Out of interest - another DSLR cooling project - #5, I think. This should be entertaining reading...
It's PWM using a Teensy 2.0 with Arduino code. The Teensy native PWM frequency is much higher than the Arduino and more suitable, I think.
The idea was to cram all the electronics for cooling, dew heater and camera sensor defogger into one relatively small box (which hinges as shown for battery replacement - yet to fit the velcro strap), to make temperature control accurate and human friendly - a single switch and LED provide all the control and feedback necessary.
Cooling capacity is 20C max differential - cooling is zoned within this range to operate at 7, 5, 0 and -5C (7 is a bit arbitrary at the moment) depending on air temp. Once setpoint is assigned it doesn't change (based on temperature reducing at night). I'm not going to use it at > 25C.
Within a few degrees it will accommodate an increase in air temperature. A new session can be started within 5 -10 minutes of shutdown, giving the cold finger time to warm up to ambient allowing setpoint to be recalculated with the push of a button.
The push button on the front of the box is depressed during power-on, disabling cooling, while registering ambient temperature and assigning setpoint temperature - using a single temperature sensor attached to the cold finger. The red LED flashes morse code to report the selected setpoint - I kid you not. ... ..-. --.. -- S F Z M ??? you guessed it Seven, Five, Zero and Minus (5 only). Releasing the button activates cooling - on-temperature is indicated by the LED.
Now I want to try it out. I have fiddled with the snubber network to get the smoothest MOSFET switching possible, based on oscilloscope output. Virtual design tools are amazing.
FYI - high power, logic level MOSFET, Vgs 4.5 and very low Rds(on) - IRF2804 from utsource. I was using a NXP PSNM1R1-PL30, but these are difficult to source and expensive to ship. Both run cool without a heatsink - barely warm to touch. However, I've used a Freetronics SMD MOSFET in this project, with similar characteristics.
I know that there are differing views on the use of PWM, but I like it. Linear is naturally noise free but, having spent many hours working with it, it's a pain to manage in a nights imaging, where consistency is needed, particularly for calibration libraries.
Thanks Jo. A few refinements and hopefully this package can be used across several platforms. When I'm happy with it, I'll move it across to a cheap second hand 450D I bought off ebay.
Next job is a custom light weight heatsink to meld into the black anodized look. I'm not sure how to go about this - machining an aluminium block or creating a mold and casting. A machined block might be easier. I cant find a heatsink with the right specs on-line.
Thanks Jo. A few refinements and hopefully this package can be used across several platforms. When I'm happy with it, I'll move it across to a cheap second hand 450D I bought off ebay.
Next job is a custom light weight heatsink to meld into the black anodized look. I'm not sure how to go about this - machining an aluminium block or creating a mold and casting. A machined block might be easier. I cant find a heatsink with the right specs on-line.
The breadboard will go eventually.
How about a PC CPU heatsink ? I've got heaps of small weird shaped alum\copper slugged, coolant piped critters with radiators I've rescued from old PC's.
Have you got any specs of your requirements ? Even to take something already made and mod it to fit might be an option.
I'm using an old Phenom cooler which does a pretty good job, but no aesthetics.
Rule of thumb is twice the power rating of the TEC - 140w. I have a 200watt Alpha Novotec, but it's a brick.
Lightweight, compact, lots of surface area and high volume, silent, vibration free airflow is the criteria. I think thats why the Phenom cooler works so well - but it could be flatter.
Lot nicer than my cooling job need to update me thread.
Awesome work
Jo
Jo, I think yours is superior in several ways. I am not skilled in machining and I don't have a workshop - stuck with a small bench, vice, hacksaw, file, drill and soldering iron - sort of have to adapt.
I tried out a new cooling algorithm last night. It's more dynamic than the previous. Setpoint didn't budge, except for the usual momentary spikes in readings - very happy with it.
Just playing around I programmed the LED on-temperature indication to illuminate at set point +/- 0.2C, and, it remained on and steady. I don't know, but that doesn't seem too bad for a cooled DSLR. I suspect it will do better with aref 3.3v - a bit more accurate than the present 5v.
I have sourced a heatsink to replace the ugly looking thing, that hangs off the coldfinger. Lower profile and will take 2 x 50 or 60mm fans.
Code calculates a target pwm value to maintain setpoint and then drives TEC 100% duty cycle to setpoint. After that pwm value is recalibrated based on any temperature differential from setpoint and dithered either side of setpoint.
New heatsink arrived today. It has good fin density, is light and low profile compared to the cpu cooler. It looks better too. Flat side is nicely machined. Balance is much improved.
Differential increased by 4C to 24C with a 70mm fan. The tape is temporary until I make a shroud to prevent light leaks.
Well this has progressed and you may have read about my agonizing efforts to find a quiet switching solution to eliminate EMI and get rid of those nasty lines in images. Digital cameras are very sensitive to EMI.
I have tried 4 or 5 different logic level n-channel very low rds on mosfets driving direct from the PWM pin and an MCP1415/16 gate driver with mixed results. Having come up with all kinds of snubber networks, poor, fair and bad, rigorous testing revealed flaws in all .... back to square one.
In a nutshell the easy method, producing the cleanest images is a 22uf Tantalum capacitor Gate to Source. This does not send the mosfet linear from what I can see. The system is slow and temperature control sloppy, although useable, which is to be expected. The NXP PSMN1R1-30PL runs cool - unlike others. It's smd cousin the 1R3-30LY runs in liquid solder - don't be distracted when testing these things - it stopped working when I tilted the board and it slid off. So much for home brew smd, for now.
The Android SPICE based program I been using seems OK and tends to verify practical observations. Images below. Time base is 20us. 12v side is 10mv resolution and offset ~5mv ripple. Low side is 1v resolution. The other Android simulator had a bug producing incorrect results. The developers appear to have fixed it.
I understand that a 22uf cap gate to source may not be optimal as far as switching losses are concerned, however, I am at a loss to find another workable solution. I thought I had EMI solved on several occasions, but there was always anomolies. I will be less hasty to cry Eureka in the future - DRAIN turn-off produces some nasty voltage spikes which are easily eliminated with a high speed diode source to drain, or so I have found.
If I have some spare time today, I will play with one or two notions that come to mind, based around the 22uf Tantulum capacitor, to improve the system response.
Changed direction and abandoned any hope of a snubber solution. Instead, in a moment of unusual electronic insight I decided to use ground planes to isolate analog and digital ground and went a step further adding a second copper plate for the mosfet with planes for each pin - source is connected to digital ground.
Analog and digital ground are connectd by a 100R resistor - added a power plane as well?
To be honest, I'm not certain which part of this system is most effective, but temp control is rasor sharp and images show no traces of switching events.
NO CAPACITORS... NO NOISE... Even driving directly from the pwm pin.
Sadly, my beloved NXP mosfet broke a leg and had to switch to a StrongIRFET, which runs in the palm of your hand - no heatsink.
First of all, I apologise for the hand drawn diagrams in this age of Eagle CAD and other software, but it is a matter of time and opportunity.
I've attached the general layout of the system. Other services could be added, expanding on the concept of keeping all functions in a single box. Shutter control can be moved to the camera as well.
At this stage sensor cooling, defogging and dew heater are in place, with plans to implement automatic control to these functions - having resolved the EMI problems producing a workable flexible self regulating system.
Control and indicating is minimalist. 1 button and 1 LED with setpoint and on-temperature indications - the addition of temperature sensor failure indication with automatic cessation of cooling and other services is in the pipeline.
EDIT: Following a little maintenance the camera chassis was unintentionally isolated from the cold finger, which generated EMI in live view. Bonding the cold finger to analog ground resolved the interference. However, both camera chassis and cold finger should be connected to analog ground for silent operation.
This is the general layout/principle for grounding the camera chassis and cold finger. Absolutely essential.
EDIT: I noticed an error in the diagram, which would have been obvious to the trained eye - I missed it. Corrected and new image uploaded. The fan power was connected to the heat sink and looked confusing - removed for clarity.
Gee whizz Rowland your really perfecting this camera cooling nice work
Love your hand drawn diagrams the grounding layout is very helpful for my camera.
What arduino module are you using? I don't know anything about them but they certainly look interesting and I was thinking about getting one to experiment with and hopefully gain a better understanding of how they all work. I'm hoping someday to make an electronic filter wheel.
My camera cooling is going well, getting around 35C below ambient regulated cooling, unfortunately I wrecked the sensor while removing some leftover CFA so it's out of action for a bit.
You've done a really well considered job on your camera, Jo. It's very good. I have some spare TMP36 temp sensors. Your welcome to one if it will help. No charge.
It's taken a while to acquire the learning to get anywhere near the results to date. It fell into place one day - I am tempted to do an EE course at Uni. I think it would be fascinating.
I'm using a Teensy 2.0 - it's not Arduino, but it is compatible and it was necessary to install the Teensyduino environment alongside the Arduino environment to use Arduino code - a bit easier for me than C++. Teensy can be run with C++ or the Arduino Wiring program.
The Teensy probably has better hardware, but that's a matter of opinion. It's not so easy to configure for external voltage sources as the Arduino. Handy if you want 3.3v for analog sensors and 5v for other stuff - motors, pwm, leds etc.
I would go with the Arduino for now of similar size such as the Leonardo and Nano, which is a bit smaller. Maybe grab a bigger model such as the Uno, which might be easier to handle first up. Jaycar sell the Freetronics version along with a bunch of convenient ancillaries, break out boards etc.
Some of the guys on the forum are real whizzes with programming. Mine is very basic - nothing flash, but it does what I want.
Next will probably be a double sided PCB. I will do the artwork by hand - roughed it out already. I'm looking at dual redundancy power and mosfets.
Load sharing mosfets - mosfet and pwm source fail modes
I am thinking of load sharing the TEC current with two mosfets, as well as single mosfet operation (one failed) and pwm source redundancy.
Two pwm sources are hard wired to the mosfet gates a primary and secondary. The secondary pwm source is shorted through a transistor. On failure of the primary pwm source, the secondary circuit is closed, kicking it into action.
Conceivably, in a nights imaging, loss of one or the other mosfet or the primary pwm source should ensure, uninterruped operation, replacing failed components in the hangar, instead of in the field.
It needs failure indications and I am not sure about the correct values for the (sense?) resistors. But it works in simulation.
Made a few changes to the auto pwm switching because there is no easy way of switching the ground side of the pin. Added switches to test the loss of pwm or mosfet.
EDIT: This is just being clever. In reality a mosfet is likely to fail before a pwm pin, in my experience, and mosfet failure is very unlikely. However, from what I can see, the mosfet remains in the state at which pwm failure occured, either on or off. If on it is linear and will most likely overheat. Although it is at 5V in this case, so maybe not.
This is the rough artwork for the Arduino compatible (Teensy 2.0), PWM, MOSFET driver board.
The idea of the multiple rails is to provide 5V, analog ground or digital ground points for various sensors and devices. Not too restrictive if I want to add new features.
It makes use of ground and voltage planes, which has proved successful in reducing mosfet switching EMI to visibly indiscernable levels.
The board will fit to the left of the battery door attached to the tripod mount (through AGND), which is part of the camera chassis. The cold finger is bonded to the chassis.
I want to see whether a bigger board or an additional board is required to increase the size and effectiveness of the ground planes.
The Teensy also has its own 5V supply - a miniature smps regulator, same size as a 7805.
I have integrated 3 of these into the cooling circuit. The logic pin allows activation of dew heaters in response to temperature or humidity conditions. Rated at 1.5A, maximum current to each heater will be 0.7A - 1 unit per heater, plus a spare.
This is the Beta board. A very crappy job of etching. The production board will be factory made This will do for now. Past the experimenting stage, it needs to be tested in situ.
I have incorporated the ground planes for the MOSFET pins as well as separating digital and analog ground sources adding 3 adjustable power sources for dew heaters and an experimental 3.3V external supply for the Teensy/Arduino, as well as a 5V supply. Cooling and heating off the same board.
One side of the Teensy has all the analog pins with a third hole for decoupling capacitors to analog ground for sensor inputs.
All ground connections are grouped on the ground planes according to digital or analog inputs.
The board and case clear the battery door. The case will be clamped to the camera base through the tripod mount hole and fixed to the heatsink for rigidity with double sided tape. The MOSFET is located so that it can contact the main heatsink.
Eventually the case will accommodate camera and dithering control - unguided widefield setup.
. Releasing the button activates cooling - on-temperature is indicated by the LED.
Looks great.
need to update me thread.
nice work 
