Mirror cooling/sensors

[iceman]

I emailed Bird about this the other day, and I noticed Dave P also brought it up in a different thread so I thought it's worth a thread of its own for discussion.

After seeing Anthony's piece of art in action at Snake Valley, it's made me much more aware of how the mirror temperature can adversely affect the image. What I'd been blaming on bad seeing, could (in some instances) be bad temperture differences between the mirror and the surrounding air.

Seeing on Anthony's laptop, the difference in the image when the mirror was not an ambient, versus when it was after he'd had the cooling system running, was significant.

So of course it's now in my mind as a next project - to cool my mirror. However baby steps are in order, as I don't have the time or the skills to get it all done quickly.

So the first step is to monitor the temperature of both the mirror and the surrounding air, to at least log the differences and so I can see visually (through the eyepiece and via imaging) what the differences are when the temperature difference is great, what the differences are on hot vs cold days/nights, etc.

Anthony has 2 sensors measuring the temperature of the mirror, and 2 measuring the ambient temp, and feeds them into his laptop for graphing. That's where I'd like to start.

Obviously stage 2 is the actual active cooling - to bring the mirror temperature down. I'll worry about that bit later. :)

So Anthony, any pointers you can give (that we can all share in :)) regarding a setup like that.. such as:

1) The type of sensors to use
2) Where they are placed on the mirror and where they are placed for ambient
3) How you stuck them on
4) Feeding the data into the laptop
5) Graphing the data

PS: I use Windows :)

I imagine Dave P is going down the same train of thought, so hopefully between the two of us (and anyone else who may be following suit) there might be a nice how-to article as a result.

Thanks for any advice and help.

[slice of heaven]

Though I'm a few light years away from setting up an imaging setup, I'd like to know this as well.
And out of curiousity, how close to ambient should the mirror be for 'visual' viewing?

[bird]

Ok, I'll grab some pictures today and write it up, it's something I have intended to do for a long time anyway.

The sensor box spits out the 4 temps once per second over a normal serial port, so you can read it with any laptop or PC under Linux,Windows etc.

The sensors are DS18B20 one-wire temperature sensors from dallas semiconductor. They are amazing things, and only about $5 each or thereabouts.

The unit is based on a picaxe microcontroller board from www.microzed.com.au (the aussie distributor for these things). I used the 28X project board.

I also added the LCD display so that I could see the temps directly on the unit, makes it easy to verify that everything is working ok when you first plug in the sensors and power it on.

The program for the microcontroller is very simple - read the 4 temps , send them over the serial port and display them on the LCD, sleep for a bit, repeat.

I'll dig out the camera is a while and post some closeups of the sensors etc.

Bird

[bird]

Just checked on the microzed website, I used the AXE001X kit, but they are currently out of stock on the 28X cpus, I'll check and see if the AXE001 kit would be ok as a substitute and let you know.

The AXE001 project kit costs $67.50 and comes preassembled, makes it very easy to work with.

The LCD display has part number AXE033 and costs $50.

regards, Bird

[Starkler]

How accurate are they? (resolution)
I imagine step 1 before fitting would be to calibrate all sensors to read the same .

[janoskiss]

I am very interested in this too, after seeing the results in real time on Bird's laptop screen. But I'd want to do it without a computer. I'm thinking LM335 analog sensors, and an op-amp based analog 'computer' for regulating temperature; with a few LEDs to show whether mirror temp >, = , < ambient.

[bird]

The DS18B20's are calibrated to 0.06 degree C. Have a look at their datasheet:

http://lomza.czad.org/pliki/DS18B20.pdf

There's probably a copy on the dallas semiconductor website as well.

I checked the four that I received and they were all accurate to 0.06 degrees of each other. I doubt that you could match that with an analogue sensor without going to a lot of trouble.

Bird

[bird]

Here's a page that describes the temperature program that I run on the picaxe microcontroller:

http://www.acquerra.com.au/astro/software/temptest.html

Bird

[bird]

Okay, I've taken some pictures. No text yet, I'll add that later today.

http://www.acquerra.com.au/astro/equipment/temp-logger/

EDIT: I've added some text now. Have a read and see if it makes sense. Feel free to ask questions here and I'll try to answer them :confuse3:

Bird

[davidpretorius]

thanks again, so one on the back of the mirror, one near the top of the mirror, and two more for ambient?

I have learnt so much in two weeks thanks to Mr Bird!!!

[bird]

The two for ambient are actually one ambient air and one for the metal tube so I can see if the tube is at air temp. The 2 on the mirror let me see if there's a thermal gradient through the mirror which would indicate that it's still coming to equilibrium.

Bird

[Robert_T]

Hi Bird...:prey2: - very impressive! I guess this setup is easier to implement for a newtonian with access from rear compared with cassegrains (even mine which is open at front is sealed at rear). Had a thought though that perhaps some useful data could be created for these scopes through experimentation and then relating experimenal results back to real-time log of temperature just on the outside tube surface (avoiding need for cable to intrude into the scope etc). In other words setup/attach sensors temporarily on mirror, inside tube and outside etc and then expose the scope to a range of temperature profiles, recording the results. From this it might be possible then to come up with some "scope-specific" data to predict when a mirror's temp would be close to ambient air temp simply through extrapolation of the outside tube temp profile. Nowhere near as accurate as a real-time inside/outside measurement setup as you have, but probably a lot more accurate - and specific to the users scope -than the simple rules of thumb many of us would use now (eg stick SCT outdoor for at least 1hr prior to us etc).

What do you think?

cheers,

[bird]

Yes, that sounds like it might work, it would make an interesting experiment to collect the data and see. The tube temp is very volatile, so it might be a bit tricky to relate it to the mirror - the tube is thermally coupled to the outside environment more than it's coupled to the mirror. I know from the tube temp data on my scope that it can wander all over the place while the mirror temp is staying pretty steady.

As you say the first step would be to attach all 4 sensors and spend some time collecting data. Are you going to build one of these temperature logger boxes?

Bird

[Robert_T]

Hi Bird, I'm interested, but it has to get in line behind a few other projects I've been putting off...;) I guess how useful this approach would be might depend a lot on the local environment too - how stable and consistently temperature typically varies at one's prime observing location (eg. my backyard), by how much and how quickly. Brisbane might be a bit milder/predictable in terms of night time temperature variation than many other areas - I'm guessing here!


cheers,

[bird]

My personal opinion is that the temperature logger is a must-build for everyone, so you can know for sure when your scope is at equilibrium and also measure accurately how well fans or other cooling thingys are doing. For the cost, about $100 in parts and some soldering, it's good value.

I took the scope up to rockhampton in April, and even with its lovely climate the temp logger was very useful to me.

regards, Bird