the minimum delay time is well below 1msec.  Sadly, many of the inexpensive flangers that use the Matsushita chipset (MN3207/MN3102) in the simplest possible configuration may only sweep from 1msec up to 12msec.  This *will* give flanging sounds, but not the sort that make you go "Wow!". It is possible to improve upon that (produce very very short minimum delay) with additional circuitry, but then you start to move outside the range of inexpensive small products.

Some people prefer flangers that include a compander to reduce noise and distortion, while other people like to leave companders out of circuit, saying it takes away too much feeling.

(*I say audible, because if the notches don't start until 10khz or so, most of the notches produced at the shortest delay times will be beyond the range of hearing of the range of most guitar amp speakers.)

[Hal]

nice post.  I _just_ read the GEO "technology of" article about flangers/phasers.  It was really really clear and stuff...required reading :-D

[kugua]

hi mark, thanks,i design a flanger effect.and i build it. I have setting the delay time is 1ms to 5ms ,but not flanging.I think is my feedback situs selected wrong?or feedback phase wrong?I dont know why?what about the feedback trait,in-phase or another?    waiting your reply.

[kugua]

i have another question about flanger modulate signal,i wanna know  the difference about the modulate signal between chorus and flanger . :icon_rolleyes:because i have made a flanger but no flanging sound occured.only the chorus sound.

[kugua]

anyone help?

[Mark Hammer]

I can help but it is hard to understand you

If your circuit only sweeps from 1ms to 5ms, it will not make a very good flanger.  You might want to read the interview with Styeven St. Croix here: http://ampage.org/hammer/files/Device1-3.PDF

Feedback or "recirculation" can be in phase or out ofphase.  Both sound good, and usually a good flanger will let you use each kind.  I think the best way to undersatdn about flanger feedback is to look at some different designs.  Here are some good links:
http://ampage.org/hammer/files/PAiA_Phlanger.PDF
http://ampage.org/hammer/files/Anderflange1.PDF
http://ampage.org/hammer/files/Hyperflange.PDF
http://ampage.org/hammer/files/BOSS_BF-1.pdf
http://www.freeinfosociety.com/electronics/schemview.php?id=117

[stm]

...and don't forget the AMS-100 flanger project that appears on Device Magazine Issues 1-9 and 1-10. Really nice and quiet. I would describe it as intermediate difficulty.

If you just want to start experimenting you should try first with John Hollis' Ultra Flanger. A nice starter circuit with a PCB already designed.

[stm]

As Mark mentioned before, the Matsushita MN3007 (and lower voltage friends) are hard to push for delays lower than 1 msec. Nevertheless the Ultra Flanger uses a clock from 50 kHz to 1 Mhz that produces a delay range from about 0.5 msec to 10 msec. Notice that tha 1Mhz clock is out of spec--the MN3x07 states 100 kHz maximum. You won't damage your IC if you push it further, however the high frequency response is degraded, and thus the dramatic flanging sounds that should appear at low delays are way less interesting.

Someone explained in an older thread that in order to use high frquency clocks (of up to 2 MHz) with the MN3007 you need strong drivers on the clock lines, which the MN3101/MN3102 are NOT.

I had some experience with this like 12 years ago. I was experimenting with a MN3007 based chorus (driven by a CD4013 complementary outputs) and despite was obtaining flanging sounds, they were not really dramatic. The same I found with boss flanger BF-2. One day, a cloned stompbox of what today I think should have been an Electric Mistress came to my hands... It was another story! Ear-piercing highs and extreme jet flanging sounds lived in this pedal.

This baby was lend to me by a friend for one night, so without even thinking twice I opened it and found an ugly brown PCB, hand drawn with an indelible marker and etched at home. The IC's were painted in black, so could not read what they were, however, I took a cotton dampedened with alcohol and voila: SAD1024, CD4013 and a quad op-amp (perhaps an LM324 or RC4136).

It was my lucky day; My father (who ia also an EE) had a datasheet of the SAD1024, and the beast was put in a 16-pin socket. I took it out and rearranged my flanger circuit (laid out in a protoboard) in order to use the SAD1024 instead of the MN3007.  Instant success!!!

Of course the SAD1024 is way too noisy (like 70 dB S/N ratio v/s 84 dB S/N ratio for an MN3007), so its use should include a compander to tame noise. In this respect the AMS-100 flanger I cited before is a great application circuit.

The moral of this are:

1) I use the MN3007 (and friends) only for chorus circuits -> take a look at the CE-2 project posted at Tonepad!

2) I either use an SAD1024 with a Compander (NE571) for flanger, or an MN3009 BBD, which is a device capable of delays from 0.20 msec (at 600kHz clock) to 6.25 msec (at 20kHz clock), thus easily allowing a sweep ratio of 30:1. The MN3009 is a 256-stage BBD with better noise and higher clock ratings than a MN3007, and it happens to be available here at a local store (no more MN3007's though ).

[Mark Hammer]

As Mark mentioned before, the Matsushita MN3007 (and lower voltage friends) are hard to push for delays lower than 1 msec. Nevertheless the Ultra Flanger uses a clock from 50 kHz to 1 Mhz that produces a delay range from about 0.5 msec to 10 msec. Notice that tha 1Mhz clock is out of spec--the MN3x07 states 100 kHz maximum. You won't damage your IC if you push it further, however the high frequency response is degraded, and thus the dramatic flanging sounds that should appear at low delays are way less interesting.

Someone explained in an older thread that in order to use high frquency clocks (of up to 2 MHz) with the MN3007 you need strong drivers on the clock lines, which the MN3101/MN3102 are NOT

You may have read it in an earlier thread, but in case not, what Mike (Vsat) found was that the chief limiting factor in the Matsushita chips is the capacitance on the clock input pins.  That is, I gather, also one of the reasons why you will see clock drivers like the MN3101/3102 rated in terms of how many stages of delay it can drive.  Why should it matter how many stages?  Because the input capacitances of those BBDs are in parallel with each other.  Increasing the capacitance that the clock driver "sees" means that it is limited with respect to how brief/fast a clock pulse can be before it stops being a pulse and starts being more like a triangle wave (NOT very good for stepping!).  If you buffer the clock lines and provide more current in the clock pulse, it is able to overcome the problem created by the clock input pins.  That's why the best flangers you find often have things like a trio of invertors in parallel between the clock pulse output on the MN3101 and the clock-pulse input on the MN3007 (or whatever the BBD is).  As Mike has indicated to me, he was able to run a 3007 at 2mhz without having it get warm to the touch.  The ratings you normally see on the Matsushita datasheets are accurate, but they are for using one of their clock chips connected directly to one of their BBDs, without any "outside help".  I think that is fair, since they have no idea what sort of outside help you may want to use.  At the same time, it tends to mislead people about what is actually possible.

You will note that the Ultra-Flanger uses buffered clock lines.

[StephenGiles]

But noise apart, I prefer the sound of SAD 1024 flangers. The MN 3009 is a much different kettle of fish (una pava distinta de pescados!) I don't think they are available in the UK.
Stephen

[puretube]

the power-current consumption of BBDs rises @ high clock-frequencies...

[kugua]

hi mark. I could not download this pdf file.http://ampage.org/hammer/files/Device1-3.PDF
.why?
Object not found!
The requested URL was not found on this server. If you entered the URL manually please check your spelling and try again.

If you think this is a server error, please contact the webmaster.

Error 404
ampage.org
Apache
could you mailing me?thanks, my email :     kuguali@yeah.net

[Connoisseur of Distortion]

the DEVICE flanger is quite a beast. i started on a layout, but never really finished it... maybe because it's so damn big? 

[kugua]

thanks everyone,i have found these files ,a good message for everyone,my design flanger is working,