Author |
Message |
Ricko
Joined: Dec 25, 2007 Posts: 251 Location: Sydney, Australia
Audio files: 27
|
Posted: Fri Jun 23, 2017 7:40 pm Post subject:
MinMaxMix circuit |
|
|
This circuit uses a simple arrangement of diodes, kinda like a diode ring but not, to make something that is kinda like a ring modulator but not.
It "mixes" two input signals. If both are positive, it gives the max of them. If both are negative it gives the min of them. If they are mixed, it sums them. Each input has a diode drip, which provides extra distortion.
Here is the schematic, simulated, and two audio examples, also from the simulation. The first has two triangle waves close: it gives a grungy chorus. The second has them separated by a non-harminic distance, and you can hear more of the inharmonic effects, but frequencies related to the inputs dominate over any sum/difference frequencies.
(The circuit can also be run from +/-15V with no change.)
ERRATUM: The circuit has 1N1148 but what is meant is the common 1N4148. Apologies. Actually, I think any small signal diodes will work.
Description: |
|
Filesize: |
322.3 KB |
Viewed: |
834 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Download |
Filename: |
tcurve.wav |
Filesize: |
86.18 KB |
Downloaded: |
944 Time(s) |
Description: |
|
Download |
Filename: |
tcurve2.wav |
Filesize: |
86.18 KB |
Downloaded: |
941 Time(s) |
Last edited by Ricko on Wed Jul 03, 2019 8:05 am; edited 2 times in total |
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
Posted: Wed Aug 23, 2017 10:28 am Post subject:
|
|
|
nice
this is exactly the kind of circuit i like Last edited by gabbagabi on Mon Apr 01, 2019 12:07 am; edited 1 time in total |
|
Back to top
|
|
|
elcoco
Joined: Feb 12, 2016 Posts: 8 Location: London
|
Posted: Wed Sep 13, 2017 3:38 pm Post subject:
|
|
|
Wow this looks great, and simple. Gonna have to give it a try and report back. |
|
Back to top
|
|
|
Ayab
Joined: Oct 27, 2015 Posts: 180 Location: London, UK
|
Posted: Mon Sep 25, 2017 1:58 am Post subject:
|
|
|
The samples sound nice and full. Will give building your circuit a try. Thanks very much. |
|
Back to top
|
|
|
Ricko
Joined: Dec 25, 2007 Posts: 251 Location: Sydney, Australia
Audio files: 27
|
|
Back to top
|
|
|
Ayab
Joined: Oct 27, 2015 Posts: 180 Location: London, UK
|
Posted: Wed Jul 03, 2019 7:25 am Post subject:
|
|
|
Was looking at the parts needed for this circuit "minmaxmix" over the weekend but not sure about the 1N1184 diodes:
Mouser has things that look a bit like a car key in an ignition or something and are quite expensive.
Am I missing something? |
|
Back to top
|
|
|
Ricko
Joined: Dec 25, 2007 Posts: 251 Location: Sydney, Australia
Audio files: 27
|
Posted: Wed Jul 03, 2019 8:02 am Post subject:
|
|
|
Oh, that is a typo (it comes up by default in the Tina-TI simulator and I don't catch it. Thanks for catching it. (The Velocitizer schematic had the same problem.)
No, it will work with any diodes. 1N4148 was what was meant. Apologies.
Rick |
|
Back to top
|
|
|
Ayab
Joined: Oct 27, 2015 Posts: 180 Location: London, UK
|
Posted: Wed Jul 03, 2019 10:14 am Post subject:
|
|
|
Ricko - Seriously, no need to apologise! Glad the diodes are standard and something I have available.
Thanks to you for making your very interesting work available |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
Posted: Sun Jul 07, 2019 6:55 am Post subject:
|
|
|
nice use of a full-wave bridge rectifier
I like the simplicity of the circuit makes me wonder how well a passive version would work, I guess I should just try it.
There will be some loss of course but it could be build very small. It also reminds me of a disrete XOR (which is a bit like
a ringmod) with a bridge rectifier that does use a transistor but doesn't need any external power. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
Posted: Sun Jul 07, 2019 7:32 am Post subject:
|
|
|
I did a quick test with a passive version. 2 signals connected to the AC inputs of a bridge rectifier and the DC ouputs of that connected to
the outside lugs of a 4K7 pot, the wiper of the pot is the output. The effect is most prominent when the pot is not in the center position,
so that might be a useful addition to the MinMaxMix circuit. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
|
Ricko
Joined: Dec 25, 2007 Posts: 251 Location: Sydney, Australia
Audio files: 27
|
Posted: Mon Jul 08, 2019 7:03 am Post subject:
|
|
|
Yes.
* Active solution: panel, three sockets, 4 diodes, opamp, 2 caps, perhaps 2 diode to protect power, perhaps two ferrite beads, a power socket and maybe four resistors, PCB to carry these
versus
* Passive solution: panel, three sockets, 4 diodes, maybe a 1M resistor to ground to make the diodes want to conduct...
And does the active version really do anything different than the passive version? I don't see it. It is not like the diode will behave much differenty depending the presense of buffers, will it? (Honest question)
The reason I do have buffers in my home version (which has three inputs not two) is to give the outputs a little boost. This is one circuit where a BAT43 or BAT85 or germanium diode is a good idea.
I have struggled to think what value to add to productize it: I think it is nice conceptually and can be useful for waveform mixing and EG mixing, but it is kinda too simple to sell! One idea was to package it with a crossfader: but it doesn't really fit in with the one I am looking at, which is a resurrection of Franco's "Seesaw" circuit from the SalMar Construction (a seminal thing that started off many modern circuits). Another idea was to couple it with some other CV fiddlers, but what? Lofic ors.
Maybe it is just so simple it could be an assembled product: I have a couple pasive modules (the mixer and LPGs from Siam excellent) and am a fan of the idea now (once all the main modules are buffered, in-betweenies don't need it in many cases.
Are you finding MinMaxMix more fun/useful for CVs or for Audio? |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
Posted: Mon Jul 08, 2019 8:04 am Post subject:
|
|
|
Iḿ not sure if the diodes would behave any different but there can be some loss with a passive resistor mixer.
For example if one signal is +5V and the other -3V then with an active mixer (with a gain of 1) the output would be +2V,
or actually -2V because it inverts (that's a good reason to keep the inverting input buffers).
With a passive mixer the output voltage would only be +1V (if equal value resistors are used). With an active version the
inputs are also isolated from eachother preventing any crosstalk in case the signals you are using for the inputs are
also used for something else with the help of a passive multiple/splitter.
For a passive version I was thinking it could be made as a small device without a panel, resembling something like this:
https://reverb.com/item/19416231-boredbrain-splix-inline-splitter-mixer-eurorack-modular-3-5mm-slime-green
But it could also be useful as an addition to another circuit which might not need an active version.
Can't say anything about CV or audio as I only did a quick test using 2 triangle/saw waves to see how it sounds and how it
differs from just mixing the signals together. It definitely sounds different. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
|
JovianPyx
Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
|
Posted: Mon Jul 08, 2019 11:48 am Post subject:
|
|
|
An additional problem with passive resistor mixers comes from the driving impedance together with the value of the mixer resistors. Let's assume equal size resistors. When the resistors are very high value, the driving impedance is mostly negligible because the mixer resistors add to the driver impedance, but as the value of the resistors is reduced, the driving impedance becomes more and more significant. Imagine what happens if we start with 100K and then try 10K and then 1K and then 100R etc. As we decrease the resistors, we perceive a higher loudness from the mixer (because the lower resistors pass more signal current), but you can see that what we're doing by lowering the value of the mixing resistors, we are coming close to connecting two driving outputs directly together. Intermodulation distortion increases as the resistive connection between the two driving devices begins to work like a multiplier. The lower the resistors' values, the more distortion there is.
An active mixer won't give such distortion which is one of several reasons why active is used for high fidelity mixers instead of passive resistor mixing. _________________ FPGA, dsPIC and Fatman Synth Stuff
Time flies like a banana. Fruit flies when you're having fun. BTW, Do these genes make my ass look fat? corruptio optimi pessima
|
|
Back to top
|
|
|
Ricko
Joined: Dec 25, 2007 Posts: 251 Location: Sydney, Australia
Audio files: 27
|
Posted: Tue Jul 09, 2019 12:33 am Post subject:
|
|
|
So in the case for MinMaxMix, where it is only diodes and a resistor mix we have three cases:
* If both inputs are below the level of a diode drop, the diodes block, and therefore it is like an open circuit: very high impedance as if unplugged.
* When both input signals are in the same quadrant, the higher level one "wins": the other is not forward biased so does not conduct. Therefore there is no tricky impact on impedance: just whatever resistors are place.
* When the signals are in opposite quadrants, both a positive-going and a negative-going diode conducts, and if they were connected together without a resistor this will be like connecting two outputs together (except that the outputs of the producers will have 1k or so resistors on the outputs usually). Therefore you always want the resistors for the positive and negative diodes' outputs before they are summed. And in this case, there is an effect on the impedance seen by each originating producer of these summing diodes.
Is that the takeaway? |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
Posted: Tue Jul 09, 2019 12:32 pm Post subject:
|
|
|
Yep, I think that is all correct.
There is something else I am wondering about though. With the active version the diodes are referenced to virtual GND (pin 2 of the opamp).
With the passive version they are referenced to eachother if not connected to anything, or probably a resistive path to GND of whatever you
connect it to. So I think that for the passive version it would be best to add 2 resistors to GND after the diodes before mixing. I was initially
thinking just one resistor after the passive mixing but that creates another voltage divider and might still act a bit wonky. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
|
Ricko
Joined: Dec 25, 2007 Posts: 251 Location: Sydney, Australia
Audio files: 27
|
Posted: Tue Jul 09, 2019 6:07 pm Post subject:
|
|
|
Yes, I think the passive version is safer with some resistor to ground (I mentioned a 1M resistor above for this) and two would be better (resistors are cheap) but I don't know that it would make much difference if there were two or one?
If your post-diode mixing resistors are 10k and the ground resistor is 1M, that just reduces a 5V signal to 4.95V: I think it is more important to have a diodeswith a really low forward voltage/diode drop (i.e. aBAT85 or germanium) rather than 1N4148s, which could give you .4V in that scenario.
Something like that. |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Wed Jul 10, 2019 2:48 pm Post subject:
|
|
|
Nice idea, tried it in my soft synth, works nice for control signals too _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
|
Back to top
|
|
|
Ricko
Joined: Dec 25, 2007 Posts: 251 Location: Sydney, Australia
Audio files: 27
|
Posted: Wed Jul 10, 2019 6:02 pm Post subject:
|
|
|
I don't mind the idea of the pot, actually.
I think a DC-blocking cap after it too, if used for waveshaping rather than CV mixing. That would open the door for cheap modulation by adding a CV to an input signal for different offsets. |
|
Back to top
|
|
|
|