***SORRY the synth is so quiet!!! I can upload another demo with better audio if you want, but you should be able to hear what's going on.
I MADE A THING!!!!
(also, this is my first schematic with EasyEDA, which I started using just 2 days ago!)
Whilst the core circuit is pretty rad without refinements, (especially for Lunetta applications) I know I need some op-amp buffers and other stuff to make it more modular friendly. HELL! I might make a PCB, cause this is a pretty rad idea!
WHAT DO I NEED???
-something that's interesting, but probably due to lack of a buffer is that when I remove the pull down resistor on the VCO input, there is a considerable amount of slew, (slide on the CV) but I'm almost certain that it's one of the large caps I stuck across on the power lines.
-Do I need an expo/volt-per-octave converter for the input?
...not super accurate, but log-y enough...
-Decoupling caps?!?! where????
-I definitely need an op amp buffer on the VCO (4046 input/4094 clock speed)
-should the buffering op amp be bi-polar powered? what would the advantages be?
I obviously need to ground ALL the extra inputs.....but I'm not so sure about the 4094 and the 4046... what exactly are the inputs?
I got the idea from https://acidbourbon.wordpress.com/2015/02/07/clocked-8-bit-random-pattern-generator-for-cmos-synth/, (I found the post on Hack-a-day), which I adapted for more common chips/ones I had on hand, but I have also changed it quite a bit.... Later I found a circuit that the incomparable, immaculate, God-like Thomas Henry had posted, But it of course was way too complicated for me, and used some sort of $20 IC that there are only 4 of in the world. JUST KIDDING!!!!!
but yeah.
what do I need to add?
what did I miss? [/img]
Schematic_8bitnoise_Sheet-1_20181210204111.png
Description:
Filesize:
78.45 KB
Viewed:
502 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
Posted: Tue Dec 11, 2018 10:26 am Post subject:
Re: CV 8-BIT NOISE/ Random waveform (needs refining) Subject description: suggestions? Buffers? exp/v-per-octave? HELP
WHAT DO I NEED???
>>probably due to lack of a buffer
I would buffer the output for sure, and if there is a DC offset at output (there is right?), get rid of that as well. I have a simple TL072 board on my website that would do the trick. You could perf the schematic on my site--it's a really simple design.
-Do I need an expo/volt-per-octave converter for the input?
...not super accurate, but log-y enough...
CL: for this board, no, I wouldn't do that.
>>Decoupling caps?!?! where????
CL: You already have 'em on the power supply? Additionally You might consider putting something like 10uF (exact value isn't critical, but > than say 2.2uF) as close to the V++ as you can for each CMOS chip, + side to the V++ pin, - side to ground.
>>I definitely need an op amp buffer on the VCO (4046 input/4094 clock speed)
CL: not sure about that. I can't see your design while I write this, but in general, You shouldn't need to buffer stages between CMOS chips, only stages between analog stuff (say, analog ins/outs) and your CMOS gates. So if this clock speed is from another CMOS stage, an op amp buffer won't make a difference.
-should the buffering op amp be bi-polar powered? what would the advantages be?
CL: Tying a "normal" op amp V-- to ground will not let signals all the way down to ground go through. You'd need a rail to rail op amp for that which is a slightly more esoteric (and expensive) part.
So I use +/- 15V (or +/-12, etc) power for all op amp buffers unless there is a really good reason not to.
CL: reverselandfill.org has a "noise!Oscillator" CMOS board that reminds me of what you have here a bit. Check it out. Martijn makes really cool stuff.
CL: I'd consider using vactrols vs. pots and voltage control them. I always try to VC as much as I can.... _________________ Visit my AUDIODIWHY blog and website
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
Posted: Tue Dec 11, 2018 5:02 pm Post subject:
Re: CV 8-BIT NOISE/ Random waveform (needs refining) Subject description: suggestions? Buffers? exp/v-per-octave? HELP
let me go over it too, although I don't think I have much to add,..
J-chot wrote:
I MADE A THING!!!!
yep, some random stuff into a shiftregister with an audiorate clk sounds nice . I like the mixpot idea, really adds something nice to it, I do
wonder if the value might be a bit too low though. But as long as nothing goes up in smoke or gets very hot it's ok.
Quote:
WHAT DO I NEED???
-something that's interesting, but probably due to lack of a buffer is that when I remove the pull down resistor on the VCO input, there is a considerable amount of slew, (slide on the CV) but I'm almost certain that it's one of the large caps I stuck across on the power lines.
hmm, what are you controlling it with ? (what's sending the CV). There are 2 things to keep in mind; the CV has to be lower than the supply
voltage of the 4046 and it has to be positive. I would definitely add a buffer to the CV input. You could use a dual supply opamp for this and a
diode in series with the output + a resistor to GND (like the one you already have now).
Quote:
Do I need an expo/volt-per-octave converter for the input? anyway
...not super accurate, but log-y enough...
You don't need it but could be nice, it just changes the frequency responce.
Quote:
-Decoupling caps?!?! where????
A 100nF capacitor per IC close to the power pin(s) and a larger one (10uF...47uF) close to where the power enters the board (2 for a dual supply)
Quote:
-I definitely need an op amp buffer on the VCO (4046 input/4094 clock speed)
-should the buffering op amp be bi-polar powered? what would the advantages be?
There shouldn't be a need to buffer a signal between 2 CMOS chips (unless you connect a lot to 1 output).
You should however add a buffer to the bit noise output or at the very least a resistor in series for protection.
Quote:
I obviously need to ground ALL the extra inputs.....but I'm not so sure about the 4094 and the 4046... what exactly are the inputs?
going by your schematic:
4094: pin 1 (STROBE) and pin 15 (OUTPUT ENABLE) should be connectd to V+ (or you could use something to modulate them)
4046: pin 3, pin 5 and pin 14 should be connected to GND. (I like to connect pin 3 to 4, add a (squarewave) audio signal to pin 14 and take the output
from pin 2).
Joined: Jan 20, 2014 Posts: 44 Location: Decatur AL
Posted: Tue Dec 11, 2018 7:38 pm Post subject:
I don't think the output has DC offset, but Audacity could be removing it automatically....... I don't have an oscilloscope.... I know. I know.... that makes me terrible
I don't think the output has DC offset, but Audacity could be removing it automatically....... I don't have an oscilloscope.... I know. I know.... that makes me terrible
Hah, not many people starting on electronics will have a scope ... and you can get a long way without one .. theory will help a lot, read what you can about that ... a scope still is damn handy of course .. even a primitive one.
Re. the DC offset ... I'm not much of an Audacity user myself, but as it tries to implement all the features CoolEdit had ... I would think it to have a manual DC removal filter .. but not an automatic one.
In this case the DC offset will most likely have been removed before Audacity, in the input capacitor of the sound card. Some sound cards do have a DC pass option though, those usually are expensive (but there is no technical reason for DC passing to be expensive - after all .. you just have to leave the DC blocking capacitor out, basically).
Every circuit having a single sided supply will output a signal having a DC offset. What you'll usually see is that such circuits have a DC blocking capacitor on audio outputs (and inputs as well), but not always. For Lunetta style circuits for instance you might not see them (except on "final" outputs maybe). It depends on what you want to connect to .. the system it will be part off.
Posted: Wed Dec 12, 2018 4:17 am Post subject:
Re: CV 8-BIT NOISE/ Random waveform (needs refining) Subject description: suggestions? Buffers? exp/v-per-octave? HELP
PHOBoS wrote:
you could add a resistor network (like an R2R) to the outputs of the shiftregister to create a random voltage.
Which could then be used on the voltage control for the VCO part to get interesting self modulation (rungler style). _________________ Jan
also .. could someone please turn down the thermostat a bit.
Posted: Wed Dec 12, 2018 5:02 am Post subject:
Re: CV 8-BIT NOISE/ Random waveform (needs refining) Subject description: suggestions? Buffers? exp/v-per-octave? HELP
J-chot wrote:
LOOOONG time lurker, first time contributor!!
...
I MADE A THING!!!!
-something that's interesting, but probably due to lack of a buffer is that when I remove the pull down resistor on the VCO input, there is a considerable amount of slew, (slide on the CV) but I'm almost certain that it's one of the large caps I stuck across on the power lines.
-I definitely need an op amp buffer on the VCO (4046 input/4094 clock speed)
-should the buffering op amp be bi-polar powered? what would the advantages be?
...
what do I need to add?
what did I miss?
Nice toy.
Output signal in this circuit will be bipolar if coupled by capacitor. Direct CMOS output is unipolar.
When adding a buffer on the 4046 input, I would mix some noise (low pass filtered) on this input (as an switch option) to make a shift register clock little random also.
4069 is working in linearized mode so additional buffer (as in original circuit or better Schmitt trigger - 4106 or 40106) would be helpful.
If you want to add analog output buffer you may consider mixing digital output noise with analog input noise from source (analog output noise from R-2R ladder also would be great).
Joined: Jan 20, 2014 Posts: 44 Location: Decatur AL
Posted: Sun Dec 30, 2018 10:19 pm Post subject:
brief note
for the sake of documentation if nothing else. I will do a proper update of the schematic later, also after I try all the neat stuff Phobos and Blue Hell suggested. this slight modification removes all drop-outs experienced with the "noise/random waveform blend" knob, which as I suspected, was due to a lack of buffer from feeding one pin to the other on the same chip, and not due to a cheap breadboard!
My goal with this circuit is to make it into a proper synth module as a noise source if nothing else, but some of these "cheap tricks" can prove quite valuable in terms of functionality... so yeah also I need op-amp buffers in there...
small note.png
Description:
Filesize:
35.34 KB
Viewed:
397 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
I don't think the output has DC offset, but Audacity could be removing it automatically....... I don't have an oscilloscope.... I know. I know.... that makes me terrible
Since Audacity uses your “soundcard”, and I’m sure your sound input is ac coupled (though there are ways to dc couple it) the dc portion of your signal is removed.
A better, more convenient way to look at signals than Audacity is to transform your pc into a two channel scope using an app from Cristian Zeitnitz: https://www.zeitnitz.eu/Scope_en
It also gives you more ways to look at signals, there is even a trigger possibility.
_________________ my synth
Joined: Jan 20, 2014 Posts: 44 Location: Decatur AL
Posted: Wed Oct 16, 2019 2:24 pm Post subject:
One year Later and I'm jumping back into this nonsense...
OK, I've been googling for at least 30 minutes, and I'm still not sure what the difference between a cv buffer and audio buffer circuit...I think my adhd might also be flaring up.
I've got this 50% strip boarded, and I need buffers for vco inputs, possibly a clock input, and audio outputs.
I think all my research just overwhelmed/confused myself...
can I just plug in some opamps wired as unity gain? do I need resistors in there too?
Joined: Jan 20, 2014 Posts: 44 Location: Decatur AL
Posted: Wed Oct 16, 2019 2:25 pm Post subject:
I should maybe clarify that I would like to buffer the potential ins and outs of these chips to work with modular stuff that I own, as well as the diy shit i've been working on.
the difference between a cv buffer and audio buffer circuit
Actually they could be the same because a lot of cv inputs will also accept an audio signal.
I consider THIS a good intro into voltages used in a modular system, it describes the technical details of the Doepfer A100 system.
Please read it and if that arises more questions, just ask. _________________ my synth
Joined: Jan 20, 2014 Posts: 44 Location: Decatur AL
Posted: Thu Oct 17, 2019 8:47 am Post subject:
Grumble wrote:
Quote:
the difference between a cv buffer and audio buffer circuit
Actually they could be the same because a lot of cv inputs will also accept an audio signal.
I consider THIS a good intro into voltages used in a modular system, it describes the technical details of the Doepfer A100 system.
Please read it and if that arises more questions, just ask.
That article didn't answer any of my questions. I still have no idea what kind of buffer circuits to insert into my module.
Control voltage buffer works on DC and not very fast voltages only. So no capacitors on input and output. If control voltage must be precize - e.g. for pitch control - then Op Amps with low offset and low temperature drift are recommended (old cheap OP07C is the example).
Audio buffers may be DC or AC connected. When AC connected, large enough capacitors shouls be placed on input and on output (non-polar recommended). How large? Depends on input/output resistances. OpAmps for audio buffers don't need to be precize, rather fast and if used for eurorack standards, with rather high input impedance (what implies FET-input OpAmps). As typical synth signals are rather high-amplitude (10Vpp), no need for super low-noise chips.
Output resistors that protect buffers should be placed in different ways. For audio buffers - simply 1kOhm at the output, for CV buffers - resistor at the output but included in OpAmp feedback loop (to not lost output voltage level).
can I just plug in some opamps wired as unity gain? do I need resistors in there too?
Heh, the ADHD answer would simply be yes.
Quote:
I need buffers for vco inputs, possibly a clock input, and audio outputs.
You can optimize things a bit depending onthe input type, and for outputs you could use a slightly different kind of buffer.
For VCO inputs, when it is for CV signals you may want multiple inputs that are mixed, which will need a summing opamp circuit. See https://www.electrical4u.com/summing-amplifier/ for instance. Note however that this circuit inverts, you you'll want an invertor after that. When however you already have multiple mixed CV inputs, you will not need any additional bufferening at all.
For output buffering you could add a series resistot after the unity gain opamp circuit, but as most modern opamps are able to handle shorting their output .. I dont really see a need for that .. its a tradition it seems .. I'd just use a unity gain opamp. For a CMOS output howver, for triggering external stuff or instance .. just add a 100 Ohm or so resistor at the output. Are you sure you actuall need output bufferening? when the output comes from an opamp already .. I see no need for that. And the 100 ohm series resistor for CMOS would just be to protect it agains shorts, no bufereing needed there. _________________ Jan
also .. could someone please turn down the thermostat a bit.
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