SparkFun Forums

[comadrejo]

Dear aeo087

I just bought one of these cameras, could anyone share with me a schematic of the breakout board designed? I also keep hearing about undocumented registers, what should I know about those?

If you look deeply this other topic:

"TCM8240MD connector and example"

You will find a lot of info about TCM8240MD and TCM8230MD.

I think there is the schematic of the TCM8240MD interfacing.
and "almost" or "comming soon" the schematic of the TCM8230MD interfacing.

If any caritative user of this forum could add to this topic a tested shematic
of the TCM8230MD it would help sparkfun to sell many more cams.

Especially with a schematic made thinking in an open variety of
possible Microcontrolers to interface with the cam (not for a particular IC).

I hope some helpfull answer post.

[KreAture]

I think something should be said here.

"The cameras are horrible, and poorly documented.
There are other cameras with interfaces better suited for microcontrollers and one should not waste time on these."

There. That felt good.

[mwoo769]

hi there,

does someone have a footprint for the TCM8230MD that I can go off, i need to be able to solder on the sensor to a breakout board using a soldering iron and the pads underneath the sensor would be a bit too hard to get too.

I have tried to find any footprint library file for the TCM8230MD from sparkfun and other interwebs source, but to no avail.

Regards,
Michael

[shimniok]

Check the eagle files linked on this blog post that I ran across.

http://sigalrm.blogspot.com/2011/03/tcm ... akout.html

Hope this helps,

Michael

[mwoo769]

Hello all,

I just found this gem of an app note, now its not exactly exhaustive, but it does go a long way into explaining the inner workings of the camera.

enjoy

http://go900.angellfear.ru/old-files/nv ... 040127.pdf

also has anyone tried to get the camera operating at 2.5V for IOVDD and PVDD, because I want to interface a Cyclone FPGA to it, but the IO bank for the chip only operates at 2.5 or 3.0V, and I would prefer not to use 3V supply on the camera, as it would seem easier to give it the 2.5V as the camera datasheet suggests, however the sheet doesnt show what settings need to be changed (if any) and the application note doesnt mention the IOvoltage either,

TL;DR anyone used 2.5V IOVDD and PVDD?

[KreAture]

Thanks mwoo!
That really helped! Still need to tune my image settings for better contrast I think.

I'm working on a new driver for atmels new release of trhe ISI driver in mainline linux.
It is supported on their AT91 arm platforms.
I'm hoping to get the new gen2 isi driver to also support the gen1 isi in the avr32ap7000 series (nbgw100 boards and such) too, but it may need a extra level of abstraction... (Wohoo kernel work!)

wohoo.jpg

[doddy1985]

Hey I know it's been a while since people posted.. Anyways here it is.. So @mwoo769; I've built a custom PCB for Altera Cyclone iii (40k LUs/LEs) just for this purpose. Although the board itself isn't designed for TCm8230 sensor, that's the reason I wanted an FPGA board. So the board I've got works on 3.3V CMOS/TTL (you choose in Altera software for the pins). Using CMOS though the CMOS logic will trigger from the toshiba sensor. Basically just the 2.5V logic high is more than the Altera chip Vih and the camera sensor low also works with Altera chip Vil... So grabbing data and clock line from the sensor is fine. If you are interested in a board or collaborating lemme know:

rhoffman @ + uvic dot ca (get rid of the plus sign for email).

~doddy

[KreAture]

My biggest problem with these sensors is the large distance between my controller hardware and the sensor chip.
The driver in the sensor is apparantly very weak so it doesn't take much of a mismatch to get data integrity problems. The reflected signals also interfere in the sensor and deteriorates the image. (I needed 1.5 feet long cable and it's not good...)

[doddy1985]

@ KreAture

That is an interesting issue. My 'cables' are like 3 inches or so, so I haven't noticed signal integrity issues. Are you able to configure the sensor on the I2C lines? My understanding of I2C at least is that they are supposed to allow for long distances between master and slave. Of course if your signals are still weak then I'm sure it'll have the same issue. I'd suggest placing a buffer perhaps midway (just opamp with current gain) so you basically resupply the current to the line. I think that the output from the sensor is something like 20mA - 40 mA? I don't have specs in front of me but if your cable wire gauge is mismatched compared to pins then you'll get some reflection for sure. Maybe also try sizing your wires to something that more closely fits with the pin sizes of the sensor (I know those are tiny MF's). Let us know what you come up with!! It's possible also that a different sensor is in order for any practical application..

Cheers,
~doddy

[KreAture]

No. The I2C is no problem as the signal is low frq.
The data signal especially clock goes out of sync with the data flanks due to crosstalk and due to weak driver.
I am sure using a buffer close to the camera is the right solution and matching this driver and impedances of signals to the cable for a proper terminated transmission.
However, the project did not have time and so we scrapped it. The camera interface on the ngw100 is also too buggy.
Using a proper lvds camera we have no issues, and on a new project we will use usb.