Author |
Message |
egasimus
Joined: Feb 11, 2011 Posts: 113 Location: Bulgaria
|
Posted: Fri Jul 08, 2011 2:15 pm Post subject:
|
|
|
Has anyone tried using the 4050 in linear mode? |
|
Back to top
|
|
|
JovianPyx
Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
|
Posted: Fri Jul 08, 2011 2:26 pm Post subject:
|
|
|
The 4050 is a buffer, meaning it does not invert the input and as such it cannot be used in the linear mode the same way an inverter can.
However, if you supply a series input resistor and put a resistor from gate input to output, it might turn into a schmitt trigger. So there's at least that. Schmitt triggers are useful in certain kinds of oscillators.
So while you can't use it as a summing amp or a linear amplifier, you might use it as a schmitt trigger.
The CMOS buffer is really meant to be used in logic to alleviate fanout problems by beefing up the current drive capability of the input signal. _________________ 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
|
|
|
cthulu
Joined: Feb 07, 2009 Posts: 56 Location: Göteborg, Sweden
|
Posted: Fri Jul 08, 2011 3:06 pm Post subject:
|
|
|
JovianPyx wrote: |
Take a look at the 4007 IC. The internal schematic (remembering the caveat posted by mosc regarding these) would allow using 2 of the gates separately as weird VCAs. If you do this kind of thing with the 4049 (for example), you have only one audio input and 6 CV inputs.
|
I knew I'd seen that before...
from this place:
http://www5b.biglobe.ne.jp/~houshu/synth/
|
|
Back to top
|
|
|
JovianPyx
Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
|
Posted: Fri Jul 08, 2011 3:31 pm Post subject:
|
|
|
That man is an interesting designer/engineer. I've seen that schematic before, but never built it even though I'm a CMOS-ophile. That is a very interesting design indeed. His design seems aimed at emulating a sort of OTA in CMOS, which is why the warning is there. He also describes the gain cell as contributing to distortion. That can be good or bad. It probably works best with small signals.
SynthLore: There has been a long standing challenge on the Synth-DIY email forum about building a synth out of nothing but 4069UB. Mr. (Dr?) Houshu has what I think is the largest collection of circuit designs aimed directly at making music with CMOS. So if anyone is close to making an all 4069UB synth, it is he... But he also likes 4007, and that VCA design makes a lot of sense... I have to wonder if it could be used like an OTA to make a voltage controlled integrator (for a filter). If not, I know there are CMOS OTA schematics on his website. I don't think it's cheating to use things like 2N7000 either. _________________ 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
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Fri Jul 22, 2011 10:13 am Post subject:
|
|
|
Hey, i have another re-invention in Linear CMOS for you. I don't call anything an invention anymore, lol, cause there seems to be nothing new under the sun! But anyway this is a fun Linear CMOS thingie so here goes...
Take a two-input inverting gate and put a feedback resistor Rf1 from the output to input 1 and a shunt capacitor Cs from input 1 to ground. Then add a second feedback resistor Rf2 from the output to input 2 and an input resistor Ri in series with an input capacitor Ci from input 2 to a square wave or pulse input.
What's that do? Well it makes sustained oscillations in response to the input signal transitions. Specifically, it can sound in my limited testing of it like synthesized percussion or like a beeeooop alarm signal sound.
I dreamed up the circuit as a super oversimplified Lunetta form of a Karplus Strong loop, where the Rf1/Cs feedback represents the delay and filter of KS, while the Rf2/Ri represent the input and loop gain functions, plus Ci to shape the input edges into pulses. A KS loop may not be quite exactly what this beastie really does, but that's how i imagineered it.
ezekiel mentioned in the nearby Lunetta subforum thread "I propose a Challenge..." also experimenting with this concept and getting similar interesting results, so at ezekiel's suggestion I chose here as the place to discuss the circuit!
Les _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
ezekiel
Joined: Oct 17, 2009 Posts: 30 Location: Columbus Ohio
|
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Fri Jul 22, 2011 4:19 pm Post subject:
|
|
|
Thank you very much ezikeil, your schematic is worth a thousand words!
I may have forgotten this part, but I'd like to also mention that applying a pulse instead of a square wave could give really good results. Specifically, use a pulse that is a lot shorter in duration than the sustained oscillation intervals that you get from the particular values of R and C that you select and/or adjust. I say that because in the real analog Karplus Strong wirk and in simulations the short pulse approach produced the most percussive and string-like sounds. Good Luck!
Les _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Wed Aug 10, 2011 8:30 pm Post subject:
|
|
|
I have done a little bit of circuit analysis on two simple opamp-like linear CMOS circuits: the voltage follower and the non-inverting amplifier. The results were quite interesting. Here is the voltage follower analysis:
Draw a two-input inverting gate such as NAND or NOR. connect the output Vo to one input and apply an input voltage Vi to the other input. This looks just like a voltage follower but with a gate there instead of an opamp.
Now, for the equation we will model the gate's transfer function as -k times the sum of the inputs. This may not quite be a NAND or a NOR, in fact it might best be modeled as a piggyback of a NAND with a NOR, but anyway it's a good approximation I believe. Our equation is as follows:
Vo = -k * (Vi + Vo)
And a little math solves for the transfer function which is:
Vo/Vi = -k / (k + 1)
Then we can plug in larger and larger values for k and we see that as k approaches infinity, this transfer dunction approaches minus one. What this means is that the voltage follower configuration will follow the input just like an opamp would, except that it will be inverted.
Intuitively this makes sense from the perspective that the gate is kind of like an opamp that has two negative inputs.
I also solved for a non-inverting opamp configuration wired up as follows. Vo feeds back to one input with resistor Rf, another resistor Ri goes to ground from that input, and the input signal Vi gets applied to the remaining input. The solution?
Vo/Vi = -(Rf/Ri + 1)
This is the familiar expression for such a circuit in opamp form, however the transfer function is inverted.
So in general what this circuit analysis suggests is that we can use two-input gates in opamp circuit configurations and they will apparently give us the same behavior as the opamp, however with the transfer function inverted.
Please note that I have assumed that a good model for the gate is the sum of the inputs time a negative large gain, which may or may not be all that valid. It does make sense to me as a first approximation but the details may be a bit different. Anyway, just reporting my findings...
Les _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
synthesist
Joined: Feb 17, 2011 Posts: 79 Location: austria
Audio files: 2
|
Posted: Sun Aug 14, 2011 6:28 am Post subject:
|
|
|
hei!
I joined a little late to this thread, but could someone please explain me one thing:
What does "linear control" actually mean?
Is it just that you control an oscillator or filter by changing the amplitude or volume of the control voltage instead of switching it on and off?
Or does "linearity" mean something more like hertz per volt or volt per octave? |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Sun Aug 14, 2011 11:31 am Post subject:
|
|
|
synthesist wrote: | hei!
I joined a little late to this thread, but could someone please explain me one thing:
What does "linear control" actually mean?
Is it just that you control an oscillator or filter by changing the amplitude or volume of the control voltage instead of switching it on and off?
Or does "linearity" mean something more like hertz per volt or volt per octave? |
That's a really good question and I'm glad you asked it because we tend to take these basic things for granted and beginners do not know them. These circuits are linear in the sense that their behavior is not switching behavior. That is, the circuit behavior can be described with a set of linear equations.
For example, a resistor is linear in that if you graph it's current vs voltage relationship you get a line. Same for inductors and capacitors but there is a differential with respect to time involved, which is also linear.
Transistors, however, have an exponential relation between current and voltage which is non-linear.
So we are taking something non-linear, a switching transistor circuit, and adding circuitry that makes it opoerate in a linear mode. That is what makes the term linear CMOS so unique.
Les _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
synthesist
Joined: Feb 17, 2011 Posts: 79 Location: austria
Audio files: 2
|
Posted: Sun Aug 14, 2011 2:51 pm Post subject:
|
|
|
Is it like the difference between digital and anologue?
A linear curve has no steps, like a triangle wave.
Exponential is something like a stairway with just a few steps.
Did I understand it right?
One of the circuits I breadboarded was a VCO using the 4046 PLL.
I drove a LFO signal into the CV pin of the 4046 and added a capacitor to ground. So I had something like a glissando effect on the pitch that sounded pretty triangular. |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Mon Aug 15, 2011 2:04 am Post subject:
|
|
|
Yes and no, lol. In the sense of linear CMOS, specifically the CMOS part, it IS in fact the difference between analog and digital because we are using digital circuits to perform analog circuitry functions. However, that is not the distinction made in GENERAL when one uses the term linear vs non-linear.
In other words, there is a correspondence between linear being analog and nonlinear being digital, however that correspondence is approximate and not exact. It's kind of like saying black is darkness and white is brightness. While black is a darker color and white is a brighter color, darkness is not the same as black and brightness is not the same as white. In the same way, there is a correspondence between digital as nonlinear and analog as linear but it's not the whole story. You can't say "that is a nonlinear circuit therefore it is digital and this other circuit is linear therefore it is analog" like that.
To be precise, a linear circuit can be described with linear equations, including linear differential equations. Things that are ideally linear (and to confuse you, in reality almost nothing is linear and almost nothing is digital, but let's not get into that): resistors, inductors, capacitors, independent and dependent voltage and current sources, inductive and capacitive transformers, antannae of any passive variety and unless I'm forgetting something, that's about it. Things that are non-linear but may have linear modes: any semiconductor device, any tube device, Flash Gordon's laser weapons, etc. etc. etc. I think light bulbs and nixie tubes are non-linear, um, what else am I missing? Well, it's difficult to be complete and list everything, so just ask me about whatever I didn't mention.
By the way, linear and nonlinear electrical circuits are only part of the story. Other physical systems such as mechanical, optical, pneumatic, and I'm not sure what else are also either linear or non-linear depending on the same criteria, which is: if you can describe it with a linear system of equations, it is a linear system.
OK? Good!
Les _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Mon Aug 15, 2011 2:05 am Post subject:
|
|
|
p.s. Yes, the Death Star's planet destruction weapon is nonlinear! OK? OK!
Les _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Mon Aug 15, 2011 2:20 am Post subject:
|
|
|
In preparation for the Linear CMOS edition of The Les Hall Show, I decided to breadboard a test circuit and sweep some transfer functions for you. I used ChucK to generate an audio range sawtooth waveform for the input signal and my oscilloscope in XY mode to observe the results, which are shown in the three attached images that I photographed using my digital camera placed in close proximity to the screen.
Boy was I surprised! I found that almost any of these circuits will self-oscillate, even with compensation capacitors, unless you jack up the gain! This totally changes my understanding of the proper use of linear CMOS so it was a very valuable exercise for me personally. Anyway, here are the photos:
Les
Description: |
Vin and Vout connected to the input to show linearity of scope setup |
|
Filesize: |
263.36 KB |
Viewed: |
182 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
CD4011 with Gain Equal to 1.0, note self-oscillation |
|
Filesize: |
261.87 KB |
Viewed: |
173 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
Transfer function with Gain equal to 10, note linear region in the middle |
|
Filesize: |
260.65 KB |
Viewed: |
177 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
_________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
ezekiel
Joined: Oct 17, 2009 Posts: 30 Location: Columbus Ohio
|
Posted: Mon Aug 15, 2011 4:17 am Post subject:
|
|
|
The 2N7000 was mentioned a few postings back. Is that like a one-transistor version of the stuff in a CMOS 4007? |
|
Back to top
|
|
|
JovianPyx
Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
|
Posted: Fri Aug 19, 2011 2:29 pm Post subject:
|
|
|
ezekiel wrote: | The 2N7000 was mentioned a few postings back. Is that like a one-transistor version of the stuff in a CMOS 4007? |
2N7000 is an N channel MOSFET, so it's like the bottom transistor of a two transistor CMOS inverter. The top transistor is a P channel type. _________________ 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
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Fri Aug 19, 2011 9:51 pm Post subject:
|
|
|
Here is a scope photo of the transfer function of a certain dual input gate circuit that I created for use in linear CMOS boolean sequencer applications. I'm going to do a show on this circuit next week (on 8-26-2011), so tune in or catch the recording.
The circuit is simple and i'll draw it up for you later but it's got two 50k input resistors tied together to a 10k resistor to ground, a 680pF shunt capacitor on the input node for compensation, a 1uF cap to the input node itself, and a 100k feedback resistor. So it's wired up like a summing amp with a few modifications to make it work.
Now, about the transfer function, i set one input to -Vmax and swept the other input in a ramp from -Vmax to Vmax. Then increased in steps to get the plot. Really if my scope would do 3D plots I'd do one of those but it cannot do that easily, so it's a family of curves. You can see that there is a low linear slope, negative, of about -2, and as the steps on the outer loop increase the curve gets shifted to one side or the other (left i think). Also note that at the logic transition there is a jump.
This is EXACTLY the kind of stuff we want to make good music circuitry.It's linear with cutoffs at the limits and a zig-zag in the middle. Chaining these together in sequence and adjusting the resistor values a bit will transform a binary count into interesting drones and rhythmic patterns, i believe.
Anyway I'm going to give it a try and mess with the circuit next week on the show.
Description: |
Dual input gate transfer function |
|
Filesize: |
262.85 KB |
Viewed: |
177 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
_________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
Paradigm X
Joined: Feb 15, 2011 Posts: 363 Location: Null and void
Audio files: 2
|
Posted: Wed Aug 31, 2011 6:46 am Post subject:
|
|
|
Inventor wrote: | I built a pair of oscillators, one ten times slower than the other one, out of NAND gates. The each have a kill switch and the NAND gates have a feedback resistor so they will oscillate properly even with no Schmidt trigger on them. Got lucky there. They sound squareish and look it on the scope, though rather distorted looking. |
I appreciate this is well under the level of this thread but any chance you can post a schematic for this? I spent about 3 hours yesterday trying to get a nand osciallaotr to work and failed miserably, using a number of schematics and resources.
Sorry if OT. I have read this thread but a bit above my head. Many thanks
Ben |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
|
Back to top
|
|
|
Paradigm X
Joined: Feb 15, 2011 Posts: 363 Location: Null and void
Audio files: 2
|
Posted: Thu Sep 01, 2011 3:47 pm Post subject:
|
|
|
Cool man, thanks.
I spent about 3 hours with the datasheet and several other googled schematics but couldnt get anything to work at all, gets really frustrating. Nice to be able to discuss it with people !
Ill try it out and have a play; ive had loads of fun starting this cmos challenge. I learn so much easier with a known starting point than trying to start from scratch.
Thanks again.
Ben
|
|
Back to top
|
|
|
Paradigm X
Joined: Feb 15, 2011 Posts: 363 Location: Null and void
Audio files: 2
|
Posted: Fri Sep 02, 2011 1:42 pm Post subject:
|
|
|
Cool success it works!
Wicked.
However, i had to use pins two and 3 of the pot, and this worked as expected. Awesome, got another 4 oscillators to add to the mix !.
I prefered a 1meg pot for r1 as well, all good fun to experiment. Next into a 4040 or 4020...
My breadboards now full tho, with two circuits, bought the tiniest one imaginable stupidaly.
Cool, thanks for that, Ben. |
|
Back to top
|
|
|
tjookum
Joined: May 25, 2010 Posts: 360 Location: Netherlands
Audio files: 26
|
Posted: Fri Sep 02, 2011 2:22 pm Post subject:
|
|
|
Hi Paradigm X,
Really cool to see your progress on your first steps towards a lunetta. Maybe you can open a topic so we can follow your progress?
I did this when I first started and it really helped to get me motivated. There are some usefull tips in there to get started:
http://electro-music.com/forum/topic-42357-100.html
Just keep going, use and abuse that breadboard, it's great fun. _________________ There he goes. One of God's own prototypes. A high-powered mutant of some kind never even considered for mass production. Too weird to live, and too rare to die.
Hunter S. Thompson
movies
noise |
|
Back to top
|
|
|
Dougster
Joined: Sep 20, 2005 Posts: 272 Location: Tucson, AZ, USA
|
Posted: Fri Sep 02, 2011 2:35 pm Post subject:
|
|
|
Do you have an updated link for the Soundcloud files or are they gone? _________________ Once you start down the modular path, forever will it dominate your destiny!
Every DIY person should own a copy of Electronotes: http://electronotes.netfirms.com
Blue LEDs are evil. |
|
Back to top
|
|
|
Uncle Krunkus
Moderator
Joined: Jul 11, 2005 Posts: 4761 Location: Sydney, Australia
Audio files: 52
G2 patch files: 1
|
Posted: Fri Sep 02, 2011 2:47 pm Post subject:
|
|
|
Once again you are honing the cutting edge Les!
Great thread. _________________ What makes a space ours, is what we put there, and what we do there. |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Fri Sep 02, 2011 10:13 pm Post subject:
|
|
|
Gracias mi amigo, i try to do my part - you know, just playing around and all.
Les _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
|