Plenty of 'Click and Lock' type lunchboxes about the right size. I butchered ....err modified about 4 of them during my experimental session. At $4-$5 each it's cheap DIY fun. And black duct tape is VERY useful.
Yes, the internals don't leave much room to play with for all that family of cameras, 450D, 1100D and I suspect the 1200D as other than a few button moves and a smoother finish to the case it is almost identical externally to the 450D.
I'll attempt my 450D first before I consider the 1200D mod.
Unless your going for total computer control of the camera the boxes will have an impact on useability. In my case I like having manual control of the setup, including ISO, framing using Liveview (even collimation tweaks through the camera). I hate BYEOS because if used for framing and focusing it automatically changes my camera setup, defaults it to jpeg, switches ISO, etc and unless you program in capture through it, as opposed to say using and intervalmetre, it leave the camera with the framing setting. It also requires an additional USB cable to my laptop which there is no room to plug in. Yes, a new laptop or usb port device may solve that problem eventually but the workflow process I have now using Liveview suits me just fine. Thus I need access to the buttons, controls, and screen.
I will concede that a sealed box offers a big advantage in condensation control, providing a dessicant repository, and can contain dry gas purges better than other solutions. The challenge there is the sealing,specially around cold finger exits, front adaptor, etc. If you exit a coldfinger through a box it becomes a permanent part of the camera, requiring disassembly to remove the box - unless you build a slot in the box to slide it off - another sealing challenge.
Another consideration with a box is keeping the camera from moving around, and stress on the cold finger sensor joint.
I look forward to what creative minds, younger than mine, can come up with to solve these problems.
Yes this is quite exciting, thanks for getting me on the bandwagon folks
Glen, you should take a look at APT...I've never experienced what you're describing. Of course, it will change settings on the camera if you select a different ISO from the menu, but not for the purpose of live view.
I tape a USB hub to my mount and run a single cable from the mounted hub to my laptop. Seemed like the least hazardous way of doing it. My next thought is to mount the USB hub to the scope rings and reduce the risk of cable snagging. I usually have 4 USB devices connected, so far.
To avoid potential future heartache, don't forget to bond the cold finger and camera chassis and run a ground lead to the supply side ground of your pwm driver.
PWM is just a digital switch on/off at different rates to vary the voltage/current through the TEC, and it can be/is very noisy. The switching effect of various devices particularly mosfets and microcontroller pins are hideous if not suppressed, leaving lines over your images.
Proper grounding will eliminate most of this. High frequency stuff can then be smoothed using small value capacitors in the right locations. I suggest reading about analog digital mixed circuits and methods of noise suppression - e.g., ground and power planes. This will be very helpful should you decide to build your own driver circuit - particularly Arduino and derivatives.
I no longer use high frequency pwm and prefer very low frequency pulsing - 1 - 10Hz at most. Noise sources are managed through proper grounding, ground and power planes and very small capacitors to handle AC frequencies, not to mention the ubiquitous inductor, separating all sources - supply and ground.
I won't dispute Rowlands advice on grounding but will say I have had no problems with lines on my images. Doing some darks as tests will show them if your going to get them. My cold finger is earthed to the sensor board frame. The finger itself is insulated with Plasti-Dip rubber spray paint, on all surfaces except the sensor mating area and the TEC area. I did that to provide a level of electrical and thermal insulation to minimise sweating inside the camera and retain cold for the sensor. From the point where the finger exits the camera it is covered with foam mat type insulation held in place with duct tape.
Consider how you are going to build structural stiffness into the cold finger/heatsink/fan assembly as this will put torque onto the sensor board unless constrained by a physical attachment to the camera or support bracket. I built a trangulated carbon fibre support for mine and it yields a very strong assembly. The photos of the carbon fibre bracket are back in this thread.
Unless your going for total computer control of the camera the boxes will have an impact on useability. In my case I like having manual control of the setup, including ISO, framing using Liveview (even collimation tweaks through the camera). I hate BYEOS because if used for framing and focusing it automatically changes my camera setup, defaults it to jpeg, switches ISO, etc and unless you program in capture through it, as opposed to say using and intervalmetre, it leave the camera with the framing setting. It also requires an additional USB cable to my laptop which there is no room to plug in. Yes, a new laptop or usb port device may solve that problem eventually but the workflow process I have now using Liveview suits me just fine. Thus I need access to the buttons, controls, and screen.
I will concede that a sealed box offers a big advantage in condensation control, providing a dessicant repository, and can contain dry gas purges better than other solutions. The challenge there is the sealing,specially around cold finger exits, front adaptor, etc. If you exit a coldfinger through a box it becomes a permanent part of the camera, requiring disassembly to remove the box - unless you build a slot in the box to slide it off - another sealing challenge.
Another consideration with a box is keeping the camera from moving around, and stress on the cold finger sensor joint.
I look forward to what creative minds, younger than mine, can come up with to solve these problems.
Hmm, my BYE just picks up last used settings, RAW etc. If you set the camera up on 'M' and Bulb then BYE just takes over everything. The only thing I touch on the 450D and now the 1200D is the power switch. It's the main reason I continue to use BYE.
Yes, the sealing issue witha cold box is problematic but one of my earlier attempts used a 'wipe' contact so when you closed the lid the cold base wiped the TEC finger and made contact. If your only aiming for 5*c then this is workable. I'll maybe look into this again with an internal cold finger system.
Front adapter is easy with a neoprene ring seal as is any electrical connection, USB, Power etc. I'll have to see if Argon can be acquired over here, never bothered to look before.
Met up with a keen mate of mine who likes to tinker and we made good progress. A bit of a nasty hack to be fair, but it's just a prototype so far. The space between the sensor and the circuit board is very tight, and I'm sure I heard something crunch when wiggling the cold finger in.
Then next gotcha on the 1100D is that there are screws for the sensor retaining plate that need to be circumnavigated or drilled for. To make it even more tricky, the main PCB is U-shaped and fits around the sensor, so it's necessary to make a 90-degree bend outwards at the same point in space. We prevailed in the end, but it's not pretty
After all that hacking, I was pretty convinced the camera would be dead on reassembly. After refitting all the ribbon cables and internal screws, we hooked it up to power and the laptop and lo and behold it powered up, connected and took a very red tinted image Lots learned and a v2 is practically a certainty.
One issue with the image is that there is a regular pattern of vertical lines. Could this be the grounding issue, or have I borked it somewhere? So far, I've haven't grounded the cold finger, but it was getting late and I'll start afresh tomorrow. Still in shock that it worked at all
Those cameras are pretty tough to kill. Make sure all your ribbon connectors are on correctly and pushed home before locking. Did you insert some thin plastic strips n either side of the finger to prevent the finger from touching a pin? Use nylon M3 screws to attach the cold finger. Any photos? Must have photos to help.
Will try and take a look again tomorrow if I have time. I had "insulated" the back of the cold finger with insulation tape, as it's a very snug fit. Given some of the physical challenges around the sensor I might attempt a v2 at the next opportunity.
Just an update...I stripped down my camera this morning as I wanted to remove all the excess baggage (mirror, viewfinder, flash) and I haven't refitted the cold finger yet.
Luckily, after putting it all back together it actually still works and the image is good i.e. without vertical line pattern. Needless to say it must have been something I did
I'm rethinking the cold finger for the 1100D now. The U-shape is hard to produce without any tools and also hard to fit, with all the protrusions from the screws around the sensor frame.
On closer observation, it should be possible to slide in the copper from the opposite (right, if viewing from the rear) as it appears unobstructed. I'd ignored this before as I wanted to keep the copper slither as short as possible. It must pass by the main circuit board but that already has some shielding on. With a little trimming of the shielding around the sensor circuit board, it should be able to slide in there with only the main PCB removed., whereas the U-shaped required the sensor assembly to be removed first.
As several people are working on cold finger cooled DSLRs I think a word about TECs is in order. They seem to be everywhere on ebay, and I used one of those initially (burnt it out pretty quick). I found that the Jaycar ones, while more expensive, are good quality and I have used the same one for over a year now without it missing a beat (PWM speaking). The one in my colour Canon is a 51W/6amp version, and I have a 68W/8.5 amp one as well which may go into my mono camera shortly.
The 51W/6amp TEC was good for a delta T of -25C for most of the year, but now that summer is approaching again it can struggle on warm nights to maintain -5C actual at the sensor. However, that's good enough and it's reasonably good on power consumption, and via the PWM can be dialed back on cold nights when all the cooling is not required. I am tempted to install the 68W/8.5amp version so that I have more 'low range' in summer, and can still dial it back on the PWM in winter. They are the same size (40x40mm) and same height. I have not done a performance comparison but when I get the mono camera completed that will be done.
It is possible to stack TECs to create 'two-stage' cooling, which some CCD cameras use, but the power consumption would make dark site use impossible but in a home/observatory setting it would work fine. If stacking you could go to lower rated TECs to reduce power consumption. As always with these things the heatsink/fan combo has alot to do with the ability to shed heat and if you aim at CPU type heatsinks/fans then the TEC will perform better. IMHO
I have attached a photo of the two Jaycar TECs I am using and the part # info is there as well.
Dunk. Tellurex has a very good technical page which explains all sorts of TEC stuff. You will find that stacked TECs is a science of its own - suggest a read.
Of all the components, choosing a heat sink is probably the most important, as Glen has pointed out. It will make or break the performance of the TEC.
For a 60 watt TEC use a 120watt heatsink, minimum. Rule of thumb double the power rating of the TEC. Another good check is to measure the temperature of the heatsink exhaust. If it's 10C or less than ambient with the TEC running flat out, you've chosen well - indicating that the thermal resistance of the heatsink is in the ball park.
No need to mention preparation of the copper surfaces and heat grease. Every junction presents a thermal resistance which needs to be minimized.
Thanks Rowland - my heat sink should be good for 100+ watts, but in the field I don't expect to use them stacked, although it'd be nice to have the option when on mains power.
When I was tinkering yesterday, I was getting up to 15C cool down from ambient running the TEC at about 2/3 power. Although this was with nothing on the cool side besides the temperature sensor, so not a real test. If I can replicate that with the cold finger I'll be happy, as my darks are very manageable below 10C.
From a little reading around, it looks like the Noctua NH-L12 is the combo to aim for I can't justify it with my recent old PC find, but just thought it might be of interest.
Another update...I've modified the cold finger to enter through the right hand side and avoid a bend. As a rough (really rough!) experiment, I stuck the Peltier on the end of the cold finger (about 50% contact) and the heatsink perched glancingly on the Peltier with another 50-60% contact...it's been running just over an hour and I'm seeing a temperature drop of about 10C in the exif data compared with the subs I took before switching the cooler on. I need to work on the box next, and solve the precarious kitchen worktop situation!
Another update...I've modified the cold finger to enter through the right hand side and avoid a bend. As a rough (really rough!) experiment, I stuck the Peltier on the end of the cold finger (about 50% contact) and the heatsink perched glancingly on the Peltier with another 50-60% contact...it's been running just over an hour and I'm seeing a temperature drop of about 10C in the exif data compared with the subs I took before switching the cooler on. I need to work on the box next, and solve the precarious kitchen worktop situation!
Need pictures !
I am contemplating that option rather than the bends involved in coming from the left side. I have perused all the pix but haven't opened up the 450D yet to get a proper 3D look at it all.
As the card slot is not used I figured that might be a good entry point but I need to take a proper squizz at it all myself. I don't want to start this till I have enough free time and we're heading into silly season and a company building shift over Xmas. It's going to be hectic.
But pix would be nice, angled off bottom right of the camera might give me an idea of spaces and finger entry.
Brent have a look at Gary Honis' online model pics for the 450D, even if you don't use hs model it is a good source of photis of the internals of various models, especially the 450D. Just search for 'Gary Honis DSLR modifications' and it will be near the top of the search. I don't think coming in from the SD card side (looking from the rear) is feasible because of the placement and thickness of the rear board and the main sensor board connector to it. The left side is a clear run underneath the usb connector board straight into the back of the sensor - which is why so many people go in that way.
Having said that, I've attempted to power it up this morning for its first thermal test of the new cold finger and case mod and it's dead we went through a lot together. Not 100% sure I'm ready to crack open the known working 1100D so close to New Moon.
The 450D and 1100D appear to differ in that there is a chip on the sensor assembly circuit board whose pins protrude into the cold finger space, hence entry from the hand grip side (right) is simpler than from the left.
Lots learned and a v2 is practically a certainty.
we went through a lot together. Not 100% sure I'm ready to crack open the known working 1100D so close to New Moon.
