C3088 Cam Module 5.0 volt to 3.3V DIO

Hi All, I picked up a C3088 Color Camera Module (aka. the Cam in the CMU and AVR Cameras). I am interested in connecting the 16-bit outputs as well
as the clock, and image framing lines to a 3.3 volt (not 5.0 volt tolerant NetBurner) cpu. Readiing the Omnivision 6620 the actual image sensor on the C3088) manual, I notice that there is with the Omnivision 6620 a means to power the output pins with a seperate power source at 3.3 volts - via the DOGND and DODDD pins on the 6620.
I see 5 unused through-hole solder pads on the C3088 module that correspond to the same name lines as the image sensor - DGND, DVDD, DOVDD, AGND, AVDD .... My question to the CMU and AVR ( BTW, I have an AVR CAM on a hexpod- works great) cam experts if anyone has played with these unused C3088 module lines to add an additional 3.3 volt rail, to allow the output pins to output only 3.3volts ?
My current solution is a Max 7000 CPLD chip (happen to have had a few on hand) which I have tied the Digital Output lines from the C3088 module to 3.3 volts - but it is kludgy, heavy, and sucks alot of power, and looks kindof silly - this big 'ole 88 pin PLC socket..... 7400LVC logic would take either lots of chips or a single level transciever (and surface experience) both options i do not like too much....
Any ideas ? Regards, John LeClair
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
snipped-for-privacy@yahoo.com wrote:

Why don't you first look at how they do this on the cmu and avr cams. Also, if you don't want to use chips, you might be able to get by with simple voltage dividers for going from 5v --> 3.3v in the one direction, and simply running 3.3v into the 5v cpu input pins in the other direction. Most 5v cpus will read this level ok, I think.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
I borrowed my AVR Cam from my hexpod and made some meashrements - measurements were almost at 5 volts on the Cam output pins (well more than enough to nuke my NetBurner board.... and yes, i was driving the CAMs i2c pins with only 3.3 volts no problem in the 3.3 --> 5.0 volt direction - verifying funcationality with an old green 75 ohm monitor. The CMU cam is the same - the SB-28 is also running at 5 volts. I would use resister dividers for perhaps a few pins running in the khertz range - but the CAM pixel clock is pushing data out the data pins at 17.7 MHZ - a resister diviider wouldn't work.
That is why i am trying to see if anyone has taken advantage of utilizing the unused camera module pins that run at 3.3 volts for the digital I/O.
dan michaels wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
snipped-for-privacy@yahoo.com wrote:

17.7-mhz transfer rate, I don't think so. That's the oscillator frequency, and the transfer rate to the cpu is probably determined by the cpu, not the cam, and is much slower. I'm sure the AVR can't accept data that fast, although the scenix chip in the cmucam can probably xfer that fast.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Actually, 17.7 MHz is the frequency of the crystal driving the camera, and that signal is output to one of the pins on the OV6620 to provide the required oscillator for the mega8 microcontroller. So both the camera and the mega8 are running off of the same synchronized 17.7 MHz crystal. And yes, the mega8 _is_ sampling data at this rate (the "in" instruction on AVRs run in a single clock cycle). The OV6620 doesn't provide the pixel data at the full 17.7 MHz clock rate. It is, minimally, 17.7 MHz / 4, and the pixel rate can be divided down even further through setting a register (the "Clock Rate Control" register, 0x11) in the camera.
Regarding the use of 5V to 3.3V logic: I haven't experimented with trying to sample the OV6620's 5V outputs with 3.3V logic. I had played around with hooking up the OV6620 to an FPGA previously (Xilinx Spartan family), and there were a couple of older versions that tolerated 5V inputs, but the more recent ones don't IIRC. I'd play it safe and drop down one of the commone level conversion chips.
John O. www.jrobot.net
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
snipped-for-privacy@gmail.com wrote:

So, sounds like the maximum practical transfer rate is about 4.4-Mhz. Or less if desired. Should be easy to deal with regarding interface circuitry.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Thanks John, I finally decided to use some 74HC125s for the level converters although i am still tinkering with them - having a brain cramp on connecting them. For me it would have been easier to use an Altera MAX7128 - but the last batch i bought an ebay were kaput....
I spent hours trying to figure out how the C3088 could work with 3.3 volts as the OV6620 states. I removed the solderr jumper that leads to the digital IO power pin, etc.... I think the probelm is at 3.3 volts, the i2c was not pulling high enough on idle.
Thanks all.
snipped-for-privacy@gmail.com wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
snipped-for-privacy@yahoo.com wrote:

Why not? Does the capacitance in the circuit create too long of a rise time at that frequency with the extra resistance? How about clamping the voltage with something like a diode connected to a 3V source? Would that put too much load on the 5V outputs?
--
Curt Welch http://CurtWelch.Com /
snipped-for-privacy@kcwc.com http://NewsReader.Com /
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.