[simpits-tech] ----> To Frank....... EPIC

Jean-Yves SURFIN' simpits-tech@simpits.org
Thu, 31 Jul 2003 22:15:02 +0200


Hi Mark,

Thanks for these comments.
They are welcome.



Really Thanks, 
Jean-Yves





Jean-Yves «Surfin'» Jouanny
101e ECV-FAF / France1 : Cougar Knights

101e ECV-QAF http://www.quebecairforce.com
101e ECV- FAF http://perso.wanadoo.fr/franceairforce



-----Message d'origine-----
De : simpits-tech-admin@simpits.org
[mailto:simpits-tech-admin@simpits.org] De la part de Mark Doran
Envoyé : mardi 1 juillet 2003 21:56
À : simpits-tech@simpits.org
Objet : RE: [simpits-tech] ----> To Frank....... EPIC

I know this wasn’t addressed to me so forgive me if I’m butting in but
in case it helps you, we have been using one other EPIC module that
might be of use to you and have another inbound that should be very
useful if it works as advertised.  Couple of other "lessons" we learned
that might save you some time also.

Rotary encoder support: http://www.avsim.com/bluesideup/ the module
available here is considerably cheaper than the RnR version and works
very effectively as far as we can tell.

Secondly we are expecting the arrival of an EPIC servo controller module
from MicroCockpit shortly.  Controlling servos will be of interest to
you if you plan to use Marv’s most excellent replica gauges for example.
Most people use phidgets and VB code for this part of the problem but
there is value in keeping the number of source code bases you have to
manage and synchronize for the project to a minimum and the EPIC servo
controller reduces that variable set by one, for us anyway.

EMDA boards http://www.em.ca/~cockpits/main.htm are an alternative to
the ABA/BTN64 boards for wiring integration.  Some wiring integration
solution of this sort is practically compulsory since otherwise you are
likely to be swallowed hole by the task of keeping track of all the
conductors required ;-)  One note on the EMDA though: we have seen some
quality problems where tracks on the boards have lifted and separated by
the joint pad/holes.  This is a nasty problem to trouble shoot but easy
to fix (small “blue wire” update).  Our wiring guy is a professional at
this soldering lark and still we see this issue sometimes so caveat
emptor.

One other EPIC trick to think about…  It’s not well known but you can
get a “Second expansion module”…this extends the input and output row
count that you can address by practically double.  To do this, buy a
“First expansion module” and order with it the GAL part that converts it
to the “Second expansion module” – this GAL thing is a socketed
component on the board that you can pop out and replace…the “special”
one give you access to input module addresses 3, 4 and 5 in the total
EPIC name space.

Why would you care??  Well if 304 switch inputs is enough for your
project maybe you don’t.  But we have found that “bit packing” the
switches by placing single switch wires from separate physical flight
hardware panels together in EPIC name space module/row addresses
complicates the wiring loom immensely (well, somewhat anyway).

By having ready access to more module addresses to wire to you can
afford to be more profligate with the namespace.  For example, if you’ve
got a panel that has six switches on it you can allocate all of them to
one row in one of the modules and ignore the two bits left over in that
row.  On the flip side of that, if in the case that you are tight on
MRBs to allocate and you need to have switches on separate physical
panels connected to the same EPIC row, now you have to run the row
select wire/conductor to more than one physical panel.

Of course you don't have to use this technique...it's a trade off: to us
the extra $$ for the additional EPIC expansion module more than offsets
the time it would take to do the extra wiring work.  Something to think
about perhaps though if you are at the planning stage.  The single most
labor intensive thing we have seen so far in our project work is making
up the wiring harnesses to hook up the panels.  So we definitely have
taken steps to cut down on that "time" expense and make the wiring "more
reliable" and modular to simplify downstream maintenance (it's going to
break, it's just a question of when! ;-).

Oh, and the other real time-saver -- we were able to find good quality
screw-type terminal blocks with exactly the right pin separation pitch
to match the EMDA cards...these were soldered to the card and then the
databits connect right up without a solder connection to the wiring
harness ends.  This was very handy when we discovered (ahem) issues with
the wiring plan and needed to move things about a bit.

Hope that may be of some help to you.  Good luck!!

Cheers,
 
Mark.