Redlake MotionScope 1000S


I purchased this camera off eBay for a very large sum of $21(though that was later contrasted by shipping costs), and it seemed like quite a good deal.
This is a camera made by the now defunct/merged Redlake Imaging Corp/Redlake MASD.
The model number is 1000S. It is the slowest of the range, capable of shooting only 1000 frames per second. It appears that there was also an 8000S model.
It was made in 1998, and from what little information on it I can find they were produced around 1991 to 2005. There were also some rebrands, like the Olympus Encore cameras which was pretty much identical, except for being all-plastic on the front. The prices ranged from around $10 000 to an unknown value.
I might be wrong on that, but there isn't any relevant info on these systems.
The front of the controller

Ports on the side

The imager from the side(with lens similar to the kit one)

Imager, front and back

The guts

When I opened the controller up I was highly surprised to see that inside was not a custom-made board for this product, but a(Redlake proprietary) PCI frame grabber without the PLX PCI 9080 controller. Instead, all the connectors are there for the video output, front panel, etc.
These images are clickable.

Left side of the main PCB

Right side

The IC above the socketed flash/ROM(haven't checked the model/make) marked S1000 is the Motorola MC68HC11E0CFN3 8-bit uC, which seems to be the main processor. There are two Altera FLEX EPF10K50RC240-4 FPGAs, with the little EPC1PC8(again by Altera) 1Mbit 8MHz FPGA configuration Flash chips. Now let's look at the imager. These images are also clickable.
From the top

Closeup of the sensor

The sensor is a Texas Instruments TC237(or 236).

Here's what I found after opening it the first time.

After finding this, I was quite disappointed, and I knew why the seller said he couldn't get an image to the screen. This sensor is driven by the application-specific Texas Instruments TMC57253, and it's been known to fail. This was a bad sign.
I tried finding the schematics or service manual online, but there was absolutely none available. Any phone number of the now-defunct Redlake has been either suspended or disconnected, and I spent a whole lot of money chasing red herrings.(damn you international call pricing)
But then I found a speck of hope - after talking to a lot of nice people, I found a person(whose name should remain anonymous for various reasons) who has worked as someone important in the company in the years when these cameras were made. I contacted him, and he was nice enough to provide me with the schematics for a similar imager head. I will not be posting them publicly, but feel free to contact me and I'll share a few parts if you really need them/have a similar unit. Checking it's output, I found that all CCD drive signaling was stuck high, whilst the TTL part of the chip was functioning correctly.
Now, remember what I said in the very beginning? The 1000S model number corresponds with the frame rate. Now scroll up and look at the MC68HC11 firmware storage. There is a sticker on it saying S1000(or 1000S). After I get this system working, I may or may not be reverse-engineering the firmware to upgrade to the 8000S model.
I bought a replacement for the TMC57253, which was very expensive. I got it a few months later, but replacing the TMC57253 driver did absolutely nothing. I also bought a very similar head, again on eBay, but hooking it up got me a very, very weird image, but it was definetly not image data. The other head works(apparently), but I am fairly certain they are incompatible.

As of May 2016, this project is on indefinite hiatus due to my acquisition of the NAC Memrecam fx K5 camera.
To be continued...