Author |
Message |
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Tue May 08, 2007 2:09 pm Post subject:
Still yet another sample and hold |
 |
|
Here's a page describing a novel twist on the S/H concept. It's a two-part circuit: the first part fires gate and trigger signals whenever its input crosses a user-set (and voltage-controlled) threshold. The second part is a S/H fired by the trigger signal from the first part.
You can use it: (1) as a standard S/H by inputting a trigger pulse into the first section, (2) to derive gate and trigger signals from any analog signal, (3) to sample one signal based on the amplitude of another, (4) to generate randon/chaotic timing signals from continuous random/chaotic signals, (5) to generate random/chaotic voltage steps from the same, (6) to derive first return maps and Poincare sections of chaotic systems, .........
The S/H section is a high-performance design with under 5 mV/sec droop and a 30 usec charging time, allowing precise sampling at rates up to 5 kHz and more.
Circuit diagram, analysis and a demo with a chaos generator:
http://home.comcast.net/~ijfritz/ch_cir3.htm
(Apologies to Rene and his YASH circuit) |
|
Back to top
|
|
 |
andrewF

Joined: Dec 29, 2006 Posts: 1176 Location: australia
Audio files: 4
|
Posted: Wed May 09, 2007 2:38 am Post subject:
|
 |
|
hi Ian
thanks for sharing this one too.
Are you planning a PCB?
would a LM311 do for the comparator?
your sound sample sounds like those tibetan temple horns after a heady brew of the local datura tea good for scaring off demons...or was it for attracting them??  |
|
Back to top
|
|
 |
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Wed May 09, 2007 6:35 am Post subject:
|
 |
|
andrewF wrote: | hi Ian
thanks for sharing this one too.
Are you planning a PCB?
would a LM311 do for the comparator? |
Andrew -- I would consider doing a board if there was enough interest. Perhaps people already have enough S/H's? I got a bit of flak before when I posted about this on sdiy.
Yeah, you could use LM311's, if you take proper care about decoupling and in/out isolation. I like the 2901/2903 because you get more than one on a chip, and according to National you don't even need decoupling caps.
The Tibetan horns would have sounded better if I actually used a patch instead of just raw VCO's.  |
|
Back to top
|
|
 |
Scott Stites
Janitor


Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Wed May 09, 2007 9:55 am Post subject:
|
 |
|
It attracted me Maybe it's a form of strange attracter?
IMO It's certainly different enough from any sample and hold that I've ever had to fall out of the "Yet another" category - a board would be great in my book!
Cheers,
Scott |
|
Back to top
|
|
 |
vtl5c3
Joined: Sep 08, 2006 Posts: 425 Location: PDX
Audio files: 13
|
Posted: Wed May 09, 2007 10:00 am Post subject:
|
 |
|
The cv control over threshold is a great idea. Makes me want to revisit my old Electronotes S&Hs to see if I can modify them.
BTW, I like the pulse output too... Might be interesting if an inverted pulse out could be added for doing cascaded analog shift register type stuff, especially if you're thinking of doing a quad S&H.
Romeo |
|
Back to top
|
|
 |
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Thu May 10, 2007 10:54 am Post subject:
|
 |
|
vtl5c3 wrote: | BTW, I like the pulse output too... Might be interesting if an inverted pulse out could be added for doing cascaded analog shift register type stuff, especially if you're thinking of doing a quad S&H.
Romeo |
Could you expand on that a bit? You're talking about a negative-going pulse? Or maybe something like a pulse at the end of the gate? |
|
Back to top
|
|
 |
vtl5c3
Joined: Sep 08, 2006 Posts: 425 Location: PDX
Audio files: 13
|
Posted: Fri May 11, 2007 9:12 am Post subject:
|
 |
|
I got the idea from the S&H in Electronotes 61. You take several S&Hs in series. The signal to be processed goes from left to right, the triggers/gates go from right to left. I think that the triggers need to be inverted between stages to delay the previous stage from triggering at the same time as the next one. Sorry if I'm not explaining it well, I'm having a hard time putting it into words. Probably best if you pull out EN 61 and read the article.
frijitz wrote: | vtl5c3 wrote: | BTW, I like the pulse output too... Might be interesting if an inverted pulse out could be added for doing cascaded analog shift register type stuff, especially if you're thinking of doing a quad S&H.
Romeo |
Could you expand on that a bit? You're talking about a negative-going pulse? Or maybe something like a pulse at the end of the gate? |
|
|
Back to top
|
|
 |
Photon

Joined: Mar 22, 2005 Posts: 363 Location: Boston
Audio files: 1
|
Posted: Tue Apr 01, 2008 7:14 am Post subject:
|
 |
|
I was recently building up the pcb for this one and found that all the through holes for the transistors were a bit undersized. None of my TO-92 style 2N3904s would fit. (the 3906's fit though. curious) I tried to gently open one of the holes using a miniature drill and a pin vise and managed to tear the trace off the board. I was able to repair it luckily. I ultimately solved the problem by using some transistor sockets that I had laying around. Did anyone else have this problem or did I get a bum pcb? Or maybe my NPNs had fat leads?
in any case, I'm looking forward to finishing this one up.
peter |
|
Back to top
|
|
 |
andrewF

Joined: Dec 29, 2006 Posts: 1176 Location: australia
Audio files: 4
|
Posted: Tue Apr 01, 2008 7:20 am Post subject:
|
 |
|
I had the same problem - fat trannie legs
just soldered them onto the back of the board |
|
Back to top
|
|
 |
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Tue Apr 01, 2008 9:24 am Post subject:
|
 |
|
Photon wrote: | I was recently building up the pcb for this one and found that all the through holes for the transistors were a bit undersized. |
Sorry about that -- totally my fault. I used the TO92 component out of my PCB layout program without checking the hole size. Turns out it is for devices with round leads. My suggestion is to use a fine file to reduce the lead size a bit. Another method would be to cut the leads off at a steep angle so the end fits into the hole. This error was corrected for the later runs of the board, so not everyone will have the problem.
Ian |
|
Back to top
|
|
 |
Photon

Joined: Mar 22, 2005 Posts: 363 Location: Boston
Audio files: 1
|
Posted: Tue Apr 01, 2008 9:50 am Post subject:
|
 |
|
No problem, Ian.
The transistor sockets that I used have round leads. |
|
Back to top
|
|
 |
|