A/DA Flanger does TZF?

[oldschoolanalog]

Do yourself a HUGE favor, and use NP cap(s) for the LFO. 

[markusw]

Just populated my board. After checking the LFO and VCO I also added the SAD1024 and thanks to Charlie's fantastic board it fortunately worked immediately! 

Tommorow I will go into more detail tweaking the pots. Since my DMM's frequency counter works just up to a few 100kHz I thought of adding a 4018 divide-by-10 to check VCO freq more properly.

I can understand now why this thing is so famous. It's really damn close to TZF.

Thanks a lot once more to Charlie for all your work!!!

Markus

[oldschoolanalog]

Wow Markus! Congrats! That was fast! 2 questions. Did you use all the component values in Charlies circuit diagram, or did you change/tweak anything? And; what was your choice for the quad op amps? Thanks!

[sfr]

Looking at the FET on this build, anyone know which connection is the source and which is the drain?  Looking at the schemo, the gate connects to R33 and C17, which would make it the "top" pin on Charlie's boards (looking from the silk-screened side, with the pads for the pot at the top) but I'm not certain which one of the other connections is source and which is drain.  

I socketed this position, and threw in a 2N5485 after reading through this thread.  This would be oriented backwards from the silk outline on the board; but I just wanted to confirm the source/drain connections - source connects to the 1m resistor (R35) and Vb, and drain connects to R35 and R37?  

Mine is 98% populated, just waiting on an order from Mouser. . .

[markusw]

Wow Markus! Congrats! That was fast! 2 questions. Did you use all the component values in Charlies circuit diagram, or did you change/tweak anything? And; what was your choice for the quad op amps? Thanks!

I had two days off work more or less entirely dedicated to ADA building. Otherwise it would have taken me at least one week
Re opamps: initially I planned to use MC34074s for IC1 and IC2. Unfortunately, I didn't realise that I ordered SMDs (clever me ). So I used a TL074 for IC1 and a LM324 for IC2 and IC4. IC4 is MC1458. Since I have sockets on all ICs I probably will try swapping the quads for others. On the other hand I don't see any reason at the moment since the thing works and noise isn't really an issue.

The even odd switch works with the values in Charlie's schem. There is a volume drop in odd mode at short delays but I assume that's how the odd mode works (signals simply cancel each other).

For the LFO I went with rev4 using two 4,7µ tantal caps wired +//- -//+ (R60 is 10k). VCO is also more or less rev4.

Otherwise I also mainly went with rev4 values. Some parts changed because I didn't have the exact values but I assume that they are on "not-that-critical" positions:
* C19 470p instead of 510p
* C13 1µ instead of 0.22
* C29 47p plus 1M for R69
* C28 470n instead of 100n (should smooth out the triangle a tad more if I'm correct)
* R65 68k
* C4 and C22 47µ
* C34 not added
* C1 100n instead of 10n
* C17 2,2µ tantal instead of non-pol
* R40 47k
* R13 4k7
* JFET is J201 because I had it at home
* R16/R18 voltage divider was changed to 1070/15k (gives more or less exactly 14V)
* R28 1M2 instead of 1M3
* C12 10µ tantal instead of 1µ non-pol
* C32 22µ tant instead of 1µ (I like overkill )

Otherwise I went with the values in Charlies schem.

Looking at the FET on this build, anyone know which connection is the source and which is the drain?  Looking at the schemo, the gate connects to R33 and C17, which would make it the "top" pin on Charlie's boards (looking from the silk-screened side, with the pads for the pot at the top) but I'm not certain which one of the other connections is source and which is drain.  

I socketed this position, and threw in a 2N5485 after reading through this thread.  This would be oriented backwards from the silk outline on the board; but I just wanted to confirm the source/drain connections - source connects to the 1m resistor (R35) and Vb, and drain connects to R35 and R37?  

Mine is 98% populated, just waiting on an order from Mouser. . .

I came to the same conclusion: G is the top pin. So the J201 is orientated backwards from the silk outline.
My assumption was that S and D orientation shouldn't matter since the FET is acting as a switch. Please correct me me if I'm wrong. Anyway, with the J201 the threshold pot works as advertised.

Next will be fine tuning of the trims and then I'll fire up the stereo out stage I added to the prototyping area.

Markus

[markusw]

After hooking up the 4018 I found that the VCO works up to 8 MHz (4 MHz at the SAD). The SAD gets a bit warm though 
Now I settled for a range of about 17KHz to 2,5MHz.

The stereo out option is quite OK btw. The second signal is tapped off R42/D while the D jumper is removed and fed through a second output stage. You have to turn regeneration up a bit but then it gives some nice stereo experience. Don't know yet if it's useful in a band setting though. Nevertheless, I think I'll keep it since it fits nicely on the prototype area.

Markus

[markusw]

Was just exxperimenting with the stereo option. Once you found the sweet spot between the speakers it's really amazing in odd mode. 
Edit: with regeneration turned up a bit it's pretty cool even off axis. Definitely worth a try!

Nice toy...

Markus

[sfr]

...it turns out the usual trimpots from Smallbear I have in the parts drawer don't really want to fit this board, (now to trawl through the threads and find where Charlie posted a Mouser part # for trimpots . . . ) so I'll need to order some parts before I can finish this anyway.

These fit great... 858-72PR20KLF

~ Charlie

Hrm, I ordered these, they just came in and they don't fit.  Dug out the calipers and measured the board and the trimpots I have  - according to the measurements in the mouser catalog, those pots *should* fit; it appears the ones I got have the pins spaced like the 858-25PR series, (slightly closer together, like the ones I've gotten from small bear) not like the 858-72PR series should be spaced.  The packaging says I've got what I ordered.   All three different values I ordered are like this.

Guess I'll see what I can do.  I know Mouser is always really cool about fixing mistakes like these, but I really wanted to finish this today! 

[oldschoolanalog]

Hi Markus. Thanks for sharing all this R&D info with us. Your efforts will surely add a bit of sanity to this build! BTW, I agree. Overkill is good (within reason; whatever that is... ) Thanx again, and HAPPY NEW YEAR!!!

[sfr]

Well, mine works.  Trying to set the trims now.

Trying to follow the procedures in the A/DA calibration document at Mooseapotamus' site for setting T4 and T5.  I don't know if I'm measuring frequency wrong or don't understand my meter (Craftsman 82400, says it'll measure up to 4MHz) - I assume you stick the red lead at the test point, and the black one at ground?  Move the Range pot full counterclockwise?  Anything else I should be doing?  I get numbers, sometimes they make sense, but moving the trim pots seems to change the numbers arbitrarily.  And sometimes the number disappears.  They do seem to be more steady with the SAD chipped pulled.  So I set it in the center and it sounds okay.  I'll have to swing by my father's and use his stuff and see if I can get this to work.

I have access to a scope, but not a signal generator; any recommendation for setting T1 by ear or a simple way to build a signal generator?

[sfr]

just a mention - be careful with with the CD4047!  I wired a lead to the test point leading to pin 13 so I could clip my multimeter to to the wire rather than holding the probe against the pad - it accidently came into contact with the power pin on the chip and I released the blue smoke!  Fortunately I pulled the power pretty quick; the other CD chips got rather warm too, but the 4047 was the only thing I fried. Checked for any other shorts in case I was wrong about the cause, re-did a solder joint that looked suspect and popped a new chip in and it's working again.

[markusw]

Hi Markus. Thanks for sharing all this R&D info with us. Your efforts will surely add a bit of sanity to this build! BTW, I agree. Overkill is good (within reason; whatever that is... ) Thanx again, and HAPPY NEW YEAR!!!

I'm always glad if I can contribute a bit to this forum.
Happy new year!!

Well, mine works.  Trying to set the trims now.

Trying to follow the procedures in the A/DA calibration document at Mooseapotamus' site for setting T4 and T5.  I don't know if I'm measuring frequency wrong or don't understand my meter (Craftsman 82400, says it'll measure up to 4MHz) - I assume you stick the red lead at the test point, and the black one at ground? 

Measuring at the testpoint I got up to 8MHz which means 4MHz at the SAD. However, if I had it running for some time at this high clock frequ the SAD got a bit warm. The 4049 btw too.  Sticking the red lead to the test point and the black to gnd was also my assumption how to do it.

Move the Range pot full counterclockwise?  Anything else I should be doing?  I get numbers, sometimes they make sense, but moving the trim pots seems to change the numbers arbitrarily.  And sometimes the number disappears.  They do seem to be more steady with the SAD chipped pulled.  So I set it in the center and it sounds okay.

T4 and T5 are pretty interactive on my build. I also had disappearing numbers when measuring at the test point.  Measuring at pins10/11 of the 4047 or directly at the SAD I didn't have this issue. Seems like in comparison to pins 10 and 11 at pin13 of the 4047 you don't get a 50/50 duty cycle. My assumption was that my DMMs freq counter doesn't like this. Maybe you can confirm this with your scope? 
After I hooked up the 4018 div-by-10 on breadboard (which also should result in 50/50 duty cycle) numbers on the freq counter were stable when measured from the test point.

I have access to a scope, but not a signal generator; any recommendation for setting T1 by ear or a simple way to build a signal generator?

I also don't have a signal generator. You might want to have a look at the Visual Analyzer Software at http://hacca.altervista.org/.
It also has a signal generator with sine, square, white noise,.... works pretty fine in my hands.
BTW, it's also fun feeding white noise into a flanger, running the output back into Visual analyzer and then watching the comb filter moving up and down on the spectrum analyzer.

Markus

[moosapotamus]

Just populated my board. After checking the LFO and VCO I also added the SAD1024 and thanks to Charlie's fantastic board it fortunately worked immediately! 

Well, mine works.  Trying to set the trims now.

AWESOME!!!

I'm so glad to hear it's working! Way to go Guys!

Gotta go finish painting the bathroom now...

~ Charlie

[analog kid]

WELL WELL, It sounds like a couple other people beside me have got themselves an A/DA Flanger now huh? I would love to hear of your guys experience with some things like the extreme wacky multipitching wobble at higher speeds (with the RANGE set more than 1/2 or so ) and other things like how your Odd/Even is behaving!
 I have had mine finished for quite a long time now but have still went back and did some tweaks here and there cause I always think some things can be improved!
 
 BTW  , NO NEEd to measure in reference to ground with Frequency counter on your meter IME. just Red lead to test point.
 Markus, WOW that's some freq range you've set it to !!!  Mine goes plenty far "off the horizon" when set to anything over 1.2Mhz.     And also have you noticed that the "ugliness" factor of the detuning effect  ( when set at the mentioned high Range/ Faster speeds)  is smoothed out a great deal when setting the BoTTOm freq. a bit higher than the Factor spec'd  38khz?? 
 that is one of the most important things to decide on for me! When I do set it that low freq higher to like 80khz  I get a much nicer sounding  vibrato, leslie, etc..  not so multpitched like a scifi computer going haywire,    AND DOESN"T audibly effect that WIDTH of the sweep range really.
 
 anyway I'd love to hear about some of this stuff. and btw I DO wish I'd have just "played along" and got myself a  PRO board as well now!  Dang my pride 

[analog kid]

Markus
 You said you used the Rev 4 LFO cap values. I don't recall 4.7uf being used in any revision(though I don't have my notes in front of me or a good memory )    Did you mean 47uf  there?
 I seem to recall the diff in Rev 3/4 here being that 4 only used the ONE single cap there , of the same value as used in Rev3 I thought?
 I ask cause I still don't think i've settled on what's best for a good speed range.
  my values of 47uf and 22uf (which is very close to combined Stock values of two 33ufs)  seem to give a quite FAST at full settings.  i don't think the last little bit of rotation has any useable settings there! RAPIDO
  If you didn't use the 'stock' values , what has been your experience with the feel and range of the lfo rate control??

[sfr]

Hrm, the numbers are steadier using just one lead to measure frequency- but now they aren't quite making sense - the low end I can dial in about right, but the high end isn't going up to the MHz range - again though, it could just be that it *is* up there and my meter is reading it wrong (perhaps I've gone *too* high) next week my father and I will go play with the scope and his meter and see if we can't get a better measurement on there. 

I found a tone generator program for OS X - any problem with generating a 1KHz sine wave (I can also make Triangle, Square, Sawtooth and various colours of noise if one of those would work better) with this, measuring the output on the scope and attenuating the volume out of the computer/CD/iPod/whatever until it's 2V p-p (or whatever it says in the calibration instructions) and using that to follow the rest of the instructions for calibration?


But playing around with a unit that may or may not be 100% and who's trims aren't set at all right now -

The assymetrical sweeps I can get out of this things sound very cool.  The manual knob is very fun too - disable the sweep and get some very cool metallic lead sounds.  I bypassed jack 3, having jumpered the "S" and "Sw" pads for now, but having the external control for the manual knob is going to be a definite must here when I get to boxing this up.

The threshold knob is the coolest thing too - turn up the enhance to the point where it's out of control, and turn the threshold until it stops making noise.  Then you just get cool crazy laser squeals and other insane noise as you hit notes on the guitar, but be totally silent when you aren't playing.

I have some massively slow speeds on this right now - to the point where change is almost imperceptible; and then you realize it's at a totally different timbre than it was originally.   Fastest speeds are almost ring-mod-esque in that Trem-Lune style.  Last bit of the rotation seems very bunched up.  Might want to use a different taper for that 500K speed than "C".  But it could just be the settings on T3?  (Which I haven't really touched)

All in all, I'm liking it, and I have the feeling I'm *really* going to like it when I get it tweaked. 

[analog kid]

shouldn't really be any prob with using ANYTHING to generate tones as long as they're right and that you can control the dB level.
 I don't know why the Rate is bunched up on yours w/ a rev log?  are you sure that it's not just you're getting some Hyper fast speeds in that last little rotation which makes it seems as it it's 'bunched' or a major change in feel in that last bit?
   What is your meter's freq counter spec'd for ??  Some of them seem to only like to top at 1Mhz (which ISN'T going to help you completely with this beast) though it is very helpful. AND how are you going about the calibration?

 The right way to do it is to kill LFO , set the Manual to the "low" freq and then Trim T5 to get close to the spec'd bottom freq ( say 38k-50k or whatever)  THEN rotate the Manual completely , to the high side NOW adjusting High freq for as close to the top freq spec as possible USING T4
  You go give and take with this much like setting intontation on a stringed instrument is , if you know what I mean
 
T3 Really shouldn't give you much problem being set "wrong" , It pretty much is in the basically right area or it killes the sweep.   You use it AFTER you get the previous set and then set the Full LFO setting so that it's travelling the entire spectrum of the freq range you just preset manually.
 easy as pie. huh    By ear I find I'm almost dead on when I set the low frequency (T5 to taste since it doesn't affect a hugh freq range like t4 does)  and then in Even mode use T4 to , with Manual sweep set to high side again, achieve that  "small" transistor radio effect .  I tell you i'm almost on the money every time when I set it by ear that way.
 If your freq counter doesn't go hi enough. Just do as described but only set the T4 step Right up to what your meter will read (hopefully like 800khz or more)  and then just push it a tad bit more . THEN give the Lo freq adj T5 another small tweak.  Make sense?
 Also to check your LFO slow sweep, just set full CCW and time the cycle with a second hand (either listening for the sweep to complete a cycle or with freq counter to come back to bottom freq) and it should take something like 25seconds.  Yes almost impercievable
 My Polyphase has a "glacially" slow sweep  almost in comparison to the A/da

 I'd love to hear some more Observations,   I found comparing my unit's settings to sound samples of the fx settings in the manual , wherever I found those?

[sfr]

shouldn't really be any prob with using ANYTHING to generate tones as long as they're right and that you can control the dB level.

Okay, good to know.  I didn't think there'd be a problem.

I don't know why the Rate is bunched up on yours w/ a rev log?  are you sure that it's not just you're getting some Hyper fast speeds in that last little rotation which makes it seems as it it's 'bunched' or a major change in feel in that last bit?

Yeah, I'm not really sure at this point.  I haven't even put knobs on the pots, so it's hard to really judge the "feel" of the thing without something like that.  I think it may just be a side effect of having such "glacially slow" speeds at one end and obscenely fast speeds at the other.  The actual amount of alteration as I rotate the knob is probably the same, but it's harder to perceive the difference between two incredibly slow sweeps than between two very fast sweeps.  Even with the reverse log, I think having such a large range selectable at this point probably hampers things.  Perhaps adding the LFO speed switch would be a good thing to do here. 

   What is your meter's freq counter spec'd for ??  Some of them seem to only like to top at 1Mhz (which ISN'T going to help you completely with this beast) though it is very helpful. AND how are you going about the calibration?

Manual for my meter says up to 4MHz.  I've been measuring it with the red lead (since you mentioned that I don't need the black one) clipped to a wire coming off of the test point (TP on the board, connecting to pin thirteen of the 4047.)  Starting out with the "Range" knob fully CCW  (this kills the LFO, correct?)  I used a 50K pot for the range, like the the schematic at Moosapotamus.net, even though the board says 10K.  Set the "Manual" pot CCW and adjust T5 until I get around 34-35 KHz.  Numbers are still jumping around, although only over a two or three points.  This works fine.  Then I set Manual pot CW and adjust T4, shooting for 1.3 MHz, but the numbers I get top out at about 60-80KHz.  But they jump around a lot too, sometimes saying only 7 or 8 KHz; occasionally jumping very high, at one point they showed just Hz not KHz or MHz.  I haven't even really been able to figure out which way turning the trim pot makes the numbers move.

Considering that the numbers are moving, perhaps I should just pull IC4 to make sure the LFO is killed? 

[markusw]

I would love to hear of your guys experience with some things like the extreme wacky multipitching wobble at higher speeds (with the RANGE set more than 1/2 or so ) and other things like how your Odd/Even is behaving!

I thought that at higher LFO rates (just below audio freq) it's normal to get some vibrato/detuning sounds.
Re odd/even: at lower clock freqs there is no change in volume, at high clock freqs there is a volume drop in odd mode. Suppose, this is normal since at thigh clock freqs the two signals cancel each other.   I hear the same with the Hoax, btw. At TZF there is complete cancellation.

And also have you noticed that the "ugliness" factor of the detuning effect  ( when set at the mentioned high Range/ Faster speeds)  is smoothed out a great deal when setting the BoTTOm freq. a bit higher than the Factor spec'd  38khz??

Thanks for the observation! If you turn back range a bit and set manual to higher clock ranges shouldn't this give about the same effect? I'm not 100% sure how the range and manual pots interact so I'd be glad about explanations 

Markus, WOW that's some freq range you've set it to !!!  Mine goes plenty far "off the horizon" when set to anything over 1.2Mhz.   

I definitely prefer it when the clock goes up to ~2,5MHz, it's closer to TZF 
What do you mean with "off horizon"?

You said you used the Rev 4 LFO cap values. I don't recall 4.7uf being used in any revision(though I don't have my notes in front of me or a good memory )    Did you mean 47uf  there?
 I seem to recall the diff in Rev 3/4 here being that 4 only used the ONE single cap there , of the same value as used in Rev3 I thought?
 I ask cause I still don't think i've settled on what's best for a good speed range.
  my values of 47uf and 22uf (which is very close to combined Stock values of two 33ufs)  seem to give a quite FAST at full settings.  i don't think the last little bit of rotation has any useable settings there! RAPIDO
  If you didn't use the 'stock' values , what has been your experience with the feel and range of the lfo rate control??

You're right Rev4 calls for a 2,2µ. I didn't have a np 2,2µ so I went with two 4,7µ caps in series, should be close enough. My freq counter doesn't really like low freqs but at fastest rate the LFO is at 12Hz which is pretty close to "specs" (10Hz). On the low end it takes 25-30 sec to complete a cycle.  I didn't check the rate pot response that carefully but I too have the feeling that it's somehow bunched up at the top end. Maybe a slightly different taper (30%) would give a smoother feeling?
On the other hand maybe sfr is right in that:

I think it may just be a side effect of having such "glacially slow" speeds at one end and obscenely fast speeds at the other.  The actual amount of alteration as I rotate the knob is probably the same, but it's harder to perceive the difference between two incredibly slow sweeps than between two very fast sweeps.

Considering that the numbers are moving, perhaps I should just pull IC4 to make sure the LFO is killed?

Did you try to measure clock at pin10 or 11 of the 4047. With my DMM I have massive problems getting reasonable readings at the test point. Sometimes I don't get any readings at all. At pins 10 and 11 you should have 50% duty which your freq counter also might prefer. At least with my DMM it gives very stable readings. Just my 2c 

Considering that the numbers are moving, perhaps I should just pull IC4 to make sure the LFO is killed?

Just tested freq at pin10 with the range at 0 and it didn't change at all regardless of the LFO rate.

But they jump around a lot too, sometimes saying only 7 or 8 KHz; occasionally jumping very high, at one point they showed just Hz not KHz or MHz.

Again, this might have to do with the duty cycle which isn't 50% at pin13.

Markus

[markusw]

BTW, in case some one is interested in a little modding    here's a "layout" I did for the stereo option:



V+ is tapped off the 7815's out, Vb is taken from R35's top pin. Basically there is no difference if you take the audio signal off pad C and remove the jumper at pads D or the other way round.

Have fun,

Markus