The light is on with cooling

[iceman]

It's common knowledge that for the best views/images, the mirror has to be at the same temperature as ambient.

People like Bird, and later David Pretorius, have gone to great lengths to actively cool their mirror using peltier coolers and fans to blow cold air onto the mirror to help the temperature come down quicker than just blowing ambient air.

I've had the cooling fan (http://www.iceinspace.com.au/forum/showthread.php?t=9981) on my scope for about 2 months now, just blowing ambient. I've also got the temperature sensor to measure ambient vs mirror temperature, and even just a normal fan blowing air onto the mirror does make a difference and helps the mirror cool quicker than just leaving it out on its own.

But if the ambient temperature keeps dropping, the mirror will never catch up. I've only ever seen the mirror get to within 0.8° of ambient once ambient has stabilised, but it usually sits 1-1.5° above.

Last night, the ambient didn't drop for quite a while - it was sitting on around 12°. I started imaging at around 6:30pm and as usual, the seeing was ordinary with large "waves" going across the preview screen. Over the next half hour, the mirror temperature actually came down to within 0.3° of ambient - the first time it's been that close.

And then it happened - the image stabilised incredibly. The seeing still caused the image to vibrate and distort a bit, but the large waves had completely disappeared. The image was much steadier and there was a much larger percentage of crisp frames.

It showed me for the first time, without a doubt, how important it is for the mirror to be at (or as close to) ambient temperature. What had been blamed on bad seeing was in part due to the boundary layer causing distortions in the image.

I had thought that a difference of around 1° was close enough, but now realise that it has to be even closer. Last night the change happened at around 0.3-0.4° difference, when the air in the tube stabilised and the image became incredibly still.

This experience has now fast-tracked my plans to actively cool my mirror, as Anthony and DP have done.

[RB]

Very interesting indeed Mike.
It'll be great to see the difference it'll make.
How long before you do the mod?

[iceman]

Don't know Andrew, I've gotta figure out exactly what I need, how to assemble it, how to power it, etc.

Expect a thread in ATM soon asking all the questions :)

[Lee]

Always a problem with cooling things isn't it..... heating - simple, easy, cheap.... cooling a way different proposition!
Same with tropical fishtanks - heater $50, cooler $hundreds+

Good luck Mike - I can see you hurrying along so as not to miss Jupiter as it disappears! :)

[IanW]

Active cooling isn't expensive to implement, a peltier device and control electronics can be had for around 50 bucks if you're handy with a soldering iron. The hard part is determining the method of mounting and how exactly to spread the cooling to maximise it's effect.

[netwolf]

Something on my todo list that keeps geting put off. Is there any good sites with information on the ins and outs of cooling. I know a while back Bird had a link. I would like to learn more about what is happening and understand it better before working on a solution. What are the facts and are they measurable and repeatble or random?

[IanW]

Here's a great starting point:

http://www.acquerra.com.au/personal/bird/astronomy/cooling/

Cheers
Ian

[iceman]

Yep, that's Bird's site. A wealth of information.

[asimov]

Dave P has some very interesting posts in this forum on his experiments with a peltier cooler on his 10" scope too Wolfie.

[netwolf]

Ok so i just spent the last hour+ reading through that link and the rest of Bird's site. I understand know the importance of equilibrium in such systems.

Question1: What causes the mirror temp to rise when cooling is turned off? Whats heating it?

Question2: Without cooling, If the mirror is hot why does the tube temp drop bellow ambient would not the heat from the mirror rise up the tube like a chimney? Hence should not the tube temp follow in sync with mirror temp?

I still am missing an aspect to this, i think i need more information about the flow of energy in this system.

[Dujon]

I wondered the same, netwolf. Thinking about it I wondered if the mirror's "core" is still well above ambient while the "surface" has cooled to almost that. The inference here is that the cooling system is sufficient to keep the surface at a reasonable temperature while the "core" energy is being released but as soon as the cooling effect is removed the surface once again begins to heat up relative to the ambient air temperature.

I'm probably wrong - it wouldn't be the first time. :(

[Photon]

Interesting questions! I'd imagine if the core of the mirror was still warm (a good suggestion BTW) optimum viewing would not be acheived as there would still be some surface distortion due to differences in expansion/contraction rates coupled with the temperature differentials(??)

[IanW]

Dujon's answer is right, it's heat energy stored in the core of the mirror. Pyrex and most glasses used in mirrors act a bit like a bucket of hot water, that is they loose the heat energy from the periphery first and then the energy gradually flows out from the central core region as the periphery cools.



It's a very complex issue involving the concept of black body radiation and the fact that a telescope radiates heat to space (ie: to 0 Kelvin). Also you have to factor in the specific thermal radiation properties of both tube and the material the mirror is made of (glass or ceramic).

It's also an area that gets extremely complex to model mathematically, so rather than bogging this discussion down in formulae and miles of terminology here's the lay version :)

Each component (mirror, tube, mirror cell) absorbs and radiates heat at different levels. In general a tube will absorb heat and cool faster than a mirror because it's got a greater surface area. If the mirror was in direct contact with the tube there would be heat transferrence, however there's usually about 20-30mm air gap between the mirror and tube and this air isn't sufficiently warmed by the mirror to transfer much heat to the tube. It is enough to produce eddy currents in the air column in the tube though (tube currents).

One thing Bird didn't mention in his article was the accuracy of his measuring gear.

[bird]

Hi guys,

regarding the electronics and stuff that I use for measuring temperatures, it's explained in a seperate article on my site, here:

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

The temp sensors are quite accurate, reading accuracy of 0.16C, digital output.

As for the reason that the temp rises when the cooling is turned off, there are really two heat sources to blame - one is the core of the mirror as has already been pointed out, and the other one is the hot side of the peltiers (the heatsinks). As soon as you turn off the power the "heat pump" effect stops, and heat can flow back into the system from the hot side.

The fans that are normally used to blow the cold air off the coldplate and onto the mirror then turn into fans that are blowing warm air as the heat from the hot side gets back into the system via the peltiers.

In more recent experiments I've tried turning off the cooling in 2 stages, first turn it down to about 50% for a few minutes and then turn it off. This gives the hot side time to cool down a bit before the heat pump is turned off completely.

Something to remember is that these mirrors are made in a temperature controlled environment, where the mirror as a whole is always at a constant temperature, so the mirror will only be at it's most accurate when it is at the same level of thermal stability.

If you're using (or thinking of using) peltiers then you have to remember that they will introduce a temperature gradient through the glass, from cold on the outside to warm on the inside. You have to turn them off and let the mirror equalise for about 20 mins to 30 mins before it can equalise temperature throughout the glass, and then you'll get the sort of spectacular views that Iceman has found :-)

cheers, Bird

[bird]

Also, the aim of the game here is to end up with the mirror as a whole at exactly the ambient temperature (ie the temperature of the air next to the mirror). To achieve this you have to remember that the temp sensors on the mirror are only reading the surface temperature, and you can be sure that the core of the mirror will be warmer, so it's normal practice to cool the mirror until the temp sensors on the glass are reading a few degrees *below* your expected air temperature.

That way, when you turn off the cooling and let the mirror equalise, you end up at the right temperature overall.

In particular (in Canberra) the air temp can keep falling almost all the way through the night, so I have to take that into account as well. I try and overcool the mirror by a couple of degrees so that it will be at the ambient temperature in about 30 minutes time, after it's had time to equalise :-)

cheers, Bird

[bird]

Following up on Mikes original post, it seems from my experimental data that you have to get the mirror to within 0.5C of ambient for the boundary layer currents to stop. Interestingly Damian Peach has also posted the same observations a few months ago, also claiming that the mirror had to be within about 0.5C of ambient.

The difference that it makes is very dramatic - details that were completely invisible suddenly pop out and the image as a whole becomes very stable.

cheers, Bird

[Photon]

Hi Bird,
Do you need to fine tune the collimation when this equilibrium point is established?
Regards Ian Gillespie

[bird]

Hi Ian,

Normally I'm waiting impatiently for the equilibrium point so I can start imaging - this means I have camera gear + filter wheel + barlow all fitted into the focusser and ready to go, so I don't want to pull it all out and check the collimation :-)

I've implemented a "collimation scope" which is an old 6x30 finderscope (eyepiece removed) + my firewire camera from last year, this is such a small and stable scope that it never shifts out of collimation, so I can compare the views through it to the view through the main scope, and adjust the collimation screws on the primary to keep them in agreement. In practice this works quite well, and means I can keep the camera gear in the scope all through the session.

I can have both video windows open at the same time on my laptop, and it's easy to compare the images coming from both scopes. The collimation scope fits onto the main scope just like an extra finderscope.

An extra benefit, the larger field of view in the collimation scope makes it a breeze to find things that are close-to but not-quite-in the field of the main camera.

cheers, Bird

[Dujon]

Thanks, Ian W and bird for your answers cum explanations.

I must admit that it was only yesterday that I realised that the fan on the base of my Newtonian was moving air away from the mirror. To me that meant that most of the air passing through the fan would be from 'outside' the telescope and very little would pass over the base of the mirror. Perhaps I could fabricate a shroud to ensure that the air being moved came from within the tube, over the mirror and out the back but I'm somewhat dubious about doing that.

Two other options would to be reverse the polarity of the electrical connections or to physically reverse the fan. The latter is a problem; the screws securing the fan (which are as loose as a coward in a bun fight) are on the mirror side of the mount. Sheesh!

A question for all:

Some years ago I did a couple of repairs on what are termed 'car fridges'. These, as best I can gather, use something like the 'peltier' to which some of you have referred to produce their cooling effect. Are we talking about the same thing?

Also from memory, the replacement cost of these cooling/heating units was quite reasonable - well, the people for whom I was doing the repairs were quite happy to pay for my labour plus the replacements.

[bird]

John, peltiers have become very low cost in the last 5 years or so, now you can buy the peltier chips from places like jaycar etc for about $10, but my first peltier (bought back in about 1996) cost me well over $100...

So peltiers now show up in all manner of "portable fridge" devices, exactly like you might find running off a cigarette lighter in a car.

In fact I think that's where DaveP got his peltiers from?

cheers, Bird

[netwolf]

Thanks all, I thought it might be something like that.
So my next question is. Can we start cooling indoors during the day to keep the mirror at ambient? Pre-cooling as such? would this help?

Regards

[asimov]

My idea was to keep the whole scope at at predetermined temp. The hardest part is coming to a conclusion on what temp to have it at. If it was kept at slightly below your *average minimum* temp for your area, that means the scope would have to 'warm up' to the ambient temp. I'm not sure what the consequences of that would be though..

[bird]

In summer I've tried to do that as best I can. but I don't have enough aircon available to really keep the scope cool, or pre-cool it enough to make any practical difference. In wintertime the day-to-night difference around here is about 20C, makes the job a bit harder still.

cheers, Bird

[RAJAH235]

FWIW. Oatley Electronics (http://oatleyelectronics.com/peltier.html) also supply the peltiers & the control unit for them at reasonable cost....:D L.

[GrahamL]

I freeze some water in a 10 inch pastry dish and slip it under my 10 " dob
(upright position ).. a while before viewing .. It "seems" to cool down quicker
but to be honest I really don't know .. I was just going to do this until I got hold of some more rechargeables to run the fan thats on it.

[IanW]

You sure can pre-cool and it helps a lot. Most of the major observatories run airconditioning all day to keep the optics cool. Stand inside the AAT dome and it's quite chilly.

Large observatories (and many smaller ones) also have massive air circulation (blowers / extractors) to further cool the systems down around sunset.



You'll find the average temperatures by state and month online or you can purchase more accurate data for a fairly hefty fee from the Met Bureau.

Warming up will cause exactly the same issues as cooling down, ie: tube currents (eddy currents) and mirror distortion if using pyrex or any of the normal glasses used in mirrors (float, plate, white, BK7 etc). If the mirror is made from Zerodur, Duradur(sp), cervit or Aerosital then there will be no change in the mirrors shape during heating or cooling.

[GTB_an_Owl]

anyone ever tried water cooling the mirror?

[IanW]

Hi Bird,

Just a thought.

Have you considered going the chilled liquid route for cooling the mirror? I used such a method to cool a Medium Format "cold" camera over 20 years ago.

It's a lot messier to construct but does have the advantage that the heatsinking and radiator for the hot side of the system can be kept well away from the telescope.

On the bonus side, you could use a 2-3 stage Peltier pump.

Cheers
Ian

[bird]

I've thought about it, mainly because I could separate the cooling stage (refrigerator) from the main body of the scope and save some weight. I would probably use a jacket that enclosed the rear and side of the mirror with a gasket of some sort to seal around the edge near the front surface.

Wasn't sure what coolant I would try, some light oil or water + antifreeze. Need to be sure that it doesn't affect the mirror chemically.

The down side is that it is not as portable as the current setup, and anything with liquids is messy...

cheers, Bird

[GTB_an_Owl]

simplicity gentlemen!
items -
1 pump (maybe a reservoir)
1 length of plastic pipe

fill pipe with radiator coolant (so it don't freeze)

2 wraps of the pipe around the edge of the mirror then bring it out and two or three wraps on the outside of the tube and connected to the pump

theory -
external tube picks up ambient temperature of air/scope surface and transfers it to the mirror

comments?

[GTB_an_Owl]

Mmmmmmmm!
vibration on mirror maybe from the pump circulation?
then again, when it reached temperature you could turn it off

[IanW]

Insufficient cooling! Ideally you need to climb the temperature differential between the anticipated observing temperature and the current temperature plus a couple of degrees. It requires a lot more cooling than you'd think.

Re: follow up post on vibration.

As long as you don't have air bubbles in the system and have the pump off the telescope then it's vibrationless. It's a technique that's been used for years in professional observatories for cooling camera systems both, film (Cold Camera) and CCD (cryogenic cooling). A few commercial CCD camera in the early days of CCD imaging for amateurs used fluid cooling systems, eg: Cookbook CCD camera by R. Berry.

[IanW]

Hi Bird,

Correct me if I'm wrong, but from the look of the scope it's using an alloy/metal tube? Have you thought about wacking a Peltier cooler on the tube?

A few thoughts based on the cooling systems I used on the camera.

The use of formed copper or stainless steel tubing (thin wall) works extremely well. A coil that spirals round the back of the tube beind the mirror cell (like your current cooling plate) and then around the edge of the mirror. Gasket using thin neoprene and use swaged connectors.

For the cooling pump, a double or triple peltier pump running on to a copper (best) or aluminium cylinder with copper pipe coiled around it (10~20 turns works well), lag well with insulation. A windscreen wiper or heavy duty fishtank pump with good quality silicon tube would be ideal for the interconnections between the refrigeration unit and cooling coil. One thing that is needed is a trap/reservoir to remove air from the lines, this only needs to be a 1L bottle with an inlet and outlet pipe brought the top lid with the pipes immersed at unequal lengths in to the coolant. The outlet pipe should be at a lower level.

For coolant, I used a silicon based industrial coolant which is probably overkill, something like Dynalene HC50 or possibly Paratherm HC etc would do the trick. Avoiding water based products unless they have a good corrosion inhibitor in them for obvious reasons!

Of course the ideal way to dispose of the internal temperature gradient of a mirror would be to fit coolant lines inside the mirror, something that would not be impossble to do with a mirror blank and some of the current hole drilling technologies.

Cheers,
Ian

[iceman]

Excellent discussion going here, it's definitely going to help me when I start on this peltier cooling route.

I'll be starting the project soon and will seek advice in the next few days.

I've got a few months before Saturn is back high enough, so I can get it right over the next few months and practise on the moon.

[bird]

Ian, I've thought about these possible ways to fit water cooling, but I would have to find that it provides a substantial improvement over the simple peltier-based cooling before I'd want to try it, based on the KISS principle (Keep It Simple, Stupid!).

My rough and ready peltier rig that I have now seems to give a suitable amount of cooling, enough that I'd probably want to try to improve on it rather than switch to a different design when I want more cooling.

But I do think about water cooling from time to time. Maybe one day I'll build a test rig and see how it checks out.

cheers, Bird

[Lester]

Hi everyone,

interesting disscussion. I use Ice packs held around the rear of the OTA of my 14" SCT with ocy straps. Tried a peltier but it needed a lot more than 1 for this sized scope. Five ice packs brings the OTA down to below ambient in 2 hours. I don't get the freezing conditions that some of you chaps get. Usually only gets down to 5 degrees minimum.