www.coridium.us/forum

Hi,

I'm working on my first ARMWeb project, and after working with it for a
couple of months without an issue, I've run into a weird problem.
Hopefully someone here can shed some light on it...

I thought the system was running fine on the bench, but when I installed
it the problem started. Quick summary: whenever the program sets two
outputs high to drive a couple of relay boards, I get:

EINT1 341c (or 3418, 3414)
ERROR Expected a constant (this repeats a bunch of times, and then..)
Data Abort 250

The board is running 7.36 firmware, and the application is pretty
simple; check two SPI temperature sensors (on IO 12 and 13), and based
on the result start or stop two fans (driven via 5V relay boards) using
IO 17 and 31. There is a wait(2000) between starting Fan 1 and Fan 2,
and then a wait(60000) before starting the whole process again.

The failure usually occurs right after starting Fan 2. Sometimes the
system will run for hours before it decides to switch on the fans, and
then fail. Weirdest thing is that a couple of times it has actually
switched on the fans and run for an hour or so first..

The only other background info that might be relevant is the power
supply. I have a 5V regulator feeding the relay boards, and the combined
current draw was too great for the PS supplied with the ARMWeb. I am
using an HP bench supply set to 7.5V, which is about the same as the
original, and feeding both the ARMWeb and the regulator from that.

Thanks in advance for any help,

Peter

>
> EINT1 341c (or 3418, 3414)
> ERROR Expected a constant (this repeats a bunch of times, and then..)
> Data Abort 250
>
> The board is running 7.36 firmware, and the application is pretty

EINT1 is enabled by the ARMweb firmware as a debugging aid for Coridium. It
allows us to interrupt the software and see where it was.

This line is driven by the USB dongle, and has a pullup on the ARMweb.

What you are seeing implies to me that that line is getting pulled low.

As you say this is coincident with a motor turn on, it seems to me its a power
supply issue. Can you run the motors off a seperate supply?

Motors look like a dead short when they start up.

Bruce,

Thanks for the info, and my apologies for taking so long to respond. I've spent
the last couple of weeks working on the PS angle, and have made some progress.
I was originally running the ARMWeb directly off the incoming supply, with a
regulator in parallel that was feeding the relay boards. I have changed the PS
completely, with the incoming DC of about 10V feeding two regulators, one for
the ARMWeb and the other for the relay boards. I thought this would remove any
likelihood of supply current spikes pulling down the ARMWeb PS.

The situation has improved a lot, but I've gone from problems about 50% of the
time when the relays are turned on or off to failing once or twice a day.

BTW, I did separate the AC line completely to minimize interactions as much as
possible. In other words, the controller PS is not fed from the incoming 115V
that the relays are switching.

I have diodes around the relay coils to avoid turn-on and turn-off spikes, but I
may need to increase the PS capacitance and check the ground wiring to avoid
interactions between the regulators and the ARMWeb.

Thanks for the help,

Peter

Peter :

Forgive me as I have been following your problem only at " arms length " ; but
this sound suspiciously like one that I had with controlling high volltage motor
switch gear from a computerized controller at an ICE PLANT of on of my friends .
We had to try and control a variety of 230 volt single and 3 phase motors with
horsepowers ranging from 1/3 thru 25 and ran into similar sounding problems . We
suspected everything from the wiring down to the interfacing and power supplies
. FILTERING of power was also suspect . Our local utility "promised" that we had
optimal power at the meter , but we still ended up scratching our heads . Our
problem "looked like" a BROWN OUT condition although the local grid was lightly
loaded .

As the solution presented itself ; our MIX of motor types was the actual
problem . We had a mix of 3 phase , shaded pole , ...............

The old formula regarding the Xc, Xl , phase shift ( Theta ) was the cause our
line fed DC voltage regulated power supplies effectively " coming out of
regulation " ( from 100 % down to 90 % ) due due motor start-up loading from the
panel . A few 73 MFD 240 volt caps on a couple of motors and some current
sensing " hard start " relays on the refrigeration compressors cured the problem
. The controller was NEVER the problem as it had
OPTO-Isolators on all of the switched outputs .

Perhaps send me an E-mail of your motor types and loading and maybe I could
help .

Please forgive me if I am ; yet again , " off base " on this one .

Regards , Rick ( Don't Give Up ! )