Something New from Doc Sketchy

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, luketeaford, lisa, Joe.

Post Reply
Thorsday
Common Wiggler
Posts: 241
Joined: Mon Dec 17, 2018 3:22 am

Re: Something New from Doc Sketchy

Post by Thorsday » Thu Aug 20, 2020 11:42 pm

Sweetness...
You do not have the required permissions to view the files attached to this post.

User avatar
Dr. Sketch-n-Etch
Super Deluxe Wiggler
Posts: 7890
Joined: Thu Jan 21, 2010 1:46 pm
Location: Vancouver, BC, Canada

Re: Something New from Doc Sketchy

Post by Dr. Sketch-n-Etch » Fri Aug 21, 2020 7:35 pm

Wow, it really looks great, Thorsday! And kudos to Jason for a great panel!

So, does it work?
They always said I would be nothin' but a fishhead -- look at me now!

Thorsday
Common Wiggler
Posts: 241
Joined: Mon Dec 17, 2018 3:22 am

Re: Something New from Doc Sketchy

Post by Thorsday » Fri Aug 21, 2020 10:22 pm

Dr. Sketch-n-Etch wrote:
Fri Aug 21, 2020 7:35 pm
Wow, it really looks great, Thorsday! And kudos to Jason for a great panel!

So, does it work?
Thank you for the gorgeous circuit and top build quality!

I have not played it yet.

Jason said it not only works, but everything about it is superb, and that he likes it better than his Encore Frequency Shifter.

It is currently in the care of USPS... I should receive before month's end.

User avatar
Moog$FooL$
Broke Dick Dog
Posts: 1721
Joined: Tue Jun 03, 2008 2:54 am
Location: Vancouver, BC.

Re: Something New from Doc Sketchy

Post by Moog$FooL$ » Sat Aug 22, 2020 2:41 am

:tu:
i use to drink & smoke.
that fat cat had to go..... wasn't even mine.

User avatar
Dr. Sketch-n-Etch
Super Deluxe Wiggler
Posts: 7890
Joined: Thu Jan 21, 2010 1:46 pm
Location: Vancouver, BC, Canada

Re: Something New from Doc Sketchy

Post by Dr. Sketch-n-Etch » Sat Aug 22, 2020 9:46 pm

Thorsday wrote:
Fri Aug 21, 2020 10:22 pm
Dr. Sketch-n-Etch wrote:
Fri Aug 21, 2020 7:35 pm
Wow, it really looks great, Thorsday! And kudos to Jason for a great panel!

So, does it work?
Thank you for the gorgeous circuit and top build quality!

I have not played it yet.

Jason said it not only works, but everything about it is superb, and that he likes it better than his Encore Frequency Shifter.

It is currently in the care of USPS... I should receive before month's end.
Wow, that really makes me feel good. Hopefully somebody else will order one soon. :party:
They always said I would be nothin' but a fishhead -- look at me now!

Dr Gris
Wiggling with Experience
Posts: 370
Joined: Tue Feb 14, 2017 5:14 pm
Location: Sweden

Re: Something New from Doc Sketchy

Post by Dr Gris » Wed Aug 26, 2020 2:13 pm

That MU panel looks great!!
Would be fun to compare it with my Encore FS, but I'm still hopelessly broke due to pandemic reasons :cry: :waah:

//M

User avatar
Dr. Sketch-n-Etch
Super Deluxe Wiggler
Posts: 7890
Joined: Thu Jan 21, 2010 1:46 pm
Location: Vancouver, BC, Canada

Re: Something New from Doc Sketchy

Post by Dr. Sketch-n-Etch » Tue Sep 15, 2020 12:46 pm

So, tonight I'm going to unveil my latest little project... a simple chromatic "analog" 7-bit quantizer (i.e., capable of generating 128 chromatic steps, so way more than are needed) which is based on a folding flash ADC of my own design which only requires one comparator per bit (rather than the one comparator per step that "sledgehammer" flash ADC requires). The voltage input is folded 6 times with Mancini rectifiers and gain-of-2 level shifters, and then the outputs of these 6 folders and the input are all compared to ground, and the output of the 7 comparators generates a Gray code which, when converted to Binary code, drives an R-2R ladder which generates the quantized output. No latches, no storage, no clocks. Real-time conversion of input voltage to output step.

I've just finished laying out the second prototype. I built the first prototype but wasn't happy with its performance, but the second one will be perfect.

Here's the layout (it might change slightly) -- it's pretty big -- 3.1" x 6.9" -- but really nothing but TL0XX opamps and two CD4070 quad XOR gates, so very cheap parts. Also, many of the resistors on the two TL084s (which are the comparators) may not be necessary -- I laid the circuit out so that I could apply hysteresis to the comparators, but it probably won't be necessary for opamp-based comparators. My first prototype used LM311s without hysteresis, and they chattered like crazy. I really don't need very fast comparators for this application, and quad opamps are cheap. I'll build it without hysteresis first, and if it chatters, then I'll add those 8 extra resistors.
QuantizerLayout.png
You do not have the required permissions to view the files attached to this post.
They always said I would be nothin' but a fishhead -- look at me now!

User avatar
devinw1
Super Deluxe Wiggler
Posts: 1575
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Something New from Doc Sketchy

Post by devinw1 » Tue Sep 15, 2020 1:18 pm

Sweet! Nice and simple and elegant!

Out of curiosity, did you have trouble with the LM311s on a protoboard or with the actual PCB? I've found you can easily get issues with them on a prototype breadboard because of trace lengths, capacitances, and grounding, but they will usually behave nicely on a PCB (if laid out sensibly).

User avatar
Rex Coil 7
Super Deluxe Wiggler
Posts: 7066
Joined: Thu Nov 29, 2012 11:29 am
Location: Captain Of Outer Space

Re: Something New from Doc Sketchy

Post by Rex Coil 7 » Tue Sep 15, 2020 1:53 pm

Moog$FooL$ wrote:
Fri Jun 26, 2020 7:27 pm
nice..... i like that one. u are really getting your shit together. finally.
how long did it take??
I hear ya ... he really needs to demonstrate some ambition once and a while.

Too busy making that callus on his ass grow, I reckon. :roll:






8-)
5U NORMALIZING PROJECT (for your entertainment) viewtopic.php?t=78836&highlight=

SCREW IT ... PULL THE PIN

User avatar
Rex Coil 7
Super Deluxe Wiggler
Posts: 7066
Joined: Thu Nov 29, 2012 11:29 am
Location: Captain Of Outer Space

Re: Something New from Doc Sketchy

Post by Rex Coil 7 » Tue Sep 15, 2020 2:18 pm

This whole COVIDZ thing has presented far too many opportunities for self reflection, and taking stock on what's important to me. That said, I really feel the need to pay The Doc for one of his constructs fabricated from nothing more than an idea, an image seen only by his mind's eye to ultimately taken all the way to final/finished and fully functional tangible production. Furthermore, I feel what he does with is time contributes and adds value to the modular synthesizer in the same way that talented/gifted musicians "add value" to the entirety of that art.

"What brand module is that?"

"Oh that? Pfft ... that's a Doc Sketchy, of course. He's Canadian, y'know! Like Rush 'n shit!"

I'm being silly there, but bad (very bad) jokes aside, I feel like adding one/some of his constructs would make my synth that much better for it. Better said, I believe my synth would be that much less complete without something from him. Genuinely sincere here. It would really add to the ~artisan factor~ of my synthesizer. ...thinking ... thinking ...

:)
5U NORMALIZING PROJECT (for your entertainment) viewtopic.php?t=78836&highlight=

SCREW IT ... PULL THE PIN

User avatar
Dr. Sketch-n-Etch
Super Deluxe Wiggler
Posts: 7890
Joined: Thu Jan 21, 2010 1:46 pm
Location: Vancouver, BC, Canada

Re: Something New from Doc Sketchy

Post by Dr. Sketch-n-Etch » Wed Sep 16, 2020 3:26 pm

Well, Rexy, while you're "thinking" I thought I'd report that the second prototype of my quantizer worked PERFECTLY -- the transitions are nice and steppy, and perfectly chromatic.

Here's a graphic of the final build, exactly as built (except for the fact that I ended up using TL074 for the comparators instead of TL084, and I left the 7 1M feedback resistors off of the comparators because I discovered that the opamp comparators were perfectly well behaved without hysteresis. To answer devinw1's question, yes, the LM311s were a right royal pain in the arse. They are simply too fast for this application, and they were firing about 50 times each time the triangle wave went through zero. Rather than add hysteresis resistors to all of them, I decided to just do away with them and use opamp comparators instead, which was the right call. I even managed to squeeze this completely new circuit onto exactly the same board size: 3.1" x 6.8", because I straightened up the R-2R ladder and routed the driving logic traces underneath the resistors.

You'll also notice that I added 100pF stability caps to each and every opamp output in the Mancini-rectifier-gain-of-2-level-shifting-buffer section. This really cleaned up all of those waveforms very nicely -- they are all nice and crisp now. Finally, it turns out (as suggested in H&H) that you can drive CD4000-series logic chips directly from opamp outputs (yes, even though they go to -13.5V and the datasheet says that you can't go more than half a volt below ground). All you have to do is put a current-limiting resistor in between the two, and the built-in protection diodes on the logic gates will hold the voltage to within half a volt of ground anyway, so long as you don't overstress the diodes with too much current (I used 10k resistors, so they're each receiving about 1.3mA -- this works completely fine, although I probably could have gone with 100k resistors).

I'm making a short video later today, after my online class (being a prof during the pandemic).
QuantizerBuild.png
You do not have the required permissions to view the files attached to this post.
They always said I would be nothin' but a fishhead -- look at me now!

KSS
Super Deluxe Wiggler
Posts: 3413
Joined: Mon Jan 25, 2016 7:28 am

Re: Something New from Doc Sketchy

Post by KSS » Wed Sep 16, 2020 6:42 pm

:tu:
Wonderful!

I worked on something semi-similar a long time ago but with nowhere near the elegance you have here. Same "can I do it all analog" motivation. Called it the double flash. One flash converter for volts, second for semitones, R2R output too.

This thread has to be one of the best on all of MW. Thanks Dr S-n-E.

Thorsday
Common Wiggler
Posts: 241
Joined: Mon Dec 17, 2018 3:22 am

Re: Something New from Doc Sketchy

Post by Thorsday » Wed Sep 16, 2020 7:11 pm

Hail Doc Sketchy

User avatar
Prunesquallor
Veteran Wiggler
Posts: 566
Joined: Wed Nov 11, 2009 3:08 am
Location: Valencia, Spain

Re: Something New from Doc Sketchy

Post by Prunesquallor » Thu Sep 17, 2020 7:01 am

KSS wrote:
Wed Sep 16, 2020 6:42 pm
This thread has to be one of the best on all of MW.
And is likely to end up as one of the most expensive. :mad: 8-) :party:
If at first you don't succeed read the instruction manual.

RIP Ziggy, the companion in my avatar.

KSS
Super Deluxe Wiggler
Posts: 3413
Joined: Mon Jan 25, 2016 7:28 am

Re: Something New from Doc Sketchy

Post by KSS » Thu Sep 17, 2020 7:27 am

Expensive? Or high value? ;)

Thorsday
Common Wiggler
Posts: 241
Joined: Mon Dec 17, 2018 3:22 am

Re: Something New from Doc Sketchy

Post by Thorsday » Thu Sep 17, 2020 7:30 am

KSS wrote:
Thu Sep 17, 2020 7:27 am
Expensive? Or high value? ;)
High Value.

User avatar
luchog
Crazy Pony
Posts: 1116
Joined: Sat Nov 21, 2015 2:57 pm
Location: The Emerald City

Re: Something New from Doc Sketchy

Post by luchog » Thu Sep 17, 2020 10:22 am

B0bcat wrote:
Fri Aug 14, 2020 3:01 pm
Euro DIY option would be fantastic
Seconded on that, although a finished module would be good too. And I'm not concerned about depth, since I'm not using shallow skiffs.

User avatar
Dr. Sketch-n-Etch
Super Deluxe Wiggler
Posts: 7890
Joined: Thu Jan 21, 2010 1:46 pm
Location: Vancouver, BC, Canada

Re: Something New from Doc Sketchy

Post by Dr. Sketch-n-Etch » Thu Sep 17, 2020 8:22 pm

Here's some more information about the Folding Quantizer. I'm preparing to make a video about it, and I put together some graphics.

The circuit is based on this basic circuit element:
Folder.png
This is what I call a "Mancini rectifier" after an article by Ron Mancini in the May 15, 2003 edition of EDN. In his circuit, the second opamp was a simple buffer, but I've refashioned it as a Gain-of-2 non-inverting amplifier which also level-shifts the output by +5.333V. Also, in the original Mancini rectifier, the output (and negative input) of the buffer is connected to the right side of the 10k feedback resistor on the rectifying opamp. However, I connect mine up in a more conventional fashion, and simulation suggests that there is absolutely no difference in circuit performance (of course, in a buffer, the negative and positive terminals share exactly the same voltage, so the right side of that resistor would not sense any difference -- there may be an impedance effect, though). Anyway, I can't hook up this circuit like Mancini does because it simply wouldn't work.

So, the Mancini rectifier is a full wave rectifier, also known as an "absolute value" circuit (although mine generates the negative of the absolute value, because the diode is flipped around). If you put a 10Vpp triangle wave in, you get a 5Vpp triangle at twice the frequency out which is only negative. Then, the Gain-of-2-level-shifting opamp restores this output to a 10Vpp centred triangle wave at twice the frequency of the input. Hence, each of these Mancini circuits simply doubles the frequency of an incoming triangle wave.

The third opamp, below the Mancini circuit, is a simple opamp comparator, which generates positive output when the triangle is positive, and vice versa. Also note that I have drawn these with hysteresis applied (at 1/1000 of the output) but in practice this is unnecessary for opamp-based comparators. Hence, I left out the 1M resistors. I still used the 1k resistors, but this was because I would have had to have installed wire jumpers otherwise, and so I figured that if I needed the hysteresis, this would save me having to desolder jumpers and install the 1k resistors.

Incidentally, I generated the 5.333V reference voltage by using an LM336Z5 5V reference, but I added a 10k trimpot in the manner suggested in the datasheet. This is supposed to give a range of voltages from 4 to 6V, and it worked perfectly. For the -5.333V reference, I simply inverted the +5.333V with an opamp inverter, but I put a 100-R trimpot at the junction between the two 10k resistors and the negative opamp input. This provides about a +/-50mV adjustment, which is enough to overcome any slop in the 1% resistors, and any voltage offset of the opamp.

Now let's look at the entire quantizer circuit, in which the schematic above is represented as boxes labelled "A B C":
BlockDiagram.png
The input voltage signal first comes into a simple buffer amplifier. This is because the Mancini circuit has quite a low (and variable) input impedance, so it will mangle any unbuffered input signal (I learned that the hard way). The buffered signal goes to the first comparator and into the first Mancini circuit. There are six Mancini circuits, but seven comparators. The comparators generate a Gray code based on the voltage of the input. Since the reference voltages are set at +/-5.333V, this circuit will generate 128 different binary codes within this voltage range. Each step will thus be 10.666V divided by 128, which is 83.33mV, which is 1/12V, or one semitone in a 1V/octave system. This entire circuit is known as a "folding ADC", which I learned about by reading the Wikipedia page about ADCs last weekend.

Of course, one cannot use Gray code directly in a DAC. Hence, one must convert the Gray code to normal Binary code, and that is what the cascading XOR gates are for. The output of the XOR gates is Binary code. Since these are CMOS gates, their outputs will be exactly 0V and 5.333V (the reference voltage, which also serves as VDD), which is perfect for driving the R-2R resistor ladder, which is the DAC. The non-inverting opamp at the end of the R-2R ladder is another Gain-of-2 level shifter, and this time it shifts the level down by 5.333V, so that the 0 to +5.333V range of the R-2R ladder output is expanded to -5.333V to +5.333V. This will give a total CV range of 10 octaves plus 8 semitones, which is well beyond the range of the modern piano (by nearly 50%). Actually, if the CV within the audio range, then the lowest and highest notes are significantly beyond the range that can be discerned as "in tune" or not. I wanted a 7-bit (128-step) quantizer to make it twice as useful as the MFOS circuit, which is 6-bit (64-step). Of course, with one additional Mancini circuit and XOR gate, this could be made into a 256-step quantizer, but there is no use for 21 octaves.
You do not have the required permissions to view the files attached to this post.
Last edited by Dr. Sketch-n-Etch on Sun Sep 20, 2020 1:48 pm, edited 1 time in total.
They always said I would be nothin' but a fishhead -- look at me now!

KSS
Super Deluxe Wiggler
Posts: 3413
Joined: Mon Jan 25, 2016 7:28 am

Re: Something New from Doc Sketchy

Post by KSS » Thu Sep 17, 2020 8:36 pm

Awesome! Thanks for sharing the details.


edit: looks like three 074's and a 4030 or 4070, three diodes and some R's would feed a 4 bit CTR for a 16 step VC sequencer?
Last edited by KSS on Thu Sep 17, 2020 8:49 pm, edited 1 time in total.

User avatar
Dcramer
Super Deluxe Wiggler
Posts: 5428
Joined: Fri Mar 29, 2013 12:11 pm
Location: Canada

Re: Something New from Doc Sketchy

Post by Dcramer » Thu Sep 17, 2020 8:37 pm

🙀

User avatar
Dr. Sketch-n-Etch
Super Deluxe Wiggler
Posts: 7890
Joined: Thu Jan 21, 2010 1:46 pm
Location: Vancouver, BC, Canada

Re: Something New from Doc Sketchy

Post by Dr. Sketch-n-Etch » Thu Sep 17, 2020 8:43 pm

The more astute among you may wish to point out that, had I reversed the direction of the diodes, I could have used +5.333V as the reference voltage to the level-shifters and saved myself having to invert the 5.333V reference, along with the 100R trimpot. However, this would have required the input to be inverted going into the first comparator, so it would not have saved an opamp (although it would have simplified the layout slightly). Also, all the comparator inputs would have to be swapped, which would mean that the positive inputs would be grounded, which would have required putting 1k resistors to ground to get the hysteresis (which I didn't need anyway). Hence, I may build it this way in the future, but other than the trimpot, it's really six to one, half a dozen to the other.

(Incidentally, I just noticed that the layout graphics above show a 1k trimpot on the reference inverter, but I actually installed 100R.)
They always said I would be nothin' but a fishhead -- look at me now!

User avatar
Dr. Sketch-n-Etch
Super Deluxe Wiggler
Posts: 7890
Joined: Thu Jan 21, 2010 1:46 pm
Location: Vancouver, BC, Canada

Re: Something New from Doc Sketchy

Post by Dr. Sketch-n-Etch » Sat Sep 19, 2020 5:15 pm

I made a video...



If you want me to make you one of these, then please send me a PM. I think $75 (US) is a fair price for the stuffed and tested board. It's 3" x 6.8" and can be powered by either +/-15V or +/-12V -- doesn't matter, as the onboard 5V reference drives everything.
They always said I would be nothin' but a fishhead -- look at me now!

User avatar
hox3d
Common Wiggler
Posts: 135
Joined: Fri Feb 02, 2018 6:30 am
Location: Grenoble, Alps

Re: Something New from Doc Sketchy

Post by hox3d » Sun Sep 20, 2020 10:41 am

Great video!
Very well explained, thanks!
Matte-fueled habits

User avatar
Dr. Sketch-n-Etch
Super Deluxe Wiggler
Posts: 7890
Joined: Thu Jan 21, 2010 1:46 pm
Location: Vancouver, BC, Canada

Re: Something New from Doc Sketchy

Post by Dr. Sketch-n-Etch » Sun Sep 20, 2020 1:49 pm

So, I messed up the schematic diagram of the Folder circuit in my third-to-last post (and in the video). However, I have now fixed it. Of course, the comparator takes the B output, not the A input. Sorry if that caused any confusion. The diagram there now is correct.

Also, it probably goes without saying, but if you would like me to make you a PCB for this quantizer, I will. Cost: $25 US

If you would like me to make you a stuffed and tested quantizer board, I will. Cost: $75 US

If you would like a quantizer module in 5U behind a 1U panel with an input jack, an attenuating pot, and four output jacks, then I will. Cost: $125 US.

I should have a complete module to show later today. Just gotta make a panel and a stooge bracket.
They always said I would be nothin' but a fishhead -- look at me now!

User avatar
Prunesquallor
Veteran Wiggler
Posts: 566
Joined: Wed Nov 11, 2009 3:08 am
Location: Valencia, Spain

Re: Something New from Doc Sketchy

Post by Prunesquallor » Mon Sep 21, 2020 4:38 am

Dr. Sketch-n-Etch wrote:
Sun Sep 20, 2020 1:49 pm
If you would like a quantizer module in 5U behind a 1U panel with an input jack, an attenuating pot, and four output jacks, then I will. Cost: $125 US.
Sounds great! Can I put on my 4U list?
If at first you don't succeed read the instruction manual.

RIP Ziggy, the companion in my avatar.

Post Reply

Return to “Music Tech DIY”