In case you wanted a through-zero-flanger

[Mark Hammer]

The "Infinite Flanger" is a flanger project that uses a pair of TDA1097 BBDs.  Like most BBDs other than MN3007/3207/3008 and 3208, these are a bugger to find, but if you can find some, or have an inkling how to adapt others to this project, I think you'll find it interesting.  In any event, I've never seen any other full schematics for a through-zero flanger or any published attempt to clearly explain how TZF works.

You can find it here http://www.milton.arachsys.com/nj71/index.php?menu=2&submenu=6 as a large zipfile containing each of the scanned pages.

The FoxRox "Paradox" is a very nice full-featured through-zero-flanger, and you can find a pleasing selection of illustrative sound clips at Dave Fox's site.  It is, however, a professional product for those of certain pocket depth or wallet thickness.  If you don't fit that category and sadly cannot become one of Dave's customers, take a peek at the Infinite Flanger.

[toneman]

WOW!
javascript:emoticon(':D')
Thanx! 4 the link, Mark

want 2 try this with 2  1024Rectocons.

wonder if i can mod my Paia Phlanger(?).....

so many circuits
so little time..
javascript:emoticon(':(')

[Jason Stout]

Thanks Mark! I have never been to that site.

[puretube]

nice "TZF" (reel-to-reel) listening:
"Itchycoo Park" by Small Faces,
"Sky Pilot" by Eric Burdon,
"Wasn`t Born to Follow" of the Easy Rider album...

[zachary vex]

sky pilot is TZF?  nauseating.  i'd suggest "life in the fast lane" by the eagles or "listen to the music" by doobie brothers to hear real mind-bending thru-zero flanging.

[Peter Snowberg]

How about "The Confessor" by Joe Walsh? I don't know.... I'm just guessing.

-Peter

[puretube]

sorry, how could I forget the doobies?

sky pilot was the first time I remember of ever having heard TZF
(on drums and vocals), and it was a main-contributor to get me
intersted in sound-manipulation...


breadboarders/builders: beware of beat-frequencies!

(b.t.w.: the "easy Rider" track was by the Byrds...)

p.s.: I do like the 3-color display feature in that I-F article...

[Bill Bergman]

Where are you gonna find two TDA1097 BBDs?

[puretube]

just use similar...

[Bill Bergman]

And that would be........

[puretube]

xx3207 plus xx3102...

(maybe have to adapt the circuit a little.... xx stands for MN or BL)

[MR COFFEE]

In any event, I've never seen any other full schematics for a through-zero flanger or any published attempt to clearly explain how TZF works.

Mark,
I don't recall where or when I read it, except it was a looong time ago (early 70s). IIRC, through zero flanging is "simply" a matter of using two delay lines in parallel whose output is mixed 50-50, with the additional condition that you either,
1) use one fixed delay line and one which varies the delay in standard flanger style such that the delay time ranges between less than the fixed delay to greater than the fixed delay, or
2) both delay lines shift in length with the LFO but it opposite directions, i.e., as the LFO voltage increases the clock frequency of one delay line, it decreases the clock frequency of the other.
The "through zero" effect is what happens when the time difference is passing through zero difference in delay time. Except for the unique sounds as you pass "through zero", the effect is the essentially the same as a flanger. Not to say it isn't pretty awesome around zero! :mrgreen:

[Bill Bergman]

Thank Pure.
I have some MN3005, MN3002??, and Sad1024's. Any of those work?

[Doug H]

Thanks for the link, Mark. I think Dave Fox's TZF flanger is one of the coolest things to come out in a long time. I'll probably just save my $$ and purchase that one day, but it's nice to know there is an alternative for DIY-ers too. It's also a nice resource for learning how you would go about such a thing.

Doug

[Mark Hammer]

An SAD1024 could make an excellent TZF if you use it as two separate 512-stage BBDs à la MN3010.

The trick in TZF is to have the "dry" signal be very slightly delayed enough that the shortest delay time of the swept BBD ("wet" signal) can become shorter than the "dry" signal.

Operationally, what does this mean?

The trick is to achieve a delay offset which is largely unnoticeable to the user while playing, but which allows the swept delay an opportunity to reach the "finish line" before the dry signal (fixed delay) during some parts of its sweep.

Without outside help (i.e., buffers, etc.) the standard MN3007/3207+MN3101/02 combination has a difficult time clocking high enough to reach delay times of much less than 1msec.  Assuming one used such a chip combination, the "dry" signal would then need to be offset by *at least* 1msec for the swept delay to come out before the dry signal.

For the most transparent feel, from the performer's stance, the shorter the fixed delay is, the more usable the effect is.  Put another way, if the "dry" signal had to be delayed 5msec to achieve TZF, it would throw off your playing to have that sort of lag introduced.

So, in the grand scheme of things, you want to aim for the fixed delay to be as short as it can possibly be, while still providing more delay time than what is provided by the swept delay at that point in its sweep cycle where the delay is shortest.  For example, if the swept delay went from 1msec to 8msec, and the fixed delay was 1.5msec, then a small portion of the sweep cycle would be spent with the swept delay coming out ahead in time of the fixed signal.

In truth, the through-zero effect happens at the crossover point, so as long as it goes through and comes back again, no matter how far to the other side the swept delay goes, you get to hear that classic sound-sucked-right-out tone TWICE (on the way there, on the way back).  The difference bewteen the fixed delay and shortest delay in the sweep cycle essentially determines the "feel" aspect, as near as I can tell, without affecting tone in any discernble way.

So, how does this translate into chip choices?  Since you want the lag from the fixed delay to be as imperceptible as possible, you need to use a chip for the swept delay that can go as low/short as possible so as to reduce the requirements for that fixed delay.

You CAN get a variety of BBD chips to clock ultra fast to produce short delays but you want to keep design requirements as low as possible so the default choice is a BBD with fewer, rather than more stages....up to a point.

In some respects, a chip with 256 to 512 stages is ideal.  Certainly 128 stages can get you very short delay times, but you also want the clock rate needed to attain the required delay to be high enough that you don't need to filter the life out of the audio signal.  512 stages is in the zone here.  Nailing sub-1msec delays is not difficult to do, and it is easily done with a 512-stage device using a clock rate that does not enter the audible range too much.  In this respect, an MN3002 or a 256-stage MN3009 or half of an MN3010 or SAD1024 is ideal.  The SAD1024 provides for independent clock inputs for each "half" of the chip, which suits our needs here just fine.

Does all of this make sense?

[Bill Bergman]

It makes sense, but knowing how to modify the circuit to work with any of the other suggested BBD is beyond my capabilities.

[Mark Hammer]

Sadly, me too.

[Bill Bergman]

Mark,
So what your telling me is you are just all talk.... :lol:  :lol:  :lol:  :wink:

[D Wagner]

Does anyone even know where to find the datasheet for the TDA1097?

Google didn't turn up anything useful.

[Mark Hammer]

Hell, not EVEN talk...just a lot of arms flailing about.   :lol: