SCOUG-HELP Mailing List Archives

<< Previous Message <<

Date:	Mon, 17 Jun 2002 21:36:19 PST7
From:	Peter Skye <pskye@peterskye.com >
Reply-To:	scoug-help@scoug.com
To:	scoug-help@scoug.com
Subject:	SCOUG-Help: Fwd: SCOUG-General: Re: after meeting idea

Content Type: text/plain

=====================================================
If you are responding to someone asking for help who
may not be a member of this list, be sure to use the
REPLY TO ALL feature of your email program.
=====================================================

Michael Rakijas wrote:
>
> The surges aren't so much a problem as switch "bounce" and
> I have heard (admittedly, third or fourth hand) that there
> can be some problems here. Mechanical switches do have a
> nasty habit of changing state somewhat indeterminately
> (bouncing between high and low states - hence the term
> bounce)

Yes. The term "bounce" actually *does* mean bounce -- the mechanical
metal leaves/springs/wipers inside the switch or relay actually do
"bounce" when they get thrown from one position to another, or
alternately the wiping action of one metal contact against another can
intermittently cause an "interruption" in the connection (this used to
drive young TTL designers nuts when they tried to trigger their test
circuits mechanically).

More important is the "changing state" you mentioned -- if you *do*
throw the switch while the voltage between a switch contact and ground
isn't zero you *might* get an inductive surge voltage due to the field
collapsing, which can cause a high-enough voltage (E=IR with infinite R)
to zap a transistor.

But you're not supposed to switch between the printers while they're
running because then the setup codes go to one printer and the data goes
to the other. I don't remember if any of the data lines on a parallel
port are kept high (non-zero volts) when the port isn't being used.
I'll have to get out my breakout box and check.

> This garbage [caused by switch bounce] usually won't
> affect say, dot matrix printers but could affect laser
> printers that use the bidirectional communications
> extensively.

Hmm. The bouncing should only occur during the time the switch is
actually being thrown, and should have an effect (it should "bounce") in
both directions.

I used three Transtar 130's (Silver Reed daisy wheels) and a DeskJet 500
ink jet printer for a while but don't remember if they did any return
signaling. Right now I'm running five LaserJet II's and, again, don't
know if they do any return signaling.

One other possibility on the cause of the garbage: The Centronics
(38-pins I think) parallel interface might use signaling lines in the
25-pin connection cables that *aren't* wired through some of the
mechanical switch boxes (which use 25-pin connectors), so if a cheap ink
jet printer doesn't use those lines it will continue to work just fine
but if an expensive laserjet printer wants to use those lines and
they're not connected inside the box, well it ain't gonna work.

I just opened up my spare 4-printer mechanical switch box and found that
the switch is a 9-wafer 27-pole switch, and all of the 25 lines in the
printer cable are connected between the computer and the printer. At
least with this box.

> Using an oscilloscope with a single trace trigger, you can
> see this happen with the flick of the switch which gives
> some credence to the idea.

Yes. I'm glad *somebody* around here uses an oscilloscope!

> HP for a while threatened to void warranties of those who
> used mechanical switches

If you're going to throw the switch while it's carrying data then you
ought to at least solder some zeners onto the data lines. Some of the
mechanical switch boxes might already have them (zeners are cheap).
There aren't any in the spare printer switch box I just opened up.

Alternately, don't switch it while it's printing. :)

> If you want certainty (and network printing is out), go
> with the electronic switch. The switches there are
> debounced to avoid this problem.

Good advice, thanks Rocky. (The debouncing is typically done with a
"one-shot" circuit on each data line which makes the switch look like it
is electronically closed even while it's still bouncing.)

> > - Electron Pete
>
> -Electric Rocky

- Electron, The Duck :>>

=====================================================

To unsubscribe from this list, send an email message
to "steward@scoug.com". In the body of the message,
put the command "unsubscribe scoug-help".

For problems, contact the list owner at
"rollin@scoug.com".

=====================================================

Return to [ 17 | June | 2002 ]

SCOUG, Warp Expo West, and Warpfest are trademarks of the Southern California OS/2 User Group. OS/2, Workplace Shell, and IBM are registered trademarks of International Business Machines Corporation. All other trademarks remain the property of their respective owners.