Author |
Message |
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
Posted: Tue Sep 25, 2012 8:20 am Post subject:
|
|
|
Hi again,
Thanks for the input Brock & Jingle Joe. You spotted my "deliberate" mistake Brock. That's what comes of trying to do something from memory with a heavy cold, I think. Of course, it should have been a 2R to 0V.
I have been fiddling around with the circuit a little (as I feel a bit lousy at the moment), but I still haven't found a 100% reliable solution to the problem. It does seem to be powering up in the steady state condition (everything's high on 4006, therefore the outputs on the 4077 are all high). I've tried a adding a pull-up and pull-down to the clock output (not at the same time!) but that hasn't solved it either. I'm using a 555 astable at present to clock it. Do you think that a CMOS oscillator (maybe a 40106 or a 4093 based oscillator) might make any difference? I may be grasping at straws here.
Thanks for your comments about the crappiness or otherwise of my VCO, JingleJoe. I have a habit of putting myself down all the time. I must get out of it!
Cheers,
Gary |
|
Back to top
|
|
|
JingleJoe
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
|
Posted: Tue Sep 25, 2012 10:48 am Post subject:
|
|
|
The oscillators used to clock it shouldn't make much difference, the 555 from my observations, has a longer initial pulse as the cap charges then goes into it's regular timing. This should not cause any problems. _________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Green Dungeon Alchemist Laboratories |
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
Posted: Tue Sep 25, 2012 11:36 am Post subject:
Randomizer |
|
|
That's what I thought - I said that I was clutching at straws. I just wondered if the duty cycle might affect its operation. With a 555 it's always >50%, unless you add diodes, if my memory serves me. Anyway, I'll persevere - I usually work it out in the end.
Thanks again for your help,
Gary |
|
Back to top
|
|
|
brock
Joined: May 26, 2011 Posts: 112 Location: Canada
|
Posted: Tue Sep 25, 2012 1:36 pm Post subject:
|
|
|
A quick fix, not necessarily a good fix, is to add an inverter on one of the shift register inputs or outputs. This will get you past the power on steady state condition but doesn't mean you won't create another. Another option is add a series resistor on one of the inputs and a normally open switch to ground. When the SR gets to the all high steady state just press the switch and you'll be up and running again. Or you can design some fault detection logic that automatically toggles an input when steady state is detected. |
|
Back to top
|
|
|
JingleJoe
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
|
Posted: Tue Sep 25, 2012 2:12 pm Post subject:
Re: Randomizer |
|
|
I like the idea of making fault detection logic for it that would be a fun little challenge.
analog_backlash wrote: | I just wondered if the duty cycle might affect its operation. With a 555 it's always >50%, unless you add diodes, if my memory serves me. Anyway, I'll persevere - I usually work it out in the end.
Thanks again for your help,
Gary |
That's right, but the duty cycle doesnlt matter, you can even use tiny pulses to clock things which is useful for some applications _________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Green Dungeon Alchemist Laboratories |
|
Back to top
|
|
|
RingMad
Joined: Jan 15, 2011 Posts: 427 Location: Montreal, Canada
Audio files: 4
|
Posted: Tue Sep 25, 2012 3:47 pm Post subject:
|
|
|
I rarely have problems with the 4006-4077 random generator now. I've determined the fastest clock speed that is "safe", and put a resistor in series with my speed pot so it doesn't go faster than approximately 385 Hz.
My oscillator is the standard 4093-based one, with a 4.7 uF cap and a 470 ohm res in series with a 150K pot.
The only time it might lock up now is when I turn my machine on. As mentioned, I have a switch that can cut the connection between 4006 pin 1 and 4077 pin 10. Toggling that usually unlocks it. Admittedly, this is not ideal.
The French book my friend photocopied the circuit from only had the schematic... no information at all. Maybe it was just theoretical.
brock wrote: | the circuit isn't latching up, it's just running through all possible states faster until it reaches a steady state. |
I read that in the other thread, and it seems like it could be a logical conclusion. However, I tested it, running the circuit for 3 hours non-stop using a clock about 385 Hz, but it never locked up.
One problem is that this shift register has no reset or clear pin. I guess the only way to do that is to cut the chip's power.
JingleJoe: that's cool that you are thinking about randomizers. One thing that would be good is to have a "cheap" one. i.e. low chip count, low number of clocks required.
James. |
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
Posted: Tue Sep 25, 2012 4:09 pm Post subject:
Randomizer |
|
|
Thanks for that James - I'll try your solution next (I obviously didn't read your posts thoroughly). I did once design a circuit to do that kind of thing automatically on power up (probably using a 4016 or 4066 somewhere). I'll dig that out, but it may be a sledgehammer to crack a nut. I could just use a push to break button I assume.
Cheers,
Gary |
|
Back to top
|
|
|
brock
Joined: May 26, 2011 Posts: 112 Location: Canada
|
Posted: Tue Sep 25, 2012 9:44 pm Post subject:
|
|
|
RingMad wrote: |
brock wrote: | the circuit isn't latching up, it's just running through all possible states faster until it reaches a steady state. |
I read that in the other thread, and it seems like it could be a logical conclusion. However, I tested it, running the circuit for 3 hours non-stop using a clock about 385 Hz, but it never locked up.
One problem is that this shift register has no reset or clear pin. I guess the only way to do that is to cut the chip's power. |
385 Hz is remarkably slow even for old 4000 series cmos. I wonder what's going on there? But even that rate is enough to go through a maximal sequence of (2^18) -1 at least 15 times so it probably isn't hitting a steady state condition at the higher rate. I wonder what the maximal length is of this shift register configuration?
Another solution to the powering up with all bits high is to swap the 4077 for a 4070. This will make the inputs to the 4006 all low.
This design is a low chip count randomizer. Unfortunately the 4006 is becoming unobtainium. It's too bad the 4031 doesn't have some useful outputs for taps to make a linear feedback shift register. You could combine it with a 4015 and a 4070 to make a maximal length LFSR of 71 bits with taps at 71 and 65. Might be longer than you would ever need. |
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
Posted: Wed Sep 26, 2012 2:26 am Post subject:
Randomizer |
|
|
Hi Brock,
I did wonder what would happen if I replaced the 4077 XNOR with a 4070 XOR. I'll try that as well (it's very straightforward as the pin-outs are the same). Also, I know of the problem of the 4006 becoming hard to obtain, so I did buy several before they become impossible to find. I wanted a spare anyway, since my old Transcendent 2000 uses one to produce pseudo-random white noise and you never know when it might give up the ghost. It also uses CA3080 OTAs which are obsolete now. I have a few of these spare.
Just as a side issue on obsolete components, you can see places (usually in the far-east) on eBay which seem to be selling things like the CA3080 and other obsoletes (e.g. MN3102, MN3207, PT2399 etc.). I have never had the courage to order any, since I'm worried about being ripped off. Does anyone know of any decent eBay suppliers, or should you not touch them with a bargepole?
Thanks again for your help,
Gary |
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
Posted: Wed Sep 26, 2012 7:34 am Post subject:
Randomizer Success!!!! |
|
|
Hi Guys,
I think that I've finally got it to work properly. I tried several of the ideas, including replacing the 4077 XNOR with a 4070 XOR. This half worked, but the R/2R ladder was only producing 4 voltage levels instead of 16 (2^n and all that). I assumed (correctly) that this meant that there were identical waveforms being produced on the some of the outputs of the 4070 (in fact, pin 3 = pin 10 and pin 4 = pin 11).
At first, I starting to think about inverting the pin 10 and pin 11 outputs (making the circuit more complicated), but then I thought that I'd experiment with the 4006-4070 interconnections. My first attempt, was to reverse the connections to pin 5 and 6 of the 4006 (i.e. 4006 pin 5 to 4070 pin 4 and 4006 pin 6 to 4070 pin 3) and it worked straight away! It also no longer needed the pull down resistor on pin 3 of the 4070.
I have attached a revised schematic (well, one with some red pen on it anyway). I have also shown the correction to the R/2R ladder, spotted by Brock. I haven't put any new waveforms up, as they look virtually identical to the earlier ones.
Now, I can finally try it out with the sequencer and see how it sounds. I hope that it was worth all the effort.
Cheers,
Gary
Description: |
|
Filesize: |
230.34 KB |
Viewed: |
523 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
|
|
Back to top
|
|
|
RingMad
Joined: Jan 15, 2011 Posts: 427 Location: Montreal, Canada
Audio files: 4
|
Posted: Wed Sep 26, 2012 6:48 pm Post subject:
|
|
|
I think you've got it, Gary! I made the changes you suggested and it appears to work fine. I crank the clock speed to about 1600Hz and it doesn't lock up, although it sortof becomes not-too-interesting, or at least it seems to barely change. But then maybe it's because the R/2R and VCO (which is what I'm testing with) can't follow the voltage changes that quickly? Anyway, I don't need to go near as high for what I use it for.
Truth be told about the pulldown res on the 4077's pin 3... after mentioning it, I discovered it didn't really solve the problem... I just never got around to updating the thread.
Now, if you want really fast bits there's the XOR chain or ring which can produce a kindof white noise cloud of bits when clocked fast. There was a schematic by ezekiel, but he may have posted it only over on deathlehem.
I shall update the schematic soon in the other thread. I'm going to do more tests... it seems to be a little too repetitious, but I'm not sure.
I tried to find a good way of resetting the pattern, but it's tricky... e.g. one can cut the power to either chip, and it still keeps going. Even cutting both chips, it sortof still runs! Alas, I ended up frying a precious 4006, and still don't have a good solution.
Thanks again for your work on it.
James. |
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
Posted: Thu Sep 27, 2012 6:52 am Post subject:
Randomizer |
|
|
Hi James,
Glad that yours worked as well - it wasn't just an abberation I think that I was quite lucky in my guess at which two wires to invert (although I didn't try any others afterwards). I'm just working on a mod to my 'Super Stylophone' in order to find out what it sounds like. I will post the results.
The other circuit that you are describing sounds a lot like the white noise source on my old, old synth (the Transcendent 2000). I'd probably break Tim Orr's copyright if I post it (although I'm sure it's already on the web somewhere). I did once do my own version of it to produce a very low speed stream of 1s and 0s. I'll look for that circuit and post it.
Thanks for your words of encouragement - I'm sure we'll be in contact again in the future.
Gary |
|
Back to top
|
|
|
JingleJoe
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
|
Posted: Thu Sep 27, 2012 11:36 am Post subject:
|
|
|
RingMad wrote: | maybe it's because the R/2R and VCO (which is what I'm testing with) can't follow the voltage changes that quickly?
|
I support this notion. Try an oscilloscope
P.S. if anyone needs any 4006s and can't find any anywhere, give me a shout, my local electronics shop has some. _________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Green Dungeon Alchemist Laboratories |
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
Posted: Sat Sep 29, 2012 9:05 am Post subject:
Ramdomizer Noises |
|
|
Hi All,
WARNING: DO NOT LISTEN TO THESE RECORDINGS IF YOU LIKE MUSIC
I've finally finished my mods (well nearly, see below) and now I have some unpleasant noise for you. I also have a very exciting picture of my breadboard hooked up to my SS showing the modification (inset). The noises have been very hastily recorded and edited (badly) with fade-ins and fade outs. 2 mp3 files are detailed below. Unless other specified, pitch changes are controlled by the R/2R ladder.
Randomizer Without
This uses no envelopes, VCF or VCA, so is almost a Lunetta.
Part 1: -1 octave (relative to master oscillator) alone with glide added towards the end
Part 2: -1 octave + 4011 ring mod
Part 3: -1 octave + APC
Part 4: -1 octave with very slight CV changes + drone
Part 5: Free for all
Randomizer With
This has envelopes, VCF and VCA.
Part 1: -2 octaves + some triggered envelopes controlling VCF and VCA
Part 2: -1 octave (VC from sequencer pots with glide) with VCF (CV from envelope and R/2R ladder).
Part 3: -1 octave + drone VCF controlled R/2R ladder, VCA controlled by sequencer pots
I realised whilst recording this, that my envelopes are not triggering at the same time as my pitch changes occur, from the R/2R circuit. I think I know why. I had the same problem on one of my flightcase synth circuits and it just needed an inverter to be added.
Anyway "enjoy",
Gary
Description: |
|
Filesize: |
231.71 KB |
Viewed: |
342 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Download |
Filename: |
Randomizer Without.mp3 |
Filesize: |
3.83 MB |
Downloaded: |
1012 Time(s) |
Description: |
|
Download |
Filename: |
Randomizer With.mp3 |
Filesize: |
1.52 MB |
Downloaded: |
955 Time(s) |
|
|
Back to top
|
|
|
JingleJoe
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
|
Posted: Sun Sep 30, 2012 3:11 am Post subject:
|
|
|
mmmm metroidy very nice (but my scope seems to suggest you have too much highpass filtering on your output or somewhere in your signal path for my liking ) _________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Green Dungeon Alchemist Laboratories |
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
Posted: Sun Sep 30, 2012 3:39 am Post subject:
Highpass Filtering |
|
|
Hi JingleJoe,
I'm sure you're right about the HP filtering, but it was not intentional. I'll have a look at my circuits again. I'm thinking off the top of my head, that it could be the mixer that's doing it. Also, I need to add proper bypass switches to avoid the VCF and VCA stages. For my "Randomizer Without" recording, I just put the filter on the highest frequency (but that's lowpass) and turned the gain up on the VCA.
All this, of course, wouldn't be a problem if I'd built a proper, fully patchable Lunetta in the first place. This machine is very much a hybrid of lots of things that I wanted to try out, with no real idea of what it was meant to end up like.
Anyway, thanks for your useful comments,
Gary
P.S. I hope that the move is going well. |
|
Back to top
|
|
|
JingleJoe
Joined: Nov 10, 2011 Posts: 878 Location: Lancashire, England
Audio files: 14
|
Posted: Sun Sep 30, 2012 5:08 am Post subject:
|
|
|
Hahaha thanks for the well wishing I really liked all the sounds it makes, it sounds like it all works reasonably well _________________ As a mad scientist I am ruled by the dictum of science: "I could be wrong about this but lets find out"
Green Dungeon Alchemist Laboratories |
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
Posted: Mon Oct 08, 2012 5:17 am Post subject:
Radomizer + Clock Inverter Recordings |
|
|
Hello I'm back again.
Sorry for any strange posts you've seen from me lately - I'll stick strictly to electronics & music in future...
Just adding a few recordings which I've done since 'fixing' my randomizer by adding a clock inverter. As I said before, this means that I can trigger my envelope generators at the same time as the R/2R CVs change and it gives a far better sound. The details are as shown below. On all 3 recordings the master oscillator frequency is controlled by the R/2R output, VCF is controlled by the sequencer CV output + ASD envelope 1 and VCA is controlled by ASD envelope 2.
Randomizer With Clock Inverter 1:
Fade in -1 octave signal, then user defined waveform, then APC.
Do a bit of knob twiddling.
Switch off some of the EG1 triggers, controlling the VCF.
Switch off some of the EG2 triggers, controlling the VCA (leading to silent gaps).
Do a bit more knob twiddling.
Randomizer With Clock Inverter Chimes:
Just uses -1 octave & 4011 ring modulator with both ASD triggers operating on every clock pulse. No twiddling. Fade-in and fade-out added afterwards.
Randomizer With Clock Inverter Doom:
Just uses -2 octave, user defined waveform and white noise with both ASD triggers operating on every clock pulse. No twiddling. Fade-in and fade-out added afterwards.
See what you think,
Gary
P.S. I know that EGs, VCFs and VCAs are cheating in Lunettaland, but they do sound good. I will start building a genuine Lunetta eventually...
Description: |
|
Download |
Filename: |
Randomizer With Clock Inverter 1.mp3 |
Filesize: |
1.61 MB |
Downloaded: |
955 Time(s) |
Description: |
|
Download |
Filename: |
Randomizer With Clock Inverter Chimes.mp3 |
Filesize: |
722.49 KB |
Downloaded: |
976 Time(s) |
Description: |
|
Download |
Filename: |
Randomizer With Clock Inverter Doom.mp3 |
Filesize: |
1.09 MB |
Downloaded: |
1011 Time(s) |
|
|
Back to top
|
|
|
RingMad
Joined: Jan 15, 2011 Posts: 427 Location: Montreal, Canada
Audio files: 4
|
Posted: Wed Oct 10, 2012 5:35 am Post subject:
Re: Ramdomizer Noises |
|
|
analog_backlash wrote: | WARNING: DO NOT LISTEN TO THESE RECORDINGS IF YOU LIKE MUSIC |
Ha! I like music and I listened to these anyway! I see no conflict. Some interesting sounds in there!
Sheesh, I thought I had written this reply ages ago. Oh well, better late than never.
James |
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
Posted: Wed Oct 10, 2012 6:28 am Post subject:
|
|
|
No problem about the delay in responding James - I'm glad that you appreciated the din!
Gary |
|
Back to top
|
|
|
|