RE: [Gumstix-users] RoboStix + CMUCam2 ttl serial help

[House, B. <Brandi.House@Trinity.edu>]

Sean,
=20
Sorry about that last reply... after further testing, I realized that =
were still having inconsistent results. =20
=20
So apparently on a power reset the CMUcam2 always tries to send the =
version of the firmware before responding to any other commands, so we =
found that unless the Robostix intercepts this information first, our =
first command always gets an NCK from the camera.  However, if you don't =
want to intercept the information on the Robostix, try sending a =
carriage return first ('\r') first (which should get an NCK), and then =
send the normal command (which from my testing should get an ACK).  We =
are currently set up at 38400 baud, three UART pins connected (TX, RX, =
and GND), and the camera power switch is in the On position.  We still =
get a little current from the TX line which causes the camera's power =
LED to flicker upon occasion, but for the moment, we are not having any =
other problems from it. =20
=20
I hope this is helpful, and I apologize for any confusion caused by my =
previous response.
=20
Brandi

________________________________

From: gum...@li... on behalf of House, =
Brandi
Sent: Tue 2/28/2006 6:47 PM
To: gum...@li...
Subject: RE: [Gumstix-users] RoboStix + CMUCam2 ttl serial help


Sean,
=20
We are using the CMUcam2+ to the robostix, but I think we experienced =
very similar issues.  When we hook up all four lines of the UARTs, the =
camera is powered by the robostix.  We are still not sure how to correct =
this problem, so if you discover a solution, we'd be glad to hear it.  =
However, we have had successful communication to the CMUcam2 at this =
point.  With all 4 lines connected (TX to RX, etc), and with the power =
switch in the 'OFF' position on the camera, we are able to send a 'GV' =
from the Robostix, and receive the 'CMUcam2 v.1 etc.' response in a =
buffer.  I do not understand the power conflict between the two, but =
this is the set-up that has worked for us pretty consistently, so feel =
free to give it a try.  I assume that the robostix is just transmitting =
enough amps through the serial lines to allow the camera to function =
without an external supply.  Interesting and unexpected for us... so if =
this can be corrected, I am quite open to suggestions.
=20
Good luck,
=20
Brandi
=20

________________________________

From: gum...@li... on behalf of Sean Luke
Sent: Tue 2/28/2006 6:02 PM
To: gum...@li...
Subject: [Gumstix-users] RoboStix + CMUCam2 ttl serial help



We are trying, and failing, to get the CMUCam2 to work with the=20
RoboStix over ttyS2 using TTL.  Has anyone had any experience here?

1. If we connect to the CMUCam2 over TTL using a BrainStem TTL=20
converter, it works peachy.

2. If we connect to the RoboStix ttyS2 over TTL using the Brainstem=20
TTL converter, it works peachy.

3. If we connect both to each other using two TTL converters and a=20
null modem RS232 cable, the gumstix receives the CMUCam2's response=20
but appears to be sending something which is constantly making the=20
CMUCam2 respond with a "Huh" statement (":NCK").  They share a=20
ground.  We're trying to figure out what hint this may be providing=20
us for #5 below.

4. If we connect the CMUCam2's transmit to the the Robostix's receive=20
via TTL, sharing a ground, the gumstix can hear the CMUCam2 fine.

5. If we connect the Robostix's transmit to the CMUCam2's receive,=20
(sharing a common ground) this is where things get weird:
        - The CMUCam2 receives *power* through the transmit line from =
the=20
Robostix.  Enough to weakly turn it on and freak it out.
        - The CMUCam2 does not receive any transmissions over the line
        - If you power the CMUCam2 off and on, the red "all good" light =
in=20
the front doesn't turn back on.

We've verified that what we think is the transmit line is indeed=20
*not* the V_BATT line.  It appears that the robostix is sending amps=20
into the cmucam2 through transmit.  TTL PWR and GND are both=20
unplugged on both devices.  They're sharing ground through a servo=20
wire on the robostix, which looks fine to me (same thing happens if=20
they share TTL GND).  Do either of them have to do something funky=20
with PWR?  My electronics experience is too weak to guess what's=20
happening.

Any guesses would be welcome.  The CMUCam2 is a pretty battle-tested=20
device, so I'd be really surprised if it was the problem.

Sean


-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting =
language
that extends applications into web and mobile media. Attend the live =
webcast
and join the prime developer group breaking into this new coding =
territory!
http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D110944&bid=3D241720&dat=3D=
121642
_______________________________________________
gumstix-users mailing list
gum...@li...
https://lists.sourceforge.net/lists/listinfo/gumstix-users