Post by thread:[EE] Grid of bipolar electromagnetic indicators

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Just a sanity check before I spend a bunch of time building this... Attached is a circuit diagram I did up for a grid of bipolar electromagnetic indicators, also known as dot-flip displays, that I want to build. The attached grid is a 3x3, but what I will actually build will be 56*13 Electrically speaking you apply a roughly 12V pulse of a few ms in either direction. This magnetises two bits of ferrite which hold the colored dot in either orientation, yellow side up, or black side up. I just want to make sure my multiplexing scheme will actually work, and is the simplest possible solution, never done multiplexing before with anything other than leds. Of course, the high-side FETs will have to be PMOS, but other than that I can't think of any gotchas. - -- http://petertodd.org 'peter'[:-1]@petertodd.org -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFIOfjN3bMhDbI9xWQRAkuFAJwOZ1lXRLW4DiakKGJXv7ePara7mwCaAxrV b18AGvgd9/HxQ1C9I4aMvYk= =7hKW -----END PGP SIGNATURE-----

part 1 1113 bytes content-type:text/plain; charset=us-asciiOn Sun, May 25, 2008 at 07:39:57PM -0400, Peter Todd wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Just a sanity check before I spend a bunch of time building this... > > Attached is a circuit diagram I did up for a grid of bipolar > electromagnetic indicators, also known as dot-flip displays, that I want > to build. The attached grid is a 3x3, but what I will actually build > will be 56*13 > > Electrically speaking you apply a roughly 12V pulse of a few ms in > either direction. This magnetises two bits of ferrite which hold the > colored dot in either orientation, yellow side up, or black side up. > > I just want to make sure my multiplexing scheme will actually work, and > is the simplest possible solution, never done multiplexing before with > anything other than leds. Of course, the high-side FETs will have to be > PMOS, but other than that I can't think of any gotchas. The attachment... Also FWIW this is the project: http://petertodd.org/art/entropy-oscillator/ -- http://petertodd.org 'peter'[:-1]@petertodd.org part 2 11494 bytes content-type:image/png (decode) part 3 35 bytes content-type:text/plain; charset="us-ascii" (decoded 7bit)

Peter, Looks OK to me (P channel fets excepted). You might want to add some sort or current limit protection in case something gets "stuck". - Series resistor with a capacitor bypassing it? Or just current limit the power supply? RP 2008/5/26 Peter Todd <spam_OUTpeteTakeThisOuTpetertodd.org>: {Quote hidden}> On Sun, May 25, 2008 at 07:39:57PM -0400, Peter Todd wrote: >> -----BEGIN PGP SIGNED MESSAGE----- >> Hash: SHA1 >> >> Just a sanity check before I spend a bunch of time building this... >> >> Attached is a circuit diagram I did up for a grid of bipolar >> electromagnetic indicators, also known as dot-flip displays, that I want >> to build. The attached grid is a 3x3, but what I will actually build >> will be 56*13 >> >> Electrically speaking you apply a roughly 12V pulse of a few ms in >> either direction. This magnetises two bits of ferrite which hold the >> colored dot in either orientation, yellow side up, or black side up. >> >> I just want to make sure my multiplexing scheme will actually work, and >> is the simplest possible solution, never done multiplexing before with >> anything other than leds. Of course, the high-side FETs will have to be >> PMOS, but other than that I can't think of any gotchas. > > The attachment... > > Also FWIW this is the project: > > http://petertodd.org/art/entropy-oscillator/ > > -- > http://petertodd.org 'peter'[:-1]@petertodd.org > > -

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Mon, May 26, 2008 at 02:10:16PM +1200, Apptech wrote: > > Looks OK to me (P channel fets excepted). > > Looks OK. > Feels like you should be able to reduce the FETs to row / > column but my brain refuses to reveal how. I know 'eh? I've been thinking about this one for months myself. I did find some rather broken commercial flip-dot modules, which had two dioded per element, but the actual driver circuitry was missing. One module I found had no diodes at all, but I suspect that it must have used some magnetic hysterius tricks, like how core memory works, and besides was constructed entirely from bundles of magnet wire woven through the ferrite elements, very different than the sub-modules I actually bought. Using a duel-rail power supply might help, but I'd rather avoid that complications really anyway. - -- http://petertodd.org 'peter'[:-1]@petertodd.org -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFIOh8Q3bMhDbI9xWQRAqE0AKCHsv5btZZUzMaTNXYuUma3r21/QgCgsj5W q1am3J0T/LXYtXH743XrWYo= =Vr1z -----END PGP SIGNATURE-----

At 10:10 PM 5/25/2008, you wrote: > > Looks OK to me (P channel fets excepted). > >Looks OK. >Feels like you should be able to reduce the FETs to row / >column but my brain refuses to reveal how. > > > Russell If they're all *tristate* drivers and you only flip or flop one actuator at a time. Same deal as Charlie-plexing. I don't think it's worth it. >Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" .....speffKILLspam@spam@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Sun, May 25, 2008 at 11:43:08PM -0400, Spehro Pefhany wrote: {Quote hidden}> At 10:10 PM 5/25/2008, you wrote: > > > Looks OK to me (P channel fets excepted). > > > >Looks OK. > >Feels like you should be able to reduce the FETs to row / > >column but my brain refuses to reveal how. > > > > > > Russell > > If they're all *tristate* drivers and you only flip or flop one actuator at > a time. Fortunately, I will only be flipping or flopping one flip dot module at a time. > Same deal as Charlie-plexing. I don't think it's worth it. So that would essentially replace each of the half bridge PMOS/NMOS pairs I've created with a single tristate driver? Isn't the trick with Charlie-plexing that you are exploiting that the combined forward voltage of two leds in series is high enough that a single led in parallel will take all the current? My 12V will be significantly higher than the forward voltage of the diodes used. - -- http://petertodd.org 'peter'[:-1]@petertodd.org -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFIOjwe3bMhDbI9xWQRApBtAJ91LjOALZPYBq6x+YxOrIPxnUP4aQCgmYui Bp50z1Teedocz5QZxVqE/H4= =kWdm -----END PGP SIGNATURE-----

At 12:27 AM 5/26/2008, you wrote: {Quote hidden}>-----BEGIN PGP SIGNED MESSAGE----- >Hash: SHA1 > >On Sun, May 25, 2008 at 11:43:08PM -0400, Spehro Pefhany wrote: > > At 10:10 PM 5/25/2008, you wrote: > > > > Looks OK to me (P channel fets excepted). > > > > > >Looks OK. > > >Feels like you should be able to reduce the FETs to row / > > >column but my brain refuses to reveal how. > > > > > > > > > Russell > > > > If they're all *tristate* drivers and you only flip or flop one actuator at > > a time. > >Fortunately, I will only be flipping or flopping one flip dot module at >a time. > > > Same deal as Charlie-plexing. I don't think it's worth it. > >So that would essentially replace each of the half bridge PMOS/NMOS >pairs I've created with a single tristate driver? It would be about the same, I'm wondering if the diodes can be eliminated. >Isn't the trick with Charlie-plexing that you are exploiting that the >combined forward voltage of two leds in series is high enough that a >single led in parallel will take all the current? My 12V will be >significantly higher than the forward voltage of the diodes used. I think with the sneak paths you'd get 12V/3 = 4V across the coils that are not supposed to be energized. If that's too much to guarantee *no* flipping then it won't work reliably without all those diodes. A 3-cent BAV99 each isn't so bad. Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" speffKILLspaminterlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Mon, May 26, 2008 at 01:56:24AM -0400, Spehro Pefhany wrote: Doh, this ended up in my sent-mail, but never got past that... weird... > >Isn't the trick with Charlie-plexing that you are exploiting that the > >combined forward voltage of two leds in series is high enough that a > >single led in parallel will take all the current? My 12V will be > >significantly higher than the forward voltage of the diodes used. > > I think with the sneak paths you'd get 12V/3 = 4V across the coils that are > not supposed to be energized. If that's too much to guarantee *no* flipping > then it won't work reliably without all those diodes. I did some quick experimentation, the coils can be reliably flipped with as little as 3V with a unknown by hand pulse-width, and that same 3V accross three dots fails to flip them. That said, in a private email someone found the supplier of my modules. As far as I can tell they just put their website online within the past few months: http://www.flipdots.com/EN/electromagnetic_displays/products/page-5/dot_stripes.html The 2.7" modules are the exact ones I have, which they are now making with original tooling bought from the defunct MARK IV FP Electronics. They specify a 350ma minimum current, achievable with 1ms pulse at 7.5V given the coils inductance. Anyway, I may build up a pulse tester and see how well things work in temperature extremes, and from unit to unit, although diodes would be safer. FWIW this is a quite nice work of art using these things: http://www.troika.uk.com/cloud.htm > A 3-cent BAV99 each isn't so bad. Yup, $64.74/1000x and $200 worth of labour. :) - -- http://petertodd.org 'peter'[:-1]@petertodd.org -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFIYl8w3bMhDbI9xWQRAkiwAJ4t8JL+yyA1FbSfmSbHXq5AvlMPTACeIIPh UrS/3K3RKl44RneaEKI6aDs= =HYeY -----END PGP SIGNATURE-----