ComMedia C328 Colour Camera (JPEG UART)

I have done some work with these cameras and they are excellent, but I encountered reset problems which I thought I shouls pass on.

On power on, the chip is reset with a simple R/C to ground (10K 0.1uF. The value of the R is a little too small and sometimes these devices do not reset properly.

To solve the problem, the R or C can be increased. I soldered a 5 pin 0.1 inch header to the camera pins. I then soldered a wire ended 0.47uF across the existing SMD 0.1uF and brought the reset line to the extra header pin.

This then solved the reset problem and also provided a reset line, which I have found is very useful.

I will try to post some pics.

rgds

mb

Here’s an interesting issue with these cameras I haven’t figured out yet. Maybe it has to do with the reset. I created a GUI app to get streaming photos on a PC.

The app sends the sync command until the camera acks at 115K baud. All works well. When trying the same thing from a stand alone TI processor (CC1110, but doesn’t matter), the camera never acknowledges at 115K baud.

Using a dual trace o’scope, the waveforms are identical between both sources. I tried to recreate all conditions that the PC generates through the USB uart, but the camera will only acknowledge when the baud rate is dropped to around 7K baud. Wierd huh?

I will try the reset solution.

Ron

Hi Ronny

Some things I have learnt (you may know already but it may help).

To be sure I wired in my reset (took a bloody long time to find that little trick - grrr!!). If you dont have it, try powering on and off - sometimes its ok, sometimes it not (on mine anyhow).

I guess you give the camera plenty of ‘sync tries’ - I found it needs about 20 to 30.

The camera draws 45mA after it has sync’ed but if you dont talk quickly, it will go into sleep mode and needs another sync cycle.

I have found downloading at 115Kb works ok from a PIC 18F458.

Wierd … you get no reply data at all?

rgds

mb

I checked my code and it runs at 57Kb not 115Kb.

The micro was running at 20MHz and as it was not an interrupt driven test harness, this was all I could get out of it.

rgds

mb