[sdiy] Re: [avrchat] : Led-matrix (slightly off topic)

[Happy Harry]

Maybe you should use a larger transistor, paralleling transistors
is bad unless they are matched or have another means of forcing
current sharing.

BUT the 1A rating is continuous. There is a Transient Thermal Impedance that 
implies that if the pulse is short enough, and infrequent enough, you can 
push MUCH harder than the steady state
rating. For large semis like IGBT's etc this is a published spec.

See also the half-sine diode ratings... for a single pulse the
current is WAY higher than the steady state rating. You can't push
THIS hard more than once (before letting the junction cool...) but
you can push harder in pulse apps.

Another trick would be to use two transistors and pulse each one
on every other scan....  NAH !

Check out the small MOSFETS like the VN12 series from Siliconix... they are 
in a TO-92 style package but with a metal tab like a baby
TO-220. Lot better thermal performance.

Its a pulse with a duty cycle... push the envelope !!!

H^) harry


>From: "Theo" <t.hogers at home.nl>
>To: <avrchat at dvanhorn.org>, <synth-diy at node12b53.a2000.nl>
>Subject: [sdiy] Re: [avrchat] : Led-matrix (slightly off topic)
>Date: Wed, 4 Apr 2001 05:35:26 +0200
>
>Yes your right.
>And this is also the problem with the project Jaan is talking about.
>I'm not engineering this project but at times I try to help out.
>However in this case I don't have a clue either.
>
>Most small signal transistors have 1A peak listed in the absolute max.
>ratings.
>However the question is:
>Is this 1A maximum indeed the number we have to count with?
>And if not how do we calculate the maximum pulse load?
>
>We aim for 7mA steady state equivalent.
>The scan rate is 1Khz , the dutycycle is 1/16 and there are 32 leds per
>column,
>therefore we have a pulse load of 62.5 uSec @ 3.6A.
>The steady state equivalent through the transistor is only 220mA.
>Are there small signal transistors that can handle this pulse load?
>
>One solution could be to have two or more transistors per column.
>e.g. Two transistors each driving 16 leds, in this case the pulse load 
>drops
>to 1.8A.
>Using 4 transistors the pulse load would be only 900mA, however in this 
>case
>we need 64 transistors!
>A solution with less resistors would be appreciated ;)
>Someone out there must have faced this problem before us!
>
>Explanations from math-heads are also welcome.
>Thanks in advance :)
>
>Theo
>
>(message cross-posted to synth DIY, this is a midi project so we are on
>topic)
>
>
>------------------
>----- Original Message -----
>From: Karl Quies <quies1345 at home.com>
>To: <avrchat at dvanhorn.org>
>Sent: Wednesday, April 04, 2001 3:55 AM
>Subject: Re: [avrchat] : Led-matrix (slightly off topic)
>
>
> > Help me if I'm wrong here.
> > With 3  LS138 +transistors as column drivers with one column
> > energized at a time, you have  a duty cycle of  1 / 24.
> > Now you want to save on amperage and energize only 8 rows at a time
> > instead of 24 rows.
> > That would put the duty factor at  1 / 72,  you would need a peak drive
> > current of 1.44 amp to equal your 20ma steady state current.
> > Karl
> >
> > ----- Original Message -----
> > From: "Paul Maddox" <Paul.Maddox at wavesynth.com>
> > To: <avrchat at dvanhorn.org>
> > Sent: Tuesday, April 03, 2001 6:38 PM
> > Subject: Re: [avrchat] : Led-matrix (slightly off topic)
> >
> >
> > > Jaan,
> > >
> > > >Sorry for the off topic question, but I really need some help!
> > > >In the project I'm working on, I have to drive 512 leds. (I'm using a
>8515
> > > as MCU)
> > > >I was planning on using a 32 * 16 matrix, with four uln2803 drivers
>which
> > > constantly drive the 32 lines.
> > >
> > > Kewl Iwas going to do 24by24 for a project...
> > >
> > > I was going to cheat though :-)
> > >
> > > have 3 74LS138's driving transistors for the columns, selecting only 
>ONE
> > > column at a time.
> > > and ONLY supply 8 rows at a time, this way the maximum number of LEDs
>that
> > > are on at any ONE
> > > point in time is 8... 8*20mA is only 160mA...
> > >
> > > If you refresh them fast enough you wouldn't see any flicker (in fact
>you
> > > could use the refresh rate to adjust the brightness of the display).
> > >
> > > Paul Maddox
> > > _______________________________________
> > > Wavesynth home page;-
> > >     Http://www.wavesynth.com
> > > Modulus synthesizers home page;-
> > >     Http://www.wavesynth.com/modulus
> > >
> > >
> > > ------------------------
> > >
> > > Be sure to visit www.dvanhorn.org, and www.avrfreaks.org
> > >
> > > To Unsubscribe, send a message to majordomo at dvanhorn.org with
>"unsubscribe"
> > as the message BODY.
> >
> > ------------------------
> >
> > Be sure to visit www.dvanhorn.org, and www.avrfreaks.org
> >
> > To Unsubscribe, send a message to majordomo at dvanhorn.org with
>"unsubscribe" as the message BODY.
>

_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com