A/DA Flanger does TZF?

[jmasciswannabe]

Sorry bout that racedriver....

I posted this over on the board order page but may as well do it here to.

Actual value of parts shown in Charlie's schematic. Had some time to blow at the in-laws Thurs. Might help when ordering parts....

c1 - 0.01
c4 - 22uf
c7 - 0.47
c18 - 1500p
c29 - 39p
d1 - 1n914 x 4
d5 - 1n914 x 2
p3 - 10kL
p4 - 50kL
p5 - 500k rev log
r3 - 2.7k
r5 - 1m
r9 - 20k
r11 - 47k
r13 - 10k
r28 - 1.3m
r39 - 22k
r40 - 68k
r43 - 27k
r44 - 5.1k
r51 - 22k
r55 - 47ohm
r60 - 1k
r61 - 2.7k
r63 - 20k
r64 - 20k
r65 - 82k
r68 - 47ohm
r69 - 2.2m
t5 - 100k

Can't wait to get this sucker started!!

[analog kid]

Does anyone have any sound samples of this new contraption?

   not so new a contraption I'm afraid , though I'm sure you're aware of this.
 As far as clips (of an A/DA Flanger) go to Modzero , they ARE there.
 'does it do flanging like using two tape recorders?.... Well, that's what any analog Flange unit Is MEANT TO TRY to emulate for sure! as to how well it does or you prefer the sound of an analog/digital flanging device to that of true old fashioned tape flanging , well I guess that's quite definately a matter of taste and opinion.

Insert Quote

If anyone has built one, how about a clip of distortion into the Tru-Zero version?

 
   you are looking to find sound clips of an "a/da" equipped with "thru-zero" huh? Well I'm quite sure it's quite obvious it never produced with such a feature and so few people have built it I am just as sure noone has bothered with attempting to add a board for thruzero. This is just somthing that's been talked about "what IF" and hence pads get added to allow some person so inclined to give this a shot!    So soundclips? obviously not gonna find that      although I agree with the previous that this is really uneccessary and perhaps just overkill with this effect.   to me the most obvious unique-ness of it IS the extreme width of the sweep. Mine really is too wide if not very carefully set up. Still alot of trouble with overall Fr.response but i think it'll be 'smoothed out' real soon as well.
  Uh, what's this Parts value list for??
 I think Charlie did quite a big help for all y'all by putting up a side by side part #/value list for each of the later Rev's and M.I. suggested.   a shame for me, considering the seemingly hours I spent squinting and scrolling over every schem/parts lists/pt layout document I could find to put together over the years to make my own such list awhile back!    guess I coulda waited

I think anyone building this is gonna want or have to pretty much let decide for themselves on quite a few of thse values , I could point out several of which I believe this to be the case with but  hey ... what's the fun in that!   Now I notice that Moos' is mixing a little bit of Rev 4 stuff with Rev3 in areas so it's even more of a "composite" revision than if just building a straight up Rev3 audio path,etc..with SAD mods (more little bothersome incompatibilities?i hope not) Though I'm sure he wouldn't do it unless he picked up info and advice for it somewhere along the way...I just must've missed it)  Also not to confuse but I don't understand where these 2k7' values came from for r47 and r61? though charlie's redraw IS using the setup from Rev 4 around IC4a/b  Though I'm sure he wouldn't do it unless he picked up info and advice for it somewhere along the way...I just must've missed it) big difference from the 100k for r47/61 as in rev3 but I still haven't seen 2k7 come up for these even in rev 4 doc's.       2k7 for r3 is however listed as such BUT I find it to be quite to much high pass in the pre emphasis at Input and the 4k7 listed elsewhere will likely end up being the preferred value here. 
If anyone can explain to me the r47/r61 detail that I'm obviuosly missing completely, thanks

 

[RaceDriver205]

I see. So the thu-zero option is not going to be added to this device? I was under the impression that it was going to be a main part of it?

[Mark Hammer]

The desirable aspect of the A/DA Flanger was that the clocking scheme allowed it to get very close to zero, and probably much closer than other compact floor pedals of that era and all subsequent analog pedals.  There IS a certain magic to the through-zero experience, but it has less utility in everyday playing than simply extending the sweep range a little farther.  It's like the difference between truffles (the pig-found fungus type) and a decent cup of coffee - sure, the one is more exotic and mind-blowing than the other, but a decent cuppa java is something you probably couldn't face each day without (and say your blessings for with every sleepy sip), where a truffle is something you want to reserve for special occasions.

[StephenGiles]

Absolutely Mark, and a very good morning to you, I'm never sure whether you are 5 or 6 hours behind us! From the samples I have heard of TZF, the volume drops to a level that would be unacceptable in a band situation, but that's not to say it would sound very flash in your home on it's own. Nevertheless, a building block of a circuit would always be welcome in this area, if only for something new to try, and there may well be ICs available which already contain much of the circuitry required, or maybe just require a few extra components to build say the quadrature LFO. Are there chips with built in dome filters I wonder?

[oldschoolanalog]

I see. So the thu-zero option is not going to be added to this device? I was under the impression that it was going to be a main part of it?

Go all the way back to pg. 14, post 260. Your answer awaits you there...

[moosapotamus]

Right... If you go back through this thread, I think you'll find that the consensus was to make it easy to add a number of different mods, including TZF, by incorporating additional pads in the layout. So, I did not spend any time designing the necessary additional circuits (building blocks) for things like additional buffer(s) for stereo outputs and additional delay line(s) for TZF. However, I did include those additional pads in the right places so when folks do come up with ideas for those building blocks, they can build them up on a daughter board and easily tie them in to the main PCB via those additional pads.

Personally, I'm not a big TZF fan, either. To me, it just sounds like the volume drops off. But, I am really interested in trying out the bounce mod.

Now I notice that Moos' is mixing a little bit of Rev 4 stuff with Rev3 in areas so it's even more of a "composite" revision than if just building a straight up Rev3 audio path,etc..with SAD mods (more little bothersome incompatibilities?i hope not) Though I'm sure he wouldn't do it unless he picked up info and advice for it somewhere along the way...

Pretty much all of the advice I picked up along the way is either in this thread or linked to within this thread... And, a huge portion of it came from Stephen's reports regarding his working build and his excellent schematic. What I put together is basically a mix of the Irwin/Giles SAD1024 adaptation and the ADA Rev4 schematic. Stephen reported that he actually wound up using the diode limiting configuration from Rev 3, based on Mike's recommendation, so I used that in my scheme & layout. Unlike the Irwin/Giles design, I also included the threshold section from Rev4. That's about it, aside from nailing down parts' values in some places.

I don't understand where these 2k7' values came from for r47 and r61?

Those are the values shown in the copy of the Rev4 scheme that I have. For the LFO section, I went with the Rev4 design because, aside from a few of the parts' values and the inclusion of R47, it's identical to the Irwin/Giles version. So, it's possible to populate the board either way.

2k7 for r3 is however listed as such BUT I find it to be quite to much high pass in the pre emphasis at Input and the 4k7 listed elsewhere will likely end up being the preferred value here.

analog kid - That's great to know... especially for all the bass players. You've been really providing some great info for tweaking and de-bugging this circuit. There's a wealth of information here that could be compiled into a really useful set of build notes.

Everyone, FYI... I'm now accepting payments for the PCB group order. Info is here...
http://www.diystompboxes.com/smfforum/index.php?topic=51692.msg389818#msg389818

Thanks!
~ Charlie

[analog kid]

 Yes I see, it is pretty much the LFO section that you went with Rev 4 , it took a few minutes before I realized exactly 'where'' you were using Rev3vs4
"Those are the values shown in the copy of the Rev4 scheme that I have. For the LFO section, I went with the Rev4 design because, aside from a few of the parts' values and the inclusion of R47, it's identical to the Irwin/Giles version. So, it's possible to populate the board either way".
  that's interesting. I don't recall having any rev 4 stuff with those values.  those parts are setup the same but not the values. I'm gonna have to dig back through all my notes and check that out. 
Also as far as being able to poplulate it either way, I don't see how that is( though i'm probably seeing wrong as usual ; )  For instance in this case of r47/r61 , these are set up a bit differently around IC4 , regardless of values.

I see. So the thu-zero option is not going to be added to this device? I was under the impression that it was going to be a main part of it?

  No you were asking about sound clips of TZF, as I said:   

" I am just as sure noone has bothered with attempting to add a board for thruzero. This is just somthing that's been talked about "what IF" and hence pads get added to allow some person so inclined to give this a shot !"   

      So yeah Charlie's obviously sticking xtra pads in every location feasable to allow playing with the ideas of about every mod that has been discussed or even suggested in our big thread.  up to a few months ago (and I'm still content )i'd venture to guess myself and maybe anyone else would've been so much more than content to pull off a cloning of this unit STOCK. even stripped down from stock for that matter. Like Stephen's
  Tough as a build as it may already be for some , parts placements spelled out on a premade board or not, I am curious to see how much brain meltdown takes place from adding loads of extra features at once. I predict many 'ts' threads.ah well  why not, that's how we learn.   Hopefully everyone using that board will try to get the stock flanger going first so that artwork is for sure verified. 
I am pretty interested in the bounce idea myself Charlie.  I may try it out on mine before long. but I've enough on my plate just getting freq response and too much modulation with most of T5 even using a large trim. I CAN still obvioulsy tame down the total modulation for the sweep to be very smooth indeed but WITH a large T5 trimmer. increasing r65 doesn't seem to bring this into more of a useable trim range either (?)

[StephenGiles]

And I still recommend building one section at a time, to ensure that all is well as you build. For an inexperienced builder this project would be quite a challenge, and faced with a fully populated board which doesn't work.............

[puretube]

I could not agree more than that, sir!

but we have to face the fact,
that we`re living in the age of:

"if it don`t work, BuyYourOwnClone"
or : "I Did it My E-W Bay"

 

[RaceDriver205]

Righto, so I guess if TZF isn't that great I should just buy 1 SAD1024?

[oldschoolanalog]

I cringe when I hear SAD1024 and TZF board mentioned in the same sentence .  Please save these soon to be extinct bbds for more important stuff. An MN/BL3207 should suit TZF just fine. A whole lot lighter on the wallet too!

[moosapotamus]

I don't recall having any rev 4 stuff with those values.  those parts are setup the same but not the values.

Right. That's exactly the point I was trying to make... the configuration is the same. So, you can build it either way by just using different parts' values.

Also as far as being able to poplulate it either way, I don't see how that is( though i'm probably seeing wrong as usual ; )  For instance in this case of r47/r61 , these are set up a bit differently around IC4 , regardless of values.

The way it looks to me in the documents that I have, if you take R47 out of the Rev4 LFO you get the same configuration as the Irwin/Giles LFO. Again, the only other difference being some of the parts' values.

Righto, so I guess if TZF isn't that great I should just buy 1 SAD1024?

I wouldn't say TZF is not that great. I think it's more a matter of individual preferences, likes & dislikes. Some like it a lot. Some like it a little less. But, if you've never heard it or tried playing with it, you'll never know. But yeah, use something other than a SAD1024 for the second delay line.

If you are interested in adding TZF to the A/DA, I'm guessing that Markus is going to go after putting together a second delay line for the TZF mod. So, stay tuned... it can always be added later, after you get the stock circuit working. And BTW, check out the pics and docs for Markus' TZF Electric Mistress, if you havent already.
http://www.diystompboxes.com/smfforum/index.php?topic=36392.msg369947#msg369947
Pretty damn cool.

~ Charlie

[RaceDriver205]

OK, I think ill stick to the standard ADA Flanger. Once Ive built it, my ultraflange is getting harvested and going in the bin!

[Mark Hammer]

A "gedanken" experiment...

What would happen if you had two different-capacity BBDs - one longer, one shorter - assigned to the dry and wet channels, and clocked them both identically?

So, say you had an MN3009 (256 stages) and an MN3007 (1024 stages). The (ostensibly) dry signal goes through the 3009 and the wet goes through the 3007.  When the 3007 is down to 1ms delay, the 3009 is down to 250us, yielding what is, in effect, a 750us delay.  If you clock the 3007 down to 500us, the 3009 is down to 125us, yielding a 375us effective delay.  If the "dry" delay is a larger capacity chip (like a 512-stage MN3004), the effect is more robust (i.e., what gets "subtracted" is even greater).

Since the "dry" side is clocked at the exact same rate as the wet side, there is no risk of heterodyning - even at the longest delays, although I suspect one would need buffering of the clock signal if the one clock were driving the combined input capacitances of the two BBDs.

Just seems, on the surface anyways, like a simple way to get delay times that can approach zero more closely without taxing the BBDs themselves too much.  The impact on the pitch of the "dry" signal should, in principle, be negligible.

[StephenGiles]

....and if it were possible to unmodulate the clock signal driving the BBD in the dry signal path........ consider the normal clock output from pins 10 & 11 of the 4047 passed to the buffering 4049, AND to a new circuit which can remove the modulation - an "unmodulator" thus providing a fixed delay from any BBD it drives. I have no idea how to do this of course, and Ton could no doubt find a good reason not to!!

[puretube]

Mark:
for sync-clocked BBD1 (512) & BBD2 (1024), the following goes:
if BBD1= 1ms, & BBD2= 2ms, delta-T = 1ms;
if BBD1= 0.5ms, & BBD2= 1ms, delta-T = 0.5ms;
if BBD1= 0.25ms, & BBD2= 0.125ms, delta-T = 0.125ms;

the differential delaytime won`t get shorter than the shorter one...

the only factional difference* will be, that the change of delaytime (sweep)
can be achieved with a smaller control-voltage (or -current, or -resistor, or cap)
variation.


*to a single (short) delayline using 512 stages

[Mark Hammer]

True.  I guess the question I'm asking is whether the sorts of minimum delay times that get very very close to zero are obtainable with the kind of clocking we normally do and whether it would be possible to get even closer using a "delta T" approach.  I imagine that for some signals, dropping from 250usec to 100usec could be a noticeable change...or would it?

Perhaps the advantage would show up where the capacity-difference is made as small as possible.  So, the dry signal goes through an MN3009 and MN3004 (256 plus 512 stages, all driven by the same clock as the MN3007) while the wet side goes through 1024 stages.  Under those circumstances, the functional delay of the wet side is always 1/4 that of the dry side, so 500usec from an MN3007 = 125usec "delta T".

Of course all of that might be pissing in the wind.  The "psychoacoustic confusion" that results at the through zero point is strongly influenced by the amount of high-frequency content.  As I've remarked before, TZF is not that impressive if you were using it to process a single clean guitar or bass, but much MORE impressive when used to process a multi-source wide-bandwidth signal (e.g., a whole band).  The corollary is that, depending on what you want to process and what you are going to listen to it through, the very very short delays I'm trying to achieve here may be past the point of diminishing returns.  Maybe a Les Paul through a warm-sounding amp and 12" speakers stands to gain very very little from a flanger that sweeps down to less than 500usec or even 1msec because the locus of the added cancellations is simply well out of the range of the signal source and speakers.

[puretube]

BBD1 (  256) = 2.5ms
BBD2 (  512) = 5.0ms
BBD3 (1024) =  10ms

BBD1+BBD2 = 7.5ms;
BBD3-(BBD1+2) = 10ms-7.5ms = 2.5ms = delta-T
= time-difference = effective delay = 2.5ms;

BBD1 = 1ms
BBD2 = 2ms
BBD3 = 4ms

BBD1+BBD2 = 3ms;
BBD3- (BBD1+2) = 4ms-3ms = 1ms;

see: you can`t get shorter than the shortest one.

But: your maximum time-difference will decrease (reduced sweep-range).


There is an other possible solution, though,
if you`re willing to carefully watch your clox`
controlvoltages for 2 BBDs not driven by 1 clock. 

[puretube]

....and if it were possible to unmodulate the clock signal driving the BBD in the dry signal path........ consider the normal clock output from pins 10 & 11 of the 4047 passed to the buffering 4049, AND to a new circuit which can remove the modulation - an "unmodulator" thus providing a fixed delay from any BBD it drives. I have no idea how to do this of course, and Ton could no doubt find a good reason not to!!

You`re back into heterodyning territory, because above scenario is the same as 2 separate clocks.
And for TZF, you want the clock-frequencies like:
F1=F2+/-X

X=0 is the zeropoint = identical clocks

for the "sweet spot", you want X coming from larger positive values
approaching 0, becoming 0,
and then increasing to larger negative values,
and from there again "turning round,
approaching 0, go "through" it
and go larger into positive,
and return again...