TZF3007 - TZF (short delay) module, MN3007 based

Started by Thomeeque, July 24, 2010, 02:24:30 PM

Previous topic - Next topic

Thomeeque

 Hi!

In 9V Electric Mistress retrofit with MN3007 thread I have promissed to Scruffie (and Dave) that I'll try to design short-delay module, lighter compared to the one designed by Markus W..

It took me ages but here it is finally:



TZF3007_v1.0_layout_1200DPI.png
TZF3007_v1.0_PCB_1200DPI.png
TZF3007_v1.0_PCB_1200DPI_mirror.png
TZF3007_v1.0_schematic_1200DPI.png
TZF3007_v1.0_partlist.txt

Notes:

  • I assume that desired delay time would be somewhere between 1-10ms (~ fs = 100kHz-1MHz for 1024-cell BBD)
  • fs range IMO allows to drop LPFs around delay-line
  • Clock signal buffers are required at this range on the other hand (MN3007 is "overclocked" here)
  • CD4047 is used above it's range here as well, so every given chip may behave differently - frequencies on the schemo are values I have got when I have build CD4047 clock experimentally (I have tried only one chip, CD4047BE by Texas Instruments, at VCC=10V).  You may experiment with C5, R6 and POT_DELAY values to get your range.
  • My clock ticked even when R6+POT_DELAY were set to "zero" (I've got these frequencies: 2MHz for C5=100pF, 2.3MHz for C5=68pF, 2.6MHz for C5=47pF, ICC was around 20mA)
  • All these frequencies were measured at +Q and -Q outs (~ sampling frequency)
  • I don't believe to my oscilloscope at these frequencies, so I don't know actually how nice are signals at +Q and -Q outs
  • Circuit won't have exact unity gain - it will be slightly sampling-frequency (~ delay time) dependant (see "Gi - fcp" graph in MN3007 datasheet) - but same will apply for wet path of the flanger, so maybe it's a good thing for the proper TZF effect?
  • Output expects decoupling capacitor at TZF return input
  • Don't forget to solder jumpers under MN3007 socket first ;)
  • If you don't plan to solder POT_DELAY directly to the PCB, you may cut PCB narrower and use inner pads instead.
  • Besides the clock snippet I did not build it, so it's not verified at this moment

 Please, feel free to comment.

Cheers, Tomas
Do you have a technical question? Please don't send private messages, use the FORUM!

Scruffie

That is awesome Thomeeque, cheers for taking the time on this one  :)... now I need to start collecting some parts to breadboard this...

Thomeeque

#2
Quote from: Scruffie on July 24, 2010, 02:43:22 PM
That is awesome Thomeeque, cheers for taking the time on this one  :)... now I need to start collecting some parts to breadboard this...

I have just updated initial post with link to partlist :) Good luck!

Note: for first shots you can use:
- for C1 and C4 anything from let's say 4u7 to 47u
- for C2 100n~1u
- RT1 10-100k
- POT_DELAY 25-100k
- C3, C6 - N/C
- R6, R7 - jumper
Do you have a technical question? Please don't send private messages, use the FORUM!

oldschoolanalog

#3
Thanks Tomas!
Who will be the first one to try this?  ;D ;)

I should say who else...

edit: Just a thought. Might want to put a bit of LPF on the output side of this. The LFO sweeps the delay through a wide range of the f's one would set the TZF board to. Heterodyning gives you the sum (shouldn't be a problem) and the difference of the 2 clocks when they "cross paths". What will happen when (not if) the TZF board is set at, let's say, 1MHz and the LFO sweeps the delay through the ~995KHz range?
1MHz - 995kHz = 5kHz = audible range = heterodyning (?)
Or maybe not.
Perhaps someone w/more TZF experience could chime in here?
Stephen? Mark?
Thanks!
Mystery lounge. No tables, chairs or waiters here. In fact, we're all quite alone.

StephenGiles

Puretube had something to say about prevention of Heterodyning - you first have a "nice hot cup of tea" and then do a very very careful layout :icon_biggrin:. His post is on here somewhere, but I should think keep the clock lines well separated by ground strips - perhaps even separate boards.
"I want my meat burned, like St Joan. Bring me pickles and vicious mustards to pierce the tongue like Cardigan's Lancers.".

12Bass

It is far better to grasp the universe as it really is than to persist in delusion, however satisfying and reassuring. - Carl Sagan

Thomeeque

Quote from: oldschoolanalog on July 24, 2010, 03:22:41 PM
Just a thought. Might want to put a bit of LPF on the output side of this. The LFO sweeps the delay through a wide range of the f's one would set the TZF board to. Heterodyning gives you the sum (shouldn't be a problem) and the difference of the 2 clocks when they "cross paths". What will happen when (not if) the TZF board is set at, let's say, 1MHz and the LFO sweeps the delay through the ~995KHz range?
1MHz - 995kHz = 5kHz = audible range = heterodyning (?)
Or maybe not.

Oh, probably yes :( It just did not cross my mind. I should have read Markus's thread properly first, it's discussed there, even puretube's terrifying post is linked there.

Unfortunately it makes perfect sense so we probably should not cheat anymore and put proper LPFs where they belong. Especially when the wet path is cheating already..

I'll try to do it ASAP.

Anyway, you can still try to build it as it is now (it won't be so cute with two additional LPFs anymore) - it still may work by some miracle and in the worst case you will end up with nice little MN3007 tester ;)

T.
Do you have a technical question? Please don't send private messages, use the FORUM!

oldschoolanalog

Quote from: Thomeeque on July 26, 2010, 05:20:18 AM
...even puretube's terrifying post is linked there.
Thanks for finding that. Not so much terrifying, but cautionary and enlightening. Good advice from somebody who has done "a bit" of designing in his time. :icon_biggrin:
Quote...proper LPFs where they belong.
It seems proper placement of the LPF's allows for much "lighter" filtering. Take care of the potential problem at the potential source of it. Good stuff.
QuoteEspecially when the wet path is cheating already..
Not so much cheating. It just "is what it is". (cheating? :icon_lol:)
QuoteI'll try to do it ASAP. Anyway, you can still try to build it as it is now (it won't be so cute with two additional LPFs anymore) - it still may work by some miracle...
I have no doubt it will work. I'll wait until all the elements are on one board. With a ground plane and separate rails for the clock, BBD & audio (LPF's); not to mention keeping components close, yet properly separated; a well designed TZF board is no small undertaking. Take your time on this one. "The people appreciate your efforts."
Quote...and in the worst case you will end up with nice little MN3007 tester ;)
Already have one of those. Its called a DOD FX75B flanger. Its tested numerous MN3006/7/9's in its time and also does a really nice chorus. It's well worth the $20US paid.
Thanks again.
All the Best!
Mystery lounge. No tables, chairs or waiters here. In fact, we're all quite alone.

12Bass

What would really be great is an A/DA or Mistress complete with de-bugged TZF on the same PCB.  Of course, that might be asking for trouble.... but perhaps possible with careful layout.

I might build up one of the TZF circuits and try it out with my A/DA clone.  It should be fine at 15V, correct?
It is far better to grasp the universe as it really is than to persist in delusion, however satisfying and reassuring. - Carl Sagan

Thomeeque

#9
OK, v1.1:



TZF3007_v1.1_Schematic_1200DPI.png
TZF3007_v1.1_PCB_1200DPI.png
TZF3007_v1.1_PCB_mirror_1200DPI.png
TZF3007_v1.1_Layout-Parts_1200DPI.png
TZF3007_v1.1_Layout-Values_1200DPI.png
TZF3007_v1.1_Layout-Jumpers.png
TZF3007_v1.1_Partlist.txt



TZF3007_v1.1.asc

Due to added LPFs there will be small insertion loss (about -1.3dB ~ 0.86), you may ignore it or try to compensate it externally (e.g. by lower value of R25=13k resistor in RETURN path in case of EM9V /difference value resistor should be then used for TZF bypass/).

Feel free to ask and/or comment, but don't expect big effort from my side anytime soon, I need a break :icon_eek:

:icon_mrgreen: T.

Btw. you may try to start with lighter filtering by not installing C4 and C7 (for "calmer" phase mainly).
Do you have a technical question? Please don't send private messages, use the FORUM!

oldschoolanalog

Wow, that was fast!
Quick Q. Should there be a cap between the BBD output and the post BBD LPF?
Thanks for the awesome effort. :icon_cool:
Mystery lounge. No tables, chairs or waiters here. In fact, we're all quite alone.

Thomeeque

Quote from: oldschoolanalog on July 28, 2010, 08:18:40 PM
Wow, that was fast!
Quick Q. Should there be a cap between the BBD output and the post BBD LPF?

Thanks Dave, I believe it's OK as it is. Q1 is this way DC biased by BBD output (Q1 actually still forms basically the same emitter-follower buffer as in v1.0 or as in my retrofit, it's just now "surrounded" by filtering network :))

Quote from: oldschoolanalog on July 28, 2010, 08:18:40 PM
Thanks for the awesome effort. :icon_cool:

You're welcome :)

T.
Do you have a technical question? Please don't send private messages, use the FORUM!

Thomeeque

Quote from: 12Bass on July 26, 2010, 12:36:58 PM
What would really be great is an A/DA or Mistress complete with de-bugged TZF on the same PCB.  Of course, that might be asking for trouble.... but perhaps possible with careful layout.

I might build up one of the TZF circuits and try it out with my A/DA clone.  It should be fine at 15V, correct?

Yes, go ahead. Just note that you need to decouple TZF3007 output by capacitor (cca 220nF) for A/DA. Insertion loss may be simply compensated by adjusting R41 value (maybe use 100k trimmer temporarily there to find the right value).

Good luck, T.
Do you have a technical question? Please don't send private messages, use the FORUM!

Thomeeque

Do you have a technical question? Please don't send private messages, use the FORUM!

Mark Hammer

#14
Thanks so much.  Speaking as a public servant, I would describe this as a true public service!  As a standalone module there are all sorts of other uses, too.  For example, a person could use this to stagger the stereo outputs of a sum/difference device, whether chorus, phaser, or flanger, so that the two signals do not cancel.  Also something you can use to adjust ear arrival times of PA speakers (within limits, of course).

Thomeeque

#15
Quote from: Mark Hammer on July 30, 2010, 08:51:59 AM
Thanks so much.  Speaking as a public servant, I would describe this as a true public service!

Should I arrange pay-pal account for donations? :icon_mrgreen:

Thanks for kind words, Mark! :) T.

..and for extension of your post (about other possibilities), interesting!
Do you have a technical question? Please don't send private messages, use the FORUM!

Mark Hammer

You're most welcome.

Think about how many producers and engineers will talk about  how they use multiple amps and cabs to "thicken up" the sound of a recorded guitar.  Now, a big part of that will be because all those amps and cabs have different resonances that show up at different points, post-strum.  But it will also be the case that the sound produced by all those amps, mic'd up separately, will not be flawlessly or perfectly aligned in time.  The difference may be a fraction of a millisecond, or two, and, HEY wait a second!

Yup, this makes an excellent addition to a splitter so that two amps, or even two distortion pedals, can be fed with signals that are staggered juuuuuuuust ever so slightly, to add some richness.  You won't hear it as chorus or flanging, because nothing is being swept, and you certainly won't hear it as delay, because it is way too short.  But you just might hear it as a tonal thickening or simply difference in character, arising from the re-alignment or misalignment of harmonics.

oldschoolanalog

Quote from: Mark Hammer on July 30, 2010, 08:51:59 AM
...I would describe this as a true public service!
Yes Tomas, you truly have earned your Halo this time! ;D
Quote from: Thomeeque on July 30, 2010, 09:04:18 AM
Should I arrange pay-pal account for donations? :icon_mrgreen:
Uh, lets not get too crazy here.  :icon_lol:
I still owe you for the switches, come to think of it.

BTW; this circuit also can also be used as a nice & ez comb filter. Perhaps add a regen loop for that use. Put the delay pot on a pedal. Whatever...
Mystery lounge. No tables, chairs or waiters here. In fact, we're all quite alone.

12Bass

Quote from: oldschoolanalog on July 30, 2010, 11:57:08 AMBTW; this circuit also can also be used as a nice & ez comb filter. Perhaps add a regen loop for that use. Put the delay pot on a pedal. Whatever...

Or build two, and have TZF controlled by a pedal....   :o
It is far better to grasp the universe as it really is than to persist in delusion, however satisfying and reassuring. - Carl Sagan

Thomeeque


Oh man, how could world ever exist without my modules? :icon_eek:

Please, don't answer :icon_mrgreen: T.
Do you have a technical question? Please don't send private messages, use the FORUM!