<< return to Pixycam.com

Problem with SPI with SS

Pixy Software
#1

So I’m using two Pixy’s with the TI CC3200 launchpad. When I hook up one Pixy cam and use the SPI with Slave Select, it works fine. However, when I add a second camera, the moment I connect the second camera to either the SPI Clock OR one or both of MOSI and MISO, it no longer functions correctly. The data I get into the microcontroller is incorrect and seems to in some cases be offsetted from my expected values (e.g. A255 or D255 instead of AA55). If I disconnect the SPI Clock from the second camera, I can get data from the first camera again, or if I disconnect both MOSI and MISO from the second camera I get valid data from the first pixy again.

I’m asking here and I’ll also ask on the TI forums, but any help would be appreciated!

#2

Hello Christopher,
This is a dumb question, but you have separate slave select signals going to each Pixy, correct?

Edward

#3

Not a stupid question at all! Need to know everything when diagnosing and all.

Yes, and I’ve verified that they both are high when the CS is disabled and low when it’s not. I’m operating at 3.3v logic, but Pixy has worked with that before now.

#4

Hello Christopher,
I can’t think of any easy tests or suggestions. It sounds like maybe a good way to debug is to look on the scope. If you get see a simple transmit request from your processor to Pixy on the scope (working) and then compare to when two Pixys are connected (not working), you can probably shed more light on the issue.

Edward

#5

I just did a test with a Tiva C and it worked fine. I personally believe it’s simply the microcontroller I’m currently using, CC3200, as the libraries for SPI with it are not very well documented. I’ve posted on TI’s forums as well in hopes of finding a solution.

I’ll try to view the actual signals in a scope on Thursday, though I’m not accustom to the instruments in my school’s lab so I’m not sure I’ll find anything there.