Super Cooled Mono 350D Project

[2stroke (Jay)]

Thats coming along sweet nice work man Great to see more guys on here growing some big ones and diving into extreme modding, love how your sealed it off while keeping the eos ef mounting system. Lol this thread makes me want to revive the 1100d's XD

[rcheshire (Rowland)]

Jo. Just use a thermometer to measure the ambient temperature and then hold it in the heatsink exhaust. 10C differential is good. If you are doing better than that then you could use a bigger TEC without modifying things too much.

PWM or linear are the two methods of control. I find PWM much easier to manage having spent a lot of time recently dicking about with linear. When you get out at night it's a pain to adjust to maintain a consistent set temperature.

PWM is more complex - you need some form of pulse generator, either a 555 timer and potentiometer or a microprocessor, to generate PWM an NPN transistor, MOSFET, resistors capacitors and an inductor. I will post an updated circuit, once I get it tuned up.

PM me if you want to discuss some options.

I'm getting setpoint +/-0.2C - which is great for darks.

[nebulosity. (Jo)]

Originally Posted by rcheshire

Jo. Just use a thermometer to measure the ambient temperature and then hold it in the heatsink exhaust. 10C differential is good. If you are doing better than that then you could use a bigger TEC without modifying things too much.

PWM or linear are the two methods of control. I find PWM much easier to manage having spent a lot of time recently dicking about with linear. When you get out at night it's a pain to adjust to maintain a consistent set temperature.

PWM is more complex - you need some form of pulse generator, either a 555 timer and potentiometer or a microprocessor, to generate PWM an NPN transistor, MOSFET, resistors capacitors and an inductor. I will post an updated circuit, once I get it tuned up.

PM me if you want to discuss some options.

I'm getting setpoint +/-0.2C - which is great for darks.

Thanks Rowland,

I'm getting around 5c difference between the heat sink and ambient air temp.

How does the PWM know how much to adjust the the power to keep the temp the same? Is it connected to something else?

I got a 12v PWM temperature controller on ebay thats got a temp sensor, and also a screen and buttons so you can set the temp differences and stuff, I think it can keep with in 1C so I'm interested to see what it's like. It might not work but as it was only 15 bucks I thought see what its like.

Should arrive next week, I can post a link if you like.

Sorry, love to PM you but I don't have an email address

Cheers
Jo

PS. The temperature controller is this

[rcheshire (Rowland)]

Looks good Jo. 5C is very low indeed. My set up is controlled by a microcontroller that I program and a few additional components. I have come up with an algorithm to keep the temperature on target/setpoint with changing ambient temperature.

Your set up should do the same. I see some up/down buttons and a Set button. I'm guessing it will read cold finger temperature and maintain it from what I can see. Should work a treat.

The PWM value is the same for a given temp differential for a particular set up. Same energy requirements.

My program estimates the PWM value to maintain setpoint based on the difference between ambient and setpoint temperatures and drives cold finger temp down at 100% and calibrates PWM based on any tendency for cold finger temp to wander up or down from set point - it looks at the difference between setpoint and cold finger temperature and adds or subtracts from the estimated value and uses that as the new PWM value and so on. The tendency is to dither either side of setpoint

I just like the pain of designing my own system - glutton for punishment.

[nebulosity. (Jo)]

Thanks for the explanation Rowland it makes a lot of sense now.

I accidently broke a blade on my fan so I had a look in my junk box and found another one that blows a LOT more air, it's got the heat sink down to 3C above ambient with the TEC running flat out.

Cheers
Jo

[wasyoungonce (Brendan)]

Originally Posted by nebulosity.

....PS. The temperature controller is this

Yep they work well. I have one (a variant of this one) in my own cooler box and managed to get it to cool to -1C, but, it would cycle sine wave to 0C due to hysterisis of the controller loop.

In anycase, anything at or below 5C (actually ~8C) is more than enough to radically reduce DSLR sensor thermal noise.

[nebulosity. (Jo)]

I made a new copper cold finger yesterday and installed it into my camera along with the new TEC I got (60W), anyway now I’m getting weird lines across my images, any idea what it could be?
I had the same problem with my old TEC but solved it by putting a 2200uf capacitor across the power wires, but no go with this new one.
Any help would be appreciated.

Cheers
Jo

[wasyoungonce (Brendan)]

try the tec cooling off and take an image

[nebulosity. (Jo)]

Originally Posted by wasyoungonce

try the tec cooling off and take an image

Tried that and there are no lines, must be something to do with the TEC power it's 12V 8A.

Any idea on what I could do?

[wasyoungonce (Brendan)]

Try a different +12V supply for the TEC. Maybe try the TEC cooling without control...aka full on.

Is you camera running on batteries? If not try above but use a battery to isolate DSLR power fro the TEC power.

[rcheshire (Rowland)]

Ahhhhhh......Jo. I have been working on a solution. In my case, the noise/lines is generated by the switching of the MOSFET gate.

Make sure you have grounded the camera chassis and the cold finger and the case.

The capacitor is preventing the MOSFET gate from going low enough to switch off. This means the MOSFET is being driven linearly and gets hot as well as being difficult to maintain temperature control.

A solution that seems to be working is to keep the MOSFET gate from going too low when in its low state. I have had success by setting two voltage sources to the gate. One 5V and the other ~2.5v and alternating them on and off at about 1khz. The MOSFET runs cold.

PWM works much the same by keeping the low voltage above zero. I will post the circuit following more testing.