Camera with SONY IMX291, some tinkering

[bojan]

Last year my camera to be used as guider (https://www.ebay.com.au/itm/173047013192?hash=item284a67df48:g: aFcAAOSwtwxfDrmS) stopped working (black frame) so I bought a new one (https://www.aliexpress.com/item/1005001998432509.html?spm=a2g0s.904 2311.0.0.5a734c4dtPwBVO)
It is supposed to be more sensitive, both in visual and IR.

At the moment I am working on enclosure...

[etill]

Looks like a cool project, I went down a similar path to make a cooled planetary camera from what is probably the same board (IMX291 from AliExpress).

I tried a few different heat sinks and peltier devices but eventually found that a small 5v TEC plus that heatsink could get down to near zero even when powered from USB (with very careful current limiting and control though).

The camera itself was a bit flakey so I shelved it, not sure if it had to do with all the heatsink compound on the back of the board to try and get the heatsink to work a bit more effectively, but as long as its not got anything conductive in it, it didn't seem to hurt.

[bojan]

Yes, it is the same, but I removed the extra PCB brim, to make it fit into that enclosure.

[etill]

If only they were on round boards..

I had that camera earmarked for an attempt at debayering, but I might mill an enclosure for it once lockdown is over and I can go back to capral and get some off cut stock for it. I never added temperature regulation, it was just a simple LM317 current limiter but for something USB powered it was surprisingly effective at 400ma (with a powered USB3 hub though). It would get down to between 0c and a few degrees above sitting on my desk after some time.

I have a set of thread mills now too that I'm looking for something to make with to get some practice, so I should be able cut the threads for a 1.25 filter into the nose of it.

[bojan]

Well, it is possible to cut PCB corners (carefully) to internal side of 4 holes, This can bring down the diameter to ~ø37 mm.
However, Jaycar enclosure is good enough.

[bojan]

Working on adapter (made from M42-EOS insert)..
I tried 3D print, but it proved to be hard to cut the thread (M42x1 or M42x0.75) with reasonable accuracy in plastic, so I decided to go with 1/2 off-the-shelf solution.
I mounted it on my modified 8/500 Rubinar.

Because of reduced distance to sensor, the lens is working at ~ 300mm or so (will check tonight if clouds allow).

[JA]

Lovely job Bojan :thumbsup:

You could always do a daytime check for infinity focus on a distant chimney or something.
One thing though: if that lens is an M42 mount from Pentax/Praktica/Zeiss/Zenit/etc, then its Flange focal distance should be around 45.5mm and I have to say that looking at the distance between the lens flange and the sensor in your picture that it looks VERY short. Perhaps even too short to achieve infinity focus. Maybe your lens mods accommodate for that or there is enough latitude in the lens helicoid movement and things go your way. Fingers crossed as they say.:thumbsup:

Best
JA

[bojan]

JA,

Those lenses are modified - I removed focus stop, so they can be focused on sensor even couple mm apart from thread (yes, Rubinar has M42x1 and Samyang has M42x0.75 thread).
Field curvature will be another issue of course.. but for guider not detrimental I hope.



I also plan do make adapter to ø31.8 mm for camera to fit into 1.25" eyepiece... just to enable another possibility.

[AdamJL]

This is a terrific project :)

[bojan]

Plastic pipe seat and Flexiclamp, screwed on wooden bar 20x40mm..
Ready for star test.
The conifers on test pic were ~5 metres away, shot through window (glass needs washing :P)

[xelasnave]

I love seeing your various projects Bojan and they must be inspirational for many folk.

All I wanted to mention is making a camera container or whatever really, and I have done this many times, you can make it out of cardboard or balsa wood and then give it a couple of coats of fibre glass resin...you could add fibre glass but I use black tile grout...I am not sure if it adds to the strength which it doesn't have to really after a couple of coats of resin but it looks great...the other thing is on your last coat cover it with a nice paper .. just office paper and that gives a reasonably smooth surface that usually does not need much sanding...the paper soaks in the resin and then adds to the strength...
Adding paper is invaluable if you use fibre glass matt as it can be a pain to sand...also the paper is not paper anymore but plastic impregnated wood fibre ;) ...to be convinced resin up just three sheets of paper and stick them together and note the strength.. same with balsa wood...resin up some thin stuff and observe how strong it becomes...I know you are into metal work but nothing like having fibre glassing ( or resin reinforced stuff) under your belt.
I have made a pier this way but used form ply.. resin glass and paper it was as strong as steel..or do I believe:P

Alex

[bojan]

Alex,

Thank you for suggestions :-)
Yes, I know about paper - when I was a teenager, I made couple of model boats of wood (rib construction properly done using veneer and 3x3 mm rods). At the end the finish was done by applying very thin paper (we used to call it "Japanese paper") on surface and then spraying with lacquer ( I managed to almost poison myself with thinner because I was doing it in small closed and on-ventilated room).

These days I prefer 3D printing of enclosures.. or off-the-shelf plastic or metal boxes if suitable (they usually are).
Paper is actually amazing material, especially when impregnated with epoxy (or polyester resin) - mechanically very stable and almost indestructible in all weather conditions. Glass matt is of course better.. :-) but personally I don't like to deal with liquid chemicals too much...

[xelasnave]

3d printing ....of course...I should get into it...
Alex

[bojan]

Some results from last night...
There is a lot of noise in the image, maybe I need to ground the enclosure. Or play with settings a bit (everything (Gain, exposure time) here was at maximum)

Calculated from this picture, Rubinar 8/500 works at f/6.7, 355 mm FL.
Star magnitudes are marked in red.
So FOV is ~ø1°.. maybe too smal for guider...

[bojan]

Because Rubinar's "shortened" FL (355 mm) is too long to be practical for finder and/or guider, I will try next with binocular lens (~150 mm).. it is ø40 mm, but maybe OK.
The scope is mounted on adjustable bracket.

[JA]

Certainly could be some interference that grounding might help with, or maybe try some power supply decoupling / filter caps just in case there's noise in the DC, especially if going for use at high gain where everything will be more noticeable. I suppose just check for noise in the DC from the USB(?) out of curiosity.

If you want to borrow a zwo290mc to perhaps help with your project, for comparison, let me know.

Best
JA

[bojan]

JA,

Thank for offer :-)
As to noise.. it is curious because the mob using the similar camera (same sensor, camera is with LAN interface) in GMN project (https://globalmeteornetwork.org/wiki/index.php?title=Main_Page#Optional_ RMS_Software_Installation) says this sensor is capable of detecting mag 5 stars with wide angle lens (ø3 mm?)..

I tried grounding the enclosure, but saw no change.

[JA]

Could it instead be noise on the DC line from the USB DC source (the PC)? Or possibly noise pickup in the USB cable - it doesn't appear that the shield is connected at the camera end of the cable as per the picture in the advert - see below. I know that sometimes shields are not always connected at both ends, but you never know. Maybe the cable doesn't even have a shield?????


Best
JA

[bojan]

Yep...
I have shielded cable, but with different connector (camera uses 2 mm pitch). I have to crimp the new connector and then we shall see :-)

[etill]

I have similar noise and generally poor performance with the boards I got. USB is a differential signal, so electrical noise effects both signals and is not generally a big issue. The noise is coming from the board, but still possibly poor ground routing or design. I have a few very clean well regulate power supplies I've built over the years, but even a simple LM317 circuit configured with a tracking pre-regulator is super clean in this use case (I have two versions of this sensor, one with ethernet, one USB-supposedly-3), and it produced the same result. I've poked around the boards with a scope looking for the serial comms to the sensor to try and see how its configured, but never found it. I went as far as popping the eprom off the ethernet board (I think it was an eprom, I don't recall the exact part number) and managed to recover the firmware, and to my shock it very obviously contained components from another vendors cameras (in fact two vendors names were found in it).

The boards probably aren't well laid out, use cheap components, etc. A case of you get what you pay for I think. The ASI290 was 10x more than I paid for either of these (but also a more expensive sensor).

I didn't try PHD, but it never reached its quoted frame rate in sharpcap and randomly stopped responding so I put it aside for later.

[hamishbarker]

I have a GMN camera running one of these IMX291 boards (but power over ethernet/IP camera basis not USB), they work well for it.

The GMN software captures 250 frames of the video output to stack for the 10 second images for the meteor detection. it keeps just the differential between frames and uses some tricks to get quite decent depth with only a 6mm lens (but the lens is f0.95 so QUITE fast, but of course only a 7mm aperture.).

Is the image you showed just a single frame or stack of multiple video frames?

I don't know if it can output raw images which you probably want for guiding, although PHD can use a USB webcam input. I also have a usb version, could give it a go with PHD and report back.

[bojan]

It was a single frame only.
BTW, GMN is on my to-do list since last year.. :thumbsup:

[bojan]

Finished..

After cutting corners of the PCB, IMX291 camera PCB fits into round enclosure left over after I thrown away the faulty one (https://www.ebay.com.au/itm/173047013192?hash=item284a67df48:g: aFcAAOSwtwxfDrmS) (CMOS sensor stopped outputting data).
I also made 1.5 mm lower spacer for PCB and 3D-printed spacer that holds both PCB and USB socket in place.