Just wanting to know if it is possible to make a BBD out of discrete components like caps, resistors, transistors, etc. Just curious...
Any input welcome.... :)
In theory, yes.
In practice: you won't live long enough to put many stages together, and stray parasitics will give bad performance compared to a chip.
It is literally the same question as: can you build computer memory at home? Yes, a lot of computer memory was built by hand from common parts. It took a government contract to put a few K of memory together. Whenever possible they used methods that scaled better. Writing on a CRT. Weaving iron beads. A more direct technology for audio delay: bouncing ultrasound in a bucket of Mercury.
I remember a thread from like 10 years ago or more where a user named stm came up with a 0.5ms delay with something like 24 cascaded all-pass sections.
If I recall it was intended for a TZF flanger. So maybe 8000-10,000 opamp filter stages for 500ms...? :icon_lol:
found it!
https://www.diystompboxes.com/smfforum/index.php?topic=25168.0
photos are gone, but it's a fun read anyways!
Blast from the past, sometimes I wonder what happened to some of the old-school posters like Ton aka puretube.
Ok.. sad though, i wanted to build a delay with out bbd chips .... heh maybe next time :)
Quote from: 11-90-an on June 24, 2020, 03:05:57 AM
Ok.. sad though, i wanted to build a delay with out bbd chips .... heh maybe next time :)
PT2399? FV-1? There are plenty of non-BBD options out there.
There's plenty of BBDs also....
You can also buy yourself a 50ft garden hose, coil it up, stick a speaker at one end and a mic at the other. You'll get a reasonable slapback echo.
Of course not exactly gig-friendly.
Quote from: Mark Hammer on June 24, 2020, 10:37:34 AM
You can also buy yourself a 50ft garden hose, coil it up, stick a speaker at one end and a mic at the other. You'll get a reasonable slapback echo.
Of course not exactly gig-friendly.
It would be smaller and lighter than the amp I used to carry to gigs, i.e., a trivial addition to my standard gear load. The stage visual would be memorable.
"Ravenous Bark Beetle*? They were the band with the garden hoses. Remember?"
*I'm pretty sure that name is available if anyone wants to use it for a band (or a pedal). :icon_wink:
Quote from: Mark Hammer on June 24, 2020, 10:37:34 AM
You can also buy yourself a 50ft garden hose, coil it up, stick a speaker at one end and a mic at the other. You'll get a reasonable slapback echo.
But I have a 50ft garden hose... :o
What type of mic would be good for this, anyway? :icon_question:
Quote from: 11-90-an on June 24, 2020, 11:03:15 AM
But I have a 50ft garden hose... :o
What type of mic would be good for this, anyway? :icon_question:
A waterproof one. Just in case... ;)
Plenty of inexpensive electret condenser capsules available. If you're serious about doing this, I'd recommend buying one of those plastic end-caps that screw onto the end of a hose or spigot, with the standard thread. Drill a suitable hole in it that allows you to insert the mic capsule in the middle.
The other end of the hose is to be treated somewhat the same as a talk box. A speaker of reasonable output level would have some sort of funnel or bezel that directs its output into the hose.
Note that delay time will depend on tube length, but as tube length increases, bandwidth will decrease. As well, because sound will bounce around inside the hose, it won't be discrete repeats, so much as a blend of one repeat and a bit of reverb-ish mush. It may be musically valid, but will be no substitute for a simple PT2399 circuit.
If you're up for the experiment, I tip my hat to you. But realistically, it's an awful lot of trouble to go to for something that may sound interesting but won't sound especially good.
Quote from: Mark Hammer on June 25, 2020, 09:14:21 AM
Plenty of inexpensive electret condenser capsules available. If you're serious about doing this, I'd recommend buying one of those plastic end-caps that screw onto the end of a hose or spigot, with the standard thread. Drill a suitable hole in it that allows you to insert the mic capsule in the middle.
The other end of the hose is to be treated somewhat the same as a talk box. A speaker of reasonable output level would have some sort of funnel or bezel that directs its output into the hose.
Note that delay time will depend on tube length, but as tube length increases, bandwidth will decrease. As well, because sound will bounce around inside the hose, it won't be discrete repeats, so much as a blend of one repeat and a bit of reverb-ish mush. It may be musically valid, but will be no substitute for a simple PT2399 circuit.
If you're up for the experiment, I tip my hat to you. But realistically, it's an awful lot of trouble to go to for something that may sound interesting but won't sound especially good.
Yessir, thanks for the input. Quarantine driving me nuts, i guess :P
I built a very short optical delay some years ago - the pictures and gone, but don't bother - and the thread came up with a few other innovative options for DIY delay:
https://www.diystompboxes.com/smfforum/index.php?topic=116243.0
I'v been meaning to do this as a ridiculous yet functional art piece. Kinda like this;
(https://i.postimg.cc/DW3CBpq7/1-2x.jpg) (https://postimg.cc/DW3CBpq7)
There are plenty of BBDs out there so this FAR from practical. I'm in it for the the amusement and to educate on these delay circuits.
I feel the market of in-production devices are also expanding as demands increase - between XVIVE and CoolAudio.
-KM
BBD chips are relatively rare and scarce now. But one easy way to do this is with a CCD camera chip. Modulate a LED with audio, focus it on a particular pixel and let a relatively show (for video) clock send it to the output. You have access to all the pixels so you can use any particular pixel on the sensor to select the time delay, as well as modulating the clock frequency.
Building the optics necessary to focus an LED requires a lens (which, if you are using a camera, is going to be there) and a restrictor plate with an opening one pixel in size. But the latter could be simulated by just clocking the output through and taking only the sample that came out at a specific time as the valid output and dropping it into a sample-and-hold. That way, there would be no need for an optical restrictor plate - you just select the output that came out at a particular time.
There are a lot of cameras that have been scrapped due to low resolution and a number of phones that have this feature. Camera CCD's should not be hard to come by.
Quote from: amptramp on June 25, 2020, 03:42:13 PM
BBD chips are relatively rare and scarce now.
???
Only if you're looking for 512 stage - single or dual BBDs. Everything else is currently back in production.
-KM
Could a sample-and-hold circuit hold a certain note...? ???
Quote from: 11-90-an on June 25, 2020, 11:56:05 PM
Could a sample-and-hold circuit hold a certain note...? ???
I think you're mixing this thread with the one I had posted.
They "hold" a voltage. That voltage can translate to a note through a VCO (voltage controlled oscillator) as a synth voice or relevant to my thread - the thread you probably meant to post in, it was more about holding a voltage for generic recall purposes. Which is the same principal.
The sample/hold stuff we're used to here for stompboxes is mostly in respect to the FSH-1 where the circuit uses a transistor as a noise source for *randomish* voltages to sample and holds to control a filter causing the sporadic filter effect. Imagine playing a wah pedal with severe parkinson's disease.
-KM
Quote from: Kevin Mitchell on June 25, 2020, 03:50:04 PM
Quote from: amptramp on June 25, 2020, 03:42:13 PM
BBD chips are relatively rare and scarce now.
???
Only if you're looking for 512 stage - single or dual BBDs. Everything else is currently back in production.
-KM
+1 agree with Kevin. Far from being scarce, there are far more BBDs now than there were ten years ago.
There are still things I'd like to see that we don't have, and I don't understand why modern chips seem to have worse clock input capacitance than the originals, but there's definitely plenty of options.
> buy yourself a 50ft garden hose ............... .............. ............. an awful lot of trouble to go to for something that may sound interesting but won't sound especially good. linkback (https://www.diystompboxes.com/smfforum/index.php?topic=124779.msg1185171#msg1185171)
And yet: this was a commercial product taken-up by a Big Brand. Cooper Time Cube. ("timecube" is totally a different thing.)
https://media.uaudio.com/blog/2009/05/doctors_03.jpg
https://reverb.com/item/2486446-vintage-urei-cooper-time-cube-stereo-acoustic-delay
https://www.uaudio.fr/blog/wiggly-frequency-response-cooper-time-cube/
https://forum.fractalaudio.com/threads/cooper-time-cube-a-delay-through-a-garden-hose-can-we-axe-it.141354/
Quote from: ElectricDruid on June 26, 2020, 08:40:15 PM
There are still things I'd like to see that we don't have, and I don't understand why modern chips seem to have worse clock input capacitance than the originals, but there's definitely plenty of options.
Hold up, did I miss that in the new data sheets? Or is that from experience.
Or did you mean Reticon vs. Panasonic.
Any way, I built a CMOS based BBD to play with some sample rate reduction, fun, but rubbish. Can't see why you'd want a home-brew for any real world purpose.
Quote from: Scruffie on June 27, 2020, 08:14:57 AM
Quote from: ElectricDruid on June 26, 2020, 08:40:15 PM
There are still things I'd like to see that we don't have, and I don't understand why modern chips seem to have worse clock input capacitance than the originals, but there's definitely plenty of options.
Hold up, did I miss that in the new data sheets? Or is that from experience.
No, it's in the data sheets. Compare CoolAudio with the old MN-series:
MN3208: 1400pF, V3208: 2800pF
Interestingly, the difference disappears when you get to the 4096-stage device:
MN3208:2800pF, V3205: 2800pF
MN3207: 700pF, V3207: ???
I'm pretty sure I've seen a similar disparity for the 3207, but the V3207 datasheet seems to have dropped that information now. I can't even find datasheets fro the Xvive clones, so if they know, they're not telling.
Honestly though, I'm not sure I trust Coolaudio's datasheets. They look look copy-pasted from somewhere else, so I don't know that I believe all the figures. Maybe they're really the same, I don't know. I should test it, I guess.
Quote from: ElectricDruid on June 27, 2020, 02:49:28 PM
Quote from: Scruffie on June 27, 2020, 08:14:57 AM
Quote from: ElectricDruid on June 26, 2020, 08:40:15 PM
There are still things I'd like to see that we don't have, and I don't understand why modern chips seem to have worse clock input capacitance than the originals, but there's definitely plenty of options.
Hold up, did I miss that in the new data sheets? Or is that from experience.
No, it's in the data sheets. Compare CoolAudio with the old MN-series:
MN3208: 1400pF, V3208: 2800pF
Interestingly, the difference disappears when you get to the 4096-stage device:
MN3208:2800pF, V3205: 2800pF
MN3207: 700pF, V3207: ???
I'm pretty sure I've seen a similar disparity for the 3207, but the V3207 datasheet seems to have dropped that information now. I can't even find datasheets fro the Xvive clones, so if they know, they're not telling.
Honestly though, I'm not sure I trust Coolaudio's datasheets. They look look copy-pasted from somewhere else, so I don't know that I believe all the figures. Maybe they're really the same, I don't know. I should test it, I guess.
The Xvive datasheets bar the MN3005 just link to the original Panasonic ones.
But so it does, I even have an older copy of the v3208 datasheet saved and it says it on that too... how weird.
Xvive datasheet download page: http://www.xviveaudio.com/art/download-a0044.html (http://www.xviveaudio.com/art/download-a0044.html)
Just thought about something right now and wanted to ask you folks...
So apparentlay how an octave up works using CMOS is using a XOr gate as an edge detector by delaying the signal a bit with some NOT gates...
(https://i.postimg.cc/9zNPgZMv/810-CB700-ECD4-4-C5-F-B135-4-CA253-D6-F97-F.png) (https://postimg.cc/9zNPgZMv)
Just wondering what would happen with, say, 36 of those NOT gate stages (without XOR)? I know that this would be very fuzzy... so probably could be fixed(ish) with a square wave to sine wave converter? Would add a mix pot as well.. what do you guys think? (Obviously would be impractical but hey, its experimentation... 8) )
EDIT: googled a bit, and came to this realization...
https://en.m.wikipedia.org/wiki/Propagation_delay
- 1 logic gate has a genaeral maximum delay of about 10 nanoseconds(ns) so for a 500ms delay you would need 5000 NOT gates... about 834 cd4069s...
- it would be better to use wire, as it states in wikipedia that 15cm of wire has about 1ns delay... so that would be 750meters of wire for 500ms delay.... about $20 if you use very cheap wire...
Hope this would enlighten other people trying to attempt a no-bbd delay... haha :D
> 750meters of wire for 500ms delay....
That can't be right? You or the Wiki has slipped a decimal-bead somewhere?
It says that a mile of wire is a milli-Second, so here to Grandma's is a whole second. Even in the bad days of municipal Missouri telephone operations, she wasn't that far away.
And that calls switched through a geo-sync satellite (we used to do that!) would be most of a minute round-trip. We stopped routing voice through geo-sync satts but the latency was not THAT bad.
I learned it as "light travels 1 nanosecond per foot". This proves that GHz computers can never be made. (Not in days when computers were the size of a truck.) (Yes, I am typing at a 3GHz computer....)
That's naked light. In a wire (or a optic-fiber) is slower and I will not argue 15cm=6" is a nano in wire. (I'd expect 7 or 8 inches but what's the point?)
I get 189 miles (303km) makes 1 milliSecond. ?? 17mS across the USA, 34mS there and back. 64mS in slow wire. ???
About like 50 feet of garden hose. And as bad as hose losses are, wire-losses are worse.
While sub-mS delays can be useful in audio (and I guess are the specialty of BBDs), many-mS delay opens more possibilities.
oh wait... my calculations were wrong.... i was typing ver late at night... it should be 75,000meters, not 750.... more realistic haha :icon_redface:
chorus is about 20ms-50ms... that should be about 3kilometers to 7.5km.... :icon_mrgreen:
Geostationary is 35,786 km (22,236 mi) up and radio signals take approximately 0.25 of a second to and fro.
Quote from: PRR on July 01, 2020, 01:16:28 AM
Geostationary is 35,786 km (22,236 mi) up and radio signals take approximately 0.25 of a second to and fro.
so making a pedal that would convert guitar signals to radio signals and bounce it off geostationary twice, in theory, have 0.5 seconds delay / 500ms... hmmm ??? :o
A few years ago, we had a rather interesting conversation here about some rather impractical, yet thought-provoking delay techniques. My favorite involved the moon.
https://www.diystompboxes.com/smfforum/index.php?topic=116204.0
And a few years later, new forum members will be astonished of discussion substances made a few years ago.. :icon_lol:
Here is the thread for all of our predictions, some of which have not happened due to lack of imagination and other which have not happened due to common sense:
https://www.diystompboxes.com/smfforum/index.php?topic=105617.0
In it, I proposed a wax cylinder delay where the music is recorded onto a wax cylinder, picked off at a variable tap angle for variable delay and heated to erase then cooled to form a hard wax surface, all in one revolution.
We also had a moonbounce delay but it is a fixed 2.58 second delay and it is only available during half of the day.
Quote from: 11-90-an on July 01, 2020, 04:00:20 AM
Quote from: PRR on July 01, 2020, 01:16:28 AM
Geostationary is 35,786 km (22,236 mi) up and radio signals take approximately 0.25 of a second to and fro.
so making a pedal that would convert guitar signals to radio signals and bounce it off geostationary twice, in theory, have 0.5 seconds delay / 500ms... hmmm ??? :o
You can use the delay of a "Facebook Live" :) Several seconds!
https://www.youtube.com/watch?v=lO9d-AJai8Q
Pipe delay:
https://www.youtube.com/watch?v=RlsZnpyZmIo
https://www.youtube.com/watch?v=rJSCQapZ-M4
Moving mic phaser:
https://www.youtube.com/watch?v=EDV1GmfHJds
Chorus:
https://www.youtube.com/watch?v=rjmqM-FbTsc
Panning + phasing + tremolo
https://www.youtube.com/watch?v=t0EoQST8p18
Quote from: GFR on July 02, 2020, 08:27:18 AM
You can use the delay of a "Facebook Live" :) Several seconds!
https://www.youtube.com/watch?v=lO9d-AJai8Q
That's very cool, and a catchy tune too!
Quote from: GFR on July 02, 2020, 08:27:18 AM
You can use the delay of a "Facebook Live" :) Several seconds!
https://www.youtube.com/watch?v=lO9d-AJai8Q
I expected to watch a couple seconds, say "neat," and be on my way. Six minutes later I'm looking up more of that band. Super cool stuff.
EDIT: A couple hours later I now own that band's album :P
Quote from: antonis on July 01, 2020, 06:41:37 AM.... years later, .... discussion substances made a few years ago.. :icon_lol:
(https://i.postimg.cc/B8gzbXXJ/XOPHONIC-42.gif) (https://postimg.cc/B8gzbXXJ)
Well yeah, Here's the Video
PRR
You did not post about
The ABC computer with rotating caps used as the memory
Remember the early Xerox copiers with the rotating selenium drums? These use a selenium-coated drum that is charged up to a high voltage then the image of a document is placed on them and lit up with visible light. The light discharges the drum where it hits it and the charge left on the dark parts of the picture are carried onto toner that is printed onto the new page, making a copy.
We could use the same technology to load the selenium drum with a light level proportional to the audio voltage and pick it off with a capacitive sensor at a variable angle from where it is introduced. We would not need the full width of a Xerox drum but with some width, we could use this as a delay for multiple independent channels.
Old Xerox machines must be cheap by now.