PDA

View Full Version : ServoPod Problems


Cyris
07-12-06, 04:53 PM
Hello,

I'm no longer able to execute commands or send files from NMITerm to my ServoPod. I'm able to have "IsoMax V0.82" appear in the terminal every time I reset the ServoPod, so there is some communication. However, I am unable to type anything into the terminal, and files will not be sent.

This began after I connected quadrature encoder outputs from a motor / motor controller to the Phase A 0 and Phase B 0 pins of J10. Could this be related in any way?

I'd like to reflash ServoPod, but I don't have the required .s file. Any tips on where to find it?

Thanks much.

RMDumse
07-13-06, 12:19 PM
The problem doesn't sound like an encoder problem, although it might be. It sounds more like a bad autostart. By jumpering PE4 to ground you can force a cold reset and bypass the autostart (if any), and then type SCRUB in, and possibly recover the software.

If that doesn't work (high probability it will, though) the kernel can be reloaded using the JTAG cable, either if you have or buy one and get the kernel from us, or you can return the board and we'll reflash it, just for shipping charges.

Try the grounding of PE4 during a reset first, then let us know.

Cyris
07-13-06, 12:52 PM
The manual has the location of PE2 and PE3, but I'm unable to find PE4. Any thoughts?

RMDumse
07-13-06, 01:29 PM
I opened the scematic from the ServoPod(TM) page.

PE4 is in the I/O SPI connector J3 - pin 4, and the pin just above it on the same connector, pin 2, is gnd. So it is very easy to slip a jumper over the two pins when reseting. Even shorting them temporarily with a paper clip or a screwdriver will do while reseting will do.

Cyris
07-13-06, 06:05 PM
Unfortunately that didn't work. Do you by any chance know where I can find the .s file needed to reflash?

RMDumse
07-13-06, 07:19 PM
Pls contact Nmi Sales. We need a nondisclosure and we'll send it out.

Cyris
07-14-06, 06:28 PM
Thanks for the help. I got a copy of the S-Record file. However when I try to re-flash, flash_over_jtag complains about not finding the I/O driver. It also mentions that the driver should be in the same folder. I tried this on a couple computers and got the same error message. Is this referring to the driver for the parallel port? Should I just track down a copy of that and stick it in the same directory?

Dave
07-17-06, 11:21 AM
The readme.txt file included in the files emailed has a corrective procedure. In the portion marked "Note for XP/NT/2000 Users" it states that the "driver_remove.exe" may need to be used, as some port driver has apparently already been loaded, and needs to be cleared for the current version of zlportio to be loaded in its place. This should allow the proper operation of flash_over_jtag in these systems.

Cyris
07-17-06, 02:07 PM
I successfully reflashed the ServoPod, but the problem is still present. Any ideas?

RMDumse
07-17-06, 02:25 PM
Very strange... Well, we have the board talking, because you can see the sign on message. But we don't know if the board is receiving anything, because there is no echo. So an important question, do we know if the PC (or whatever you're talking to the board with) still works? Do you have a scope? Can you put a scope on the output of the PC? Do you have a loop back cable or connector you could verify its operation with? Or have you tried using a different PC?

Have you tried restarting the communications software, could it be hung up? (I have had that happen to me before like this. Closing and re-opening cause the sending side to work again.)

Have you removed the quadrature connections, as in, could it be a ground loop problem, with the ServoPod(TM) stuck in the middle between two devices with incompatible ground levels?

Can you give us any other clues so we can help better?

RMDumse
07-17-06, 02:28 PM
Ah, another thought. Most encoders take power to run their internal LEDs. Could it be the encoders are taking enoug power they overload the ServoPod(TM)'s regulators?

Generally, you'd see a brown out condition with this case, where the reset message would appear multiple times, usually with a truncation and repeat as power was fading in and out. So it might not be appropriate, but I wanted to mention it as another idea.