Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
Posted: Wed Apr 25, 2012 4:47 am Post subject:
GDAL Synthesizer R1 Subject description: Holy fhtagn, I've built a synthesizer.
Insteads of doing things like sleeping or eating meals or showering or putting on clothes that aren't pajamas, I have built a synthesizer.
Thats not entirely true, I did do those some of those things but it feels like it took over my life for a few days and was much more important than the necessities of life.
GDAL Synthesizer R1
The R stands for ramshackle.
I have no schematics which would make much sense to anyone but me so I'll just post these videos and you can ask me whatever you want to know about it
Only one octave is wired up in this video and it was wired up wrong
Camera battery ran out at the end of this one.
_________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
Posted: Thu Apr 26, 2012 12:32 pm Post subject:
Thanks The box will be fine, I doubt I'll have any gigs with this, it is so unstable. Also I tend not to have gigs; I'm a musician, yes, but no one knows I exist _________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
Posted: Fri Apr 27, 2012 9:43 am Post subject:
Thanks chaps, yup, 36 microswitches to be precise.
Well I don't know about giving the specifics of the circuits, a lot are my own design so it would take all day, but I can give you an overveiw of whats in it. I had to leave out alot of possible and easy to add features because of time and cost constraints.
LFO: with rise and fall control, it could be both square wave with PWM and triangle but I didn't impliment that.
VCO: square wave with PWM which can be controlled manually and by LFO, so you can get, for example, a very thin pulse wave which changes between 5% and 15% duty cycle. It can generate a wonky saw wave/triangle wave too but I didn't impliment that.
VCF: Some kind of variation on the MFOS Weird Sound Generator filter, but I've changed it a bit. It has resonance, cutoff, LFO and envelope control.
Envelope: with attack and decay control, I think I know how to make a sustain for it too but I am not certain so I left that off.
Keyboard and Tuning: difficult. There is an offset control, because the transistor for the VCO only starts conducting at about 0.6 volts so you need to add on a bit (about 0.5 or 0.6) It is not 1V/oct, it is better. I made a resistor chain from trimmers, for the keyboard, which outputs the voltages you would expect to get out of an exponential converter. This eliminates the need for an expo converter and gets rid of the temperature dependancy of the VCO circuit.
Tuning is done with a non-inverting op-amp amplifier circuit and a non-inverting op-amp summer. Tuning it is perculiar you have to adjust the amplification to be near what you want, then adjust the offset so the voltage incriments change the VCO frequency over the right frequency range.
A problem arose with the keyboard though; if you hold down two keys you don't get the standard monophonic response because what you have done is short out a chunk of the resistor chain so you get some weird frequency about one third between the two notes you are playing.
I have a solution to this already though, I just haven't figured out the resistor values for it yet. It is simpler aswell. It's hard to explain but basically it still uses a voltage divider, except you short out bits of the chain to ground to change the voltage at the output. This will avoid the problems I observed before.
The hardest circuit to make was the only one which isn't apparent; the trigger pulse generator. _________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Joined: Mar 07, 2007 Posts: 4579 Location: London, UK
Audio files: 172
Posted: Sat Apr 28, 2012 4:46 pm Post subject:
Quote:
I just haven't figured out the resistor values for it yet.
it shouldn't really matter so long as they're all the same. _________________ Garret: It's so retro.
EGM: What does retro mean to you?
Parker: Like, old and outdated. Home,My Studio,and another view
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
Posted: Sat Apr 28, 2012 5:00 pm Post subject:
EdisonRex wrote:
Quote:
I just haven't figured out the resistor values for it yet.
it shouldn't really matter so long as they're all the same.
Seem like you may have overlooked alot of my post
The resistors in my chain are precisely tuned to give an exponential increase in voltage. It's not 1 Volt per octave, its hertz per volt or whatever it is that the one with an exponential voltage increase is called _________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Joined: Mar 07, 2007 Posts: 4579 Location: London, UK
Audio files: 172
Posted: Sat Apr 28, 2012 5:05 pm Post subject:
No, I missed the part where you do it the hardest way possible. _________________ Garret: It's so retro.
EGM: What does retro mean to you?
Parker: Like, old and outdated. Home,My Studio,and another view
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
Posted: Sat Apr 28, 2012 5:23 pm Post subject:
EdisonRex wrote:
No, I missed the part where you do it the hardest way possible.
What?!?
If you have an easier way I'd like to see it old chap.
This way there is no temperature dependancy, you see? no sourcing matched transistors for an expo converter or buying rare expensice ICs, less circuitry overall, thats not even half the advantages, the list goes on.
Also:
_________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
But it's your design mate. nice videos anyway. _________________ Garret: It's so retro.
EGM: What does retro mean to you?
Parker: Like, old and outdated. Home,My Studio,and another view
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
Posted: Sun Apr 29, 2012 3:52 am Post subject:
thanks for the diagram but I don't get their resistor chain, is that some variation on the R2R laddder? ... and are they all thermistors?
I see a few temperature compensated transistor pairs in there too, no thanks. I'd love to have a korg MS20/MS10 but I'd never use thier circuits in my synthesizers if they still use thermistors and transistor exponentiators. _________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Joined: Mar 07, 2007 Posts: 4579 Location: London, UK
Audio files: 172
Posted: Sun Apr 29, 2012 4:54 am Post subject:
No, they're not thermistors. It's just a resistor ladder. I think they used metal film resistors in the keyboard. And you don't *need* a prepackaged matched transistor pair. You can match your own. But you don't need to do it that way either. It's just a lot simpler in the long run because you only have to tune it in one place.
The actual circuit is the bit to the right of the keyboard ladder, and the bit below which is a multiplier.
The whole point is if you make the keyboard itself simple it's less likely to break. And you can do the linear to exponential conversion in one place. You don't need to use that korg exponential multiplier - any old exponential VCA will do. There's a few examples in the forums - and scattered around the net. They can be pretty simple to put together.
Just suggesting it because I tried what you're doing a LONG time ago, and that's why my dad shook his head and bought me a subscription to electronotes. _________________ Garret: It's so retro.
EGM: What does retro mean to you?
Parker: Like, old and outdated. Home,My Studio,and another view
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
Posted: Sun Apr 29, 2012 8:06 am Post subject:
Quote:
Just suggesting it because I tried what you're doing a LONG time ago, and that's why my dad shook his head and bought me a subscription to electronotes.
Ha!
I wish I'd known about Electronotes back in the day....
Joined: Mar 07, 2007 Posts: 4579 Location: London, UK
Audio files: 172
Posted: Sun Apr 29, 2012 1:20 pm Post subject:
Don't get me wrong - I love the sounds this thing is making. Ramshackle or not, it has character. (I LOL'd over the accidental dubstep)
I'm just saying it'd be easier to tune with a single exponential VCA driving your oscillator. _________________ Garret: It's so retro.
EGM: What does retro mean to you?
Parker: Like, old and outdated. Home,My Studio,and another view
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
Posted: Sun Apr 29, 2012 1:45 pm Post subject:
Look at my signature, I never claim to know everything, I probably don't fully grasp the concepts I am dealing with and I never said all my circuits were better than the MS20s circuits, I'd never dream of replacing them, you speak madness, MADNESS I TELL YOU!
I merely said my method of creating an exponential voltage is better because it is simpler and more stable than all the exponential converters I have seen.
I see many of you have qualms with what I say but I don't see any circuits or explanations or anything bwhich would sway my opinion at the moment.
How would one exponential VCA be easier to tune anyways? What I've got in my circuit is basically this:
Expo resistor chain --> Sample and hold --> DC amplifier (with variable gain from 0.[something incredibly low] to 2) --> summing amp 1
and
Offset --> summing amp 1
LFO --> summing amp 1
then
summing amp 1 --> VCO
I know how to calculate the voltages needed for my VCO and thus what gain to set the amplifier to and how much offset to add (It will be simply Y=mx+c, in this case Hz=m*V+Offset, then its just a case of finding out the incline of the VCO and adjusting the gain of the DC amplifier accordingly) but I currently don't own an oscilloscope accurate enough to test its frequency.
Just FYI I've got my synth in tune eventually, just by ear It would have been easier had I done some more sums. _________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
Posted: Fri Jun 08, 2012 7:36 am Post subject:
Circuits!
Diagrams!
Schematics!
Whatever you want to call them I've finally got some together for this
I advise one does not make the keyboard resistor chain in the method I did originally, I recently came up with a better way of doing it which is independent of the total chain resistance and just a little better in a few ways.
I'll get to that later perhaps...
The circuit diagram for the trigger generator is taken from a simulator I used to test it, it is a differentiator and then an inverter and combiner and it amplifies things and removes offsets which incur along the way.
The delay of the trigger pulse is the key to eliminating key up trigger pulses, the signal is delayed and ANDed (thats logic AND) with the gate, which is not delayed. This provides key up elimination.
Synthmonger/rene schmits VCO with a PWM comparator.
The envelope and LFO are basically the same circuit, however the envelope uses the reset of the 556 as the gate input and the trigger as the delayed trigger input. This initiates the decay of the envelope when a key is lifted but re-triggers if another is pressed while the first is held down. Basically this arrangement makes it work like you would expect an envelope to.
VCF is a variation on the MFOS WSG filter/twin T combined op amp filters.
keyboard cv_delay_sample and hold_gate_key up elimination.png
Description:
Filesize:
1.81 MB
Viewed:
192 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
envelope.png
Description:
The D input should read T, it is from the output of the key up elimination AND gate
Filesize:
1.76 MB
Viewed:
169 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
TRIGGER PULSE GEN.png
Description:
Filesize:
22.32 KB
Viewed:
185 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
vco lfo.png
Description:
Filesize:
1.79 MB
Viewed:
177 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
cv summing_vcf.png
Description:
Filesize:
1.79 MB
Viewed:
175 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
_________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Joined: Feb 03, 2013 Posts: 3 Location: Montreal, Canada
Posted: Sun Feb 03, 2013 12:34 pm Post subject:
Your GDAL Synthesizer R1 is really awesome and I would really like to hear it live so I'm trying to assemble it.
I have all the components and different circuits (from your drawings) assembled already.
The problem is how to connect them to each other? Could you tell me if I got it right?
RESISTOR CHAIN KEYBOARD -> GATE
RESISTOR CHAIN KEYBOARD -> SAMPLE AND HOLD
GATE -> KEY UP ELIMINATION
GATE -> ENVELOPE
KEY UP ELIMINATION -> ENVELOPE
KEY UP ELIMINATION -> SAMPLE AND HOLD
SAMPLE AND HOLD -> VCF
SAMPLE AND HOLD -> VCO
SAMPLE AND HOLD -> AMP1
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
Posted: Fri Feb 08, 2013 9:26 am Post subject:
Contact me via my tumblr which I use as an archive of my electronics
http://gda-labs.tumblr.com/
I'm still active in all those other places too, lurking here from time to time but not posting anymore.
In reply to you, Miros2424, I would acctually suggest using some quite different circuitry now, since I built it I've improved everything to do with this synth a lot. However I'm reluctant to post anything here anymore so you're stuck with these old obsolete designs for now.
If you build this you will have the only GDAL R1 synth in the world, mine is long gone
It's hard for me to remember everything I did but I think I missed a CV scaling amplifier out of my original schamtics, which is why some output and input markers may not match, I'm looking at SCV on one of the amp potentiometer inputs, which would stand for "Scaled Control Voltage". I'll whip one up which will work (see attachments). Oh, R1 might need to be 4K7 for improved range, anyways with that you should be able to get down to 10% of your CV and above 100%.
But that whole thing is a bit iffy, I can see a few ways to do it better and I know better ways to achieve tuning now.
Seems like I was saying in above posts, certainly one can do a lot of complex circuits with "tempcos" for extra accuracy and precision and efficiency but you can make a more round about circuit which is easier to comprehend and less efficient which gives you the same results. on a larger perspective of the world I'd allways try to be more efficient but just one tinkerer building a synth can cobble together any old shite and get it working, is it so bad if the method is non-standard? Things to think about, anyways, I digress and fear I may now become misunderstood.
Everything you have there looks correct. I'll try to fill in your question marks;
sample and hold output -> TRIGGER PULSE GENERATOR -> DELAY
VCF -> main output of whole synth! can be put through a potentiometer then high pass filter to remove Dc offset. e.g. 10k pot and a passive high pass filter made with a 10uF cap and 100k resistor.
AMP1 -> Question
AMP2 -> Question
amp 1 and amp 2 went to the VCF CV input and the VCO CV input however I don't remember which went to which, look at the diagrams I am sure it says. look at the output, I think one says FCV (which stands for Filter CV).
Cv scaling amplifier.png
Description:
Filesize:
6.95 KB
Viewed:
12500 Time(s)
_________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Joined: Feb 03, 2013 Posts: 3 Location: Montreal, Canada
Posted: Wed Feb 13, 2013 7:03 pm Post subject:
Quote:
sample and hold output -> TRIGGER PULSE GENERATOR -> DELAY
As per the schematics the T which stands for Trigger is an input to the Sample and Hold circuit so I'm still puzzled;
Resistor Chain Keyboard -> Sample&Hold -> Trigger Pulse Generator -> Delay -> Key Up Elimination -> T input on Sample&Hold
This above chain does not look right to me because the Sample&Hold circuit is in a loop. Please advise. Last edited by Miros2424 on Mon Feb 18, 2013 5:40 pm; edited 1 time in total
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You can download files in this forum