I worked out the weight of my colour cold finger 450D today, it was 1.5kg complete: With T-Adaptor, UV/IR Filter, and bayonet ring, cold finger assembly including TEC, Deep Cool 200 heatsink/fan combo, the carbon fibre support bracket, battery eliminator, all wiring, insulation and Argon bag. This is lighter than I expected.
The spec sheet for the 450D lists the body weight as 475gms without battery. The weight of the Deep Cool heatsink/fan is 400gms.
Out of interest I also weighed the new Coolmaster/Noctua fan combo that I have assembled for the mono camera, and it came in at 450gms; not much difference between this 120mm solution and the 92mm Deep Cool. Probably will not be a factor on the focuser.
I'd be interested in getting some other weights from people doing these mods so that we can baseline it.
Haven't stuck mine in the scales yet but it's nowhere near that, I stripped most of the unnecessary crap for astro and the DC coupler has practically no weight compared to the battery. I'm going to give the cooler master a go as it's light and I've been happy with the performance so far and it's not so out of place on the end of the new casing.
Have had an email from CentralDS to advise that my 450D sensor has arrived safely and they will be starting the debayering this week. They have tested it to make sure it is operational and everything is ok after transport. Just be aware that shipping times from Australia to South Korea take from 2.5 to 3 weeks using Parcel Air/Aus Post. I am sure International Express would be faster but cost alot more as well. I paid $14 for Parcel Air for just my sensor board, a complete camera would obviously be much more.
Once Yun Lee advises me the debayering is complete and the sensor testing done, I will make the payment to their bank account and he ships it back (hopefully using EMS or one of the other fast Asian courier outfits).
My first build on the 450D should come in at under 1 kg. But I'm only using a small alum heatsink plus fan to start with. When I go copper finger I will up the heatsink capacity but still reckon I won't be much over 1 kg all up. I'll do some weight measurements when it's closer to completion.
I'm also thinking about the argon idea but wonder if the result justifies the effort if temps just above zero do the job.
Received an update from CentralDS late yesterday. The debayering of my 450D sensor is complete and the unit has been tested, and is ready to return. Yun Lee advised that the Canon image sensor wafer characteristics are not all the same, and up until my 450D they had only debayered 600d, 5d and 6d sensors. As they did not have experience with the 450D sensor they elected to bring in two 450D sensors for testing prior to starting mine - this is what has taken so long (the delivery of the two test units). The test units arrived this week. The CentralDS debayering process is based on a chemical solution to strip away the bayer matrix and micro lenses without scratching the surface of the sensor - unlike the process formerly used by JTW and others, which involves scraping off which can scratch the sensor and had the potential of breaking the gold wires on the edges. So after testing their chemical stripping process on the test 450Ds and finding that their process worked and the units tested OK afterwards, they went ahead and debayered my sensor. My sensor passed the QA testing and is now ready to return ship.
Here is a link to the CentralDS writeup on their debayering process, with examples of mono resolution improvements:
Today I received my now debayered 12.2 Megapixel sensor back from CentralDS. A really quick return trip of only four days from South Korea - way to go EMS.
It was returned disassembled because I asked for it that way, I did not want the original Canon cover glass reinstalled (it's too highly reflective compared to Mulitcoated alternatives), and I wanted to inspect the debayering job right to the edges and contacts. You can see in the photo that the chemical bayer matrix stripping does make a bit of mess on the edges but the pixels area is all clear and there are no scratches, stain, etc - I can tolerate the edge mess.
I also received the Astronomik MC Clear EOS Clip-in filter that I ordered to replace the Canon cover glass to create a sealed chamber in front of the sensor.
Now I can get to work on reassembling the sensor and installing the cold finger. The Anti-Alysing filter is still on the plastic filter frame but I will be removing that as well from the final assembly. The optic train will just be the MC Clear Clip-in to the sensor face.
I may cover the 'now' exposed contacts on the edge of the sensor to prevent any flaring off of those but I need to consider how I cover them up or even if, provided they are hidden by the sensor frame.
Photos attached.
BTW I will be putting a Teleskop-Express EOS Filter Drawer on the camera as well. Details here:
This filter drawer will maintain the focal distance of my current EOS T-Adaptor and allow changing of filters without removing the camera from the focuser, and provides adequate heatsink/fan clearance to allow access. It seems to be the only real option for this cooled camera. I have not ordered this unit yet, that will have to wait for after Christmas I think; I want to test the rebuilt 12.2 megapixel mono cooled camera in just luminance mode for the coming new moon period if I can get it ready in time. Looking forward to a big resolution improvement over the OSC sensor.
The camera has now been reassembled, cold finger installed, and preliminary daylight ambient testing done (the TEC is not installed yet). I departed from the colour camera design and used a straight through cold finger, this will give me a bit more clearance in the front for the filter drawer as the fan will be placed slightly to the rear; this design means leaving the small connector board out of the camera but that's ok as I will be using BYEOS to drive this one. I epoxied the temperature sensor to the cold finger just where it enters the sensor back, and inserted some insultion foam on top of the cold finger inside the camera. The cold finger, which has five coats of Plasti-dip rubber paint, sits on a carbon fibre spacer so that cold is not passed into the camera frame. I have attached a few photos.
For first light testing I stuck on an old Tamron lense and took a couple of shots on full auto just to verify that the sensor was working and there was no damage during debayering (confirming what CentralDS said of course). Hard to tell much re the resolution on the test shot but the flower when zoomed in carries detail very well for a simple jpeg. The resulting camera image is pink of course, no IR filter on it and it renders everything in pink. Simple enough to just view in B&W in Photoshop but a custom colour balance will be needed for daytime use - well not that I plan that anyway. Of course I had to dial back that image quality substantially to get in on IIS, so you can't really tell much from a "Save for Web" file.
I will get to work on the cooling in the next couple of days. If the cloud ever goes away I hope to get it on a scope soon.
Straight through finger makes it a lot easier hey ?
Interesting my 450D doesn't come up as pink as yours does. Certainly brighter but not just PINK all over. Must test further.
Straight through finger makes it a lot easier hey ?
Interesting my 450D doesn't come up as pink as yours does. Certainly brighter but not just PINK all over. Must test further.
Well I don't have a problem with the folded finger design, but if your deleting the connector board then you can go straight out with spacer support.
I think I have worked out why the test image was pink, its comes down to the camera settings. I was shooting in RGB, but it is possible to shoot in Monochrome (Style Settings allow you to choose this). From my reading on the Net it is suggested to just go ahead and shoot in RAW and simply process as Monochrome in Photoshop. The argument being that capturing in RGB preserves colour interpretation data in the RAW file, but capturing in Monochrome you lose that and are relying on the camera to do the interpretation. I will try both modes on my DSO test shots at some point. For now here is a second test image of the same plant taken with the camera set to Monochrome.
That doesn't make sense to me. With a monochrome sensor, ANY colour profile or setting should be irrelevant.
All you should need to do is shoot in RAW and NOT demosaic. Photoshop won't know how to do this, since it assumes that all .CR2 files are colour encoded. Try dcraw instead...
Your probably right Dunk, I will look at the image histogram on Photoshop. I admit to knowing nothing of mono processing right now. For example, if I shoot a RAW sub file through a narrowband filter, how do I 'assign' a colour to that data, etc. A whole new processing regime to learn.
Glen, what you get through a filter or otherwise at this point is all just "luminance" how you assign a colour to it is up to you and is just aesthetics.
Btw, try:
dcraw -4 -o 0 -D -t 0
This should output a 16-bit monochrome TIFF file, with no funky business.
The only difference from what I can tell is the physical removal of the colour matrix. The camera firmware still applies a human eye response curve to the RAW file; that is, sRGB or Adobe, whichever is preselected. Presumably, this applies to monochrome files produced by the camera as well - not sure about that one...?
hi
just take a pic of a white screen on your monitor and use that as a custom white balance. all images will then display as monochrome.
you don't need to debayer in software. just use it as a raw mono file.
Managed to get the first DSO images last night, but have some problems still to sort out. First my old camera controller power supply failed the night before and I had to switch over to another one (the Bintel 12V 7.5amp supply). The Bintel power supply was putting out moving interference lines visible on Liveview and on BYEOS, it's fine running the mount but not clean enough for the camera (and the camera cold finger is earthed). So I switched again to run the camera controller directly off the battery system and that cleaned up the interference - which was a relief as I was starting to think it was the camera.
Next I noticed that I had some internal fogging happening. I had not bothered to put the camera into the Argon bag for the test, nor had I insulated the cold finger on the outside of the camera body (other than the Plasti-dip coating), and it was a more humid night - that combo led to optic fogging somewhere. I shutdown the cooling and allowed the camera to heat up enough to evaporate the fogging, then turned the cooling back on for the test shots.
I ran a set of five 180 second subs on M42, which looked ok on Liveview and then had to shut it down due to approaching cloud.
On stacking and processing this morning I can notice a pattern on the images that appears to be something like evaporative gradient patterns - probably due to the fogging or my attempt to clean the optics prior to the test - using isoalcohol instead of my preferred acetone.
So no image to share at this stage, not until I am happy with it.
**** I spoke too soon!! This morning I tried to clean the sensor with acetone down through the open shutter, something I have done in the past with no problem, just set low ISO and bulb mode. Sadly I managed to snag the shutter curtain and seem to have damaged it - DOH! I am hoping the spring has just popped off the control arm but I will need to disassemble everything to get to it ( or maybe I can just take the front section off to get to the spring). Might have to step away from this till the new year if the shutter needs to be replaced.***
I believe I have found out what happened to that power supply. I opened up the case and found water inside - large drops moving around inside the bottom of the case and obviously shorting the bottom board out as they moved. That explains the 'trying' to work nature of the fault. I had been leaving the power supply out in the observatory and I guess as it cooled off it has been drawing in moist night air which condenses at some point and builds up into drops and gathers at the bottom of the case. I have disassembled what I can and blown it out with compressed air. It's now sitting out in the sun to try and drive out any residual moisture. Hoping its going to work again when I get it back together. It was a good supply, with good shielding, intended for LED light strings I think. Fingers Crossed.
In other news, I believe I have fixed my camera shutter and it may not need replacing. One of the shutter curtains had been pulled out of the slot it slides in (after snagging on that cotton bud), and after I got the front of the camera off and confirmed the springs and actuator arms were ok, I managed to get the curtain back in the slot without breaking anything. Cycled it a few times manually with the gear wheel so it should be ok when I get time to put it back together. Must get onto cleaning that sensor properly.
Certainly a better result than I expected.
Geez Glen, you have some weird things happen.
I would have thought that the power supply would stay dry due to internal heating ? Not that I have water problems but my 12v supply for the mount is a sealed unit like a laptop supply so no water intrusion possible. The 12v-5v supply for fans etc is a PC unit modified but that sits up clear of the floor in a desk shelf. Seems to put out enough heat to keep itself dry.
The initial test subs of M42 (non-cooled) showed some isopropanol evaporative type pattern on the surface which I assumed could be cleaned off, but more concerning was a couple of small diagonal lines and some black pixels scattered througout the image when zooming in to pixel level. The lines and black pixels only showed up on long subs of M42. I will go back and look through all of the subs shot at different exposure times but it is strange, it's not processing related.
Since that test I have completely cleaned the sensor with pure acetone and things have improved.
I have shot some auto shutter speed daylight shots, and flats against a white sheet of paper - which you would think would show up black pixels if they were dead pixels or pixels covered with residue of the bayer matrix. They did not appear in the short bright test shots. The lines, if they are there, probably won't show up in flats or bright shots.
I will probably buy one of the sensor lupe inspection tools so I can examine the sensor surface in more detail. For now it is clean enough to resume testing - when the cloud goes away.
get it put together and start testing, the Moon isn't pestering us for a reasonable part of the evening now...
