1100D cold finger sensor cooling with TEC and water cooling

[Gina]

Been doing a few odd bits to the cooling system. Cut a square of 0.3mm copper sheet and cut a narrow bit from the edge, bent it out to the side and tinned it with solder ready to solder on an earth (ground) wire. This copper square will go between the 25mm TEC and the cold finger to shield it from interference from the TEC, when switched on and off rapidly. This is shown in the first pghoto below with the earth wire (green) taken to a suitable point on the metal frame of the camera and attached with one of the original case screws. The ground line of the circuitry and TECs is also connected here.

I've also made some P shaped holders for the temperature sensors (in TO92 case). These can be seen in the second and third photos, on the intermediate copper plate (between the two TECs) and the water block respectively.

[rcheshire]

I managed to eliminate most interference by first bonding the camera chassis to GND, then shielding the TEC wires (not sure how effective that was) and finally placing a 22uf capacitor across the pull down resistor Gate to GND. Apparently, I am told this last act of heresy puts the MOSFET in linear mode??? I had to learn electronics from the ground up for this project so have no idea really, but it worked. I get the occasional interference line which Pixinsight rejects well. The other advice given was to separate [removed word]ogue and digital grounds back to the Arduino board to avoid back EMF. Nice to see experts at work.

[eddie3000]

Has anybody mentioned putting ice cubes (or ice packs) in the water reservoir? Or even adding an extra reservoir with chilled water and a second pump conected to the first with a temperature controller to keep the temperature in the firsdt reservoir constant to keep noise levels constant?

I like this project very much. I might have a go myself.

Very good work indeed! Congratulations.

[Gina]

I managed to eliminate most interference by first bonding the camera chassis to GND, then shielding the TEC wires (not sure how effective that was) and finally placing a 22uf capacitor across the pull down resistor Gate to GND. Apparently, I am told this last act of heresy puts the MOSFET in linear mode??? I had to learn electronics from the ground up for this project so have no idea really, but it worked. I get the occasional interference line which Pixinsight rejects well. The other advice given was to separate [removed word]ogue and digital grounds back to the Arduino board to avoid back EMF. Nice to see experts at work.

In my previous attempt I had the MOSFETs on the strip board with quite long wires (3"-4") to the TECs. The camera metal parts were not connected to the electrical earth, the cold finger and inbetween copper plate were not connected to anything and the only decoupling was from power lines to circuit ground. I have already screened the cold finger from the TECs and have the MOSFETs right beside the TECs. I have also earthed the camera frame to power and signal earths with the screen also connected to the camera frame earth point.

I was hoping the ferrite rings I ordered would come today but they haven't so I'll try without them for now.

[Gina]

Has anybody mentioned putting ice cubes (or ice packs) in the water reservoir? Or even adding an extra reservoir with chilled water and a second pump conected to the first with a temperature controller to keep the temperature in the firsdt reservoir constant to keep noise levels constant?

I like this project very much. I might have a go myself.

Very good work indeed! Congratulations.

Thank you

Ice has been mentioned before to cool the water even more but to my thinking it's too fiddly and messy. With the proper 120mm radiator amd same sized fan I find the water temperature is just a degree or two above ambient and with a good 40C temperature differential obtrainable between water and cold finger I think that will do. For hot climates I guess you could feed the water through a water block attached to a third TEC and then water cool (or maybe air cool) that. However, I gather that even in hot climates the air temperature at night goes well down. You might want an air conditioned obsy anyway.

[Gina]

I've had a bit of a rethink about decoupling and decided it would be better to decouple the MOSFET outputs to earth/ground rather than to the power lines (other side of TECs). Less chance of feeding interference back down the power wires. So here's a new diagram.

[yesyes]

The other advice given was to separate [removed word]ogue and digital grounds back to the Arduino board...

swear word filter fail...

[Gina]

Well... Been working on it and done a bench test - FAIL Must be missing something. Applied 12v from bench variable PSU. Gradually increased voltage while I checked the camera voltage. Rose up to 7.9v and stayed there as expected with supply up to 13.5v. So far so good Then I switched on the manual PWM control and gradually increased PWM pot. Current went up gradually as expected but suddenly at less than a quarter on, the current went up to 4A and the camera voltage went up to 8.2. I think the crowbar operated and current was being limited by the PSU - I had set it to 4A limiting. I think the over-voltage crowbar has already earned it's keep!

Here's a new diagram rearranged a bit to show the earthing and wiring runs. The real circuit lacks the RF chokes and capacitors to the TECs as I haven't yet got the ferrite rings.

[Gina]

I think perhaps I should remove the camera power supply from the box. For testing I could use a battery. I need to isolate the problem.

On more thinking I wonder if connecting supply negative to frame was such a good idea. Maybe I'd better take the camera out of the box, stick a battery in it and check it out.

[Gina]

I've taken the camera out, removed the power adapter, switched camera off and put a charged battery in. Camera seems to be OK but not tried it on the netbook as that's out in the obsy and it's raining hard. But the shutter is working judging by the sound. Camera is now back in the box. Took out the camera power fuse to avoid and funny business from the crowbar circuit and tried the 12v TEC - all working smoothly - current increases evenly as PWM control is advanced, right up to full power at 4.5A. Didn't leave it on full as the water block got hot.

Next test is to see if the camera shows any interference lines though I can't get an image on the sensor without putting it on the scope - the only connection is a 48mm thread for the focal reducer (the bayonet is right inside the box).

[Gina]

Camera seems fine Tried it indoors with EOS Utils live view without scope or lens. 12v TEC control doesn't produce interference but it didn't before. Need to get the dual supply on it. I'll try that tomorrow.

BTW - put the fuse back so the camera regulator and crowbar were on and no problem there. So the problem is with the camera connected to it. Looks like the camera doesn't like it's negative battery connection connected to the internal ground. So a PSU for the camera will have to be isolated from the camera ground.

[rcheshire]

Swear word no! it should say analogue ground. But that project did generate some colourful language at times. Not a patch on the work you guys have produced.

I think 40 C differential is more than adequate. I live near Melbourne Australia. Evening times after a 40 degree day are not conducive to AP anyway. More often though nights are 20 degrees or so. Inland, dew point might get down to 3 or 4 degrees perhaps less in winter, but seldom less than -1 or -2 from experience. Coastal areas are hideous. 25 degrees dewpoint 18C Arghh... water dripping all over, but a constant 20C is better than the sensor rocketing up to 35 or 40C. Makes for comfortable shirt sleeve AP, if you can put up with the Mossies. If you can control the condensation, particularly at the face of the sensor/filter glass, you'll have a real winner.

Someone suggested ground loop problems as a cause of interference - again, if that means anything it may well be worth investigating???

[Gina]

Swear word no! it should say analogue ground. But that project did generate some colourful language at times. Not a patch on the work you guys have produced.

The simpler things can often be the most frustrating. Have to say though that this cooling project has been plenty frustrating!!

I think 40 C differential is more than adequate. I live near Melbourne Australia. Evening times after a 40 degree day are not conducive to AP anyway. More often though nights are 20 degrees or so. Inland, dew point might get down to 3 or 4 degrees perhaps less in winter, but seldom less than -1 or -2 from experience. Coastal areas are hideous. 25 degrees dewpoint 18C Arghh... water dripping all over, but a constant 20C is better than the sensor rocketing up to 35 or 40C. Makes for comfortable shirt sleeve AP, if you can put up with the Mossies. If you can control the condensation, particularly at the face of the sensor/filter glass, you'll have a real winner.

With a virtually sealed case and several packs of silica gel around the camera, I'm hoping to get the humidity right down and stop the condensation. I'm also hoping that with the sealed lens system of the focal reducer that the outer lens will be warm enough not to dew up. If I do get a problem here I shall have to add a dew heater to the FR.

Someone suggested ground loop problems as a cause of interference - again, if that means anything it may well be worth investigating???

Good point, thank you If the battery negative is connected to the metal frame where all the circuit boards etc. are grounded/earthed then there was a loop when I connected the 0v of the circuitry to the frame. I'll try it again without that connection.

[Gina]

A bit of progress to report. Camera in box with cooling bits on back and camera running off battery. Camera working fine in EOS Utils both live view and single shot. Applied both supplies - 12v and 5v - and turned up PWM. Current in 5v line increased linearly with knob and 12v TEC heated up as before. Only tested for a few seconds as there is no heat being taken away from the water block and it gets hot pretty quickly. No interference appeared in the image. So far so good. Next step is to take the camera out again and replace the battery with the power adapter unit and try that. I think I'd better have an ammeter in the 12v supply as well as the 5v and be ready to disconnect at a moment's notice I have removed the connection from circuit ground to camera frame.

[Auntystatic]

strap a 25V 10,000uf cap across the lm317 where the 100nf is, that might help with the ripple that's coming from the fet switching.

[Gina]

strap a 25V 10,000uf cap across the lm317 where the 100nf is, that might help with the ripple that's coming from the fet switching.

I think that might just be a touch OTT I have added a 47uF electrolytic across the 12v supply though.

I now have the camera powered from the LM317T and have turned up the PWM with NO interference showing on the image This is still only with diffused light though - not an image.

[rcheshire]

Well done Gina, an euphoric moment.

[Gina]

Well done Gina, an euphoric moment.

Thank you

Today I'm planning to connect everything up, screw things in place, put several silica gel packs round the camera, screw the back on etc. and then attach it to the scope. I can then carry out a test on distant trees.

[Gina]

The ferrite rings came today so although the setup seems to be working without the RF decoupling I decided to reduce any chance of RFI by going for the decoupling RF chokes and capacitors.

Here is the circuit diagram showing the specific eathing arrangement and a couple of photos showing the implementation - winding a few turns of 5A wire through the ferrite rings. The decoupling capacitors are connected directly to the source connections of the MOSFETs.

[Gina]

Just done a quick test in obsy before tea and looking good so far

Camera and box on scope and everything connected up including the water cooling system. Flap open and scope trained on distant hill, trees and road. Running EOS Utilities and camera is fine in both single shot and live view Turned up PWM and picture is perfect on this bright subject. Checked at various PWM settings. Cooling is working but I only had time to let it get down to +10C from nearly 30C. No condensation but only 5C below ambient so didn't expect any. I'll run a longer test this evening and see if/what temperature I get consdensation. I'm hoping not to see any as the box was sealed indoors on a dry day with just 53% RH (according to my weather station) and 4 x 10g bags of silica gel around the camera.

[Gina]

Successful extended test on far hill, trees and road carried out.

Temperature down to -13C and NO condensation

[yesyes]

Well done. So you've got it all under control now.

This gives me hope. I'm still struggling getting the far side of the cold finger cold enough.

Would you be willing to write a sort of summary post showing the final designs you ended up using? (circuit diagram, TEC / cold finger layout, parameters of the TECs you're using, ...)

There's been quite a bit of experimenting and it would be good to have the stuff that's working all in one place.

[Gina]

Now a clear night would be nice to do a proper test

[Gina]

Well done. So you've got it all under control now.

This gives me hope. I'm still struggling getting the far side of the cold finger cold enough.

Would you be willing to write a sort of summary post showing the final designs you ended up using? (circuit diagram, TEC / cold finger layout, parameters of the TECs you're using, ...)

There's been quite a bit of experimenting and it would be good to have the stuff that's working all in one place.

Thank you ...and... Yes, I was thinking of doing that

[rcheshire]

Excellent Gina. Brilliant result. Definitely need the enclosure and tube connection to the scope to control condensation. I don't think the same is possible with my conventional lens set up.

The novice that I am when it comes to electronics, I have put off making a dew heater for the face of the Astrodon filter and would appreciate your thoughts, guidance as to design that you may be willing to impart.