Monday, November 9, 2015

Data dump: Dew Heater Edition

I have a number of projects going on right now, which I will hopefully have more to say about later on. Right now I think it's important that I get a few pieces of information and findings out there on the internet both for my own benefit and for other people who might be working in similar things.

First: Here in the deep, deep, swampy south the dew point outside is typically one degree or less below air temperature. This means anything outside that has a view to the open sky will build up dew on it within a few minutes of exposure (the physics of radiative heat loss, maybe another post on another day).

In my case I'm concerned about my telescopes, which means I want heating but not enough to cause visual distortion: air of two different temperatures creates a lensing effect at their boundary, think ripples of air over hot asphalt on a summer day. Remember this?

That's colder/lighter gas distorting the air. Supposedly this happens significantly enough at 2 degrees F or less, but I don't have good data on that. Either way, the right answer is to keep a telescope as close to the air temperature around it as possible.

The plan is to make a nice microcontroller circuit that measures the air temp and scope temp, and if the scope temp is low turn on a heater. This doesn't need to be terribly quick or smart or use PID since there's a lot of thermal mass involved, and I've designed the heaters to be very mild.

I collected up most of the parts to do this, but a good test night rolled around first so I grabbed a cheap temperature controller I got from Ebay a while back just to get the heater kicking. Finding information on setting this up was difficult (it probably came with printed instructions, which don't last long in my possession). So here they are, if you have a cheap Chinese controller that looks like this:

Mine is labeled W2020 temperature controller, but let's assume it's China and there are 300 model numbers for the same product. Pressing "Set" once will let you use the arrow keys to change the desired temperature. Pressing and holding it for 5 seconds will enable a menu, which works like this:

CodeCode descriptionSetting rangeFactory setting
P1Hysteresis setting0.1-302.0
P2Highest setting upper limit120120
P3Lowest setting upper limit-55-55
P4Temperature correction+10~-10C0
P5Delay start time0-10minutes0
P6buzzer switchon/offoff
P7Value holdon/offoff
P8Restore factory settingson/offoff

Change P0 to H since we're heating an object, and P1 (how far above/below desired temp it kicks on) to something nice and low since we want to keep it pretty close to our set temperature (in fact the temperature will likely spike above our set a little bit because the heater wire has some thermal mass).

Oh, and the hookups on the back: NTC is the temperature probe, OUT is the heater (neither of these are polar since they are effectively resistors) and IN is power supply, with the little "o" denoting the positive side.

There are commercially available telescope products to do all of this, but being the telescope world they are stupid expensive and probably don't have a lot of active development going on with them. Sorry astronomer friends, but that's the world I can see from here: once something "works" the discussion is done, even if it means using parts that haven't been produced in the last decade. Also for the disciples of Carl Sagan the physics of things is not often considered too deeply.

My own complaining aside, the groundwork is solid. The internet tells me commercial heaters are about 0.75W/inch of circumference. This results in a heater that will just barely get warm to the touch, but seem like it isn't even working when stuck to a cold metal telescope. If you're lazy on math like I am and using a 12v power supply:

amps * ohms = volts
amps * volts = watts
desired resistance (ohms) = 0.75watts / inches
desired resistence @ 12v = 192 ohms/inch of telescope circumference
Adjust if you're in a place that is freezing cold and humid, or if you're insulating the thing to start with, or not using 12v or whatever. I'm only here to solve my own problems.

I grabbed some rolls of nichrome laying around the lab (doesn't everyone?) and came up with nice long wind of 18 gauge 80/20 wire. Pro tip: other websites will tell you to make a mess of parallel heaters or funky ways of staggering your wire to come up with enough length for a given resistance. Easier way: buy a few different thicknesses of wire, thicker stuff will end up with a longer length for the given resistance. If you have a small bodie'd thing like a spotting scope you probably want 26+ gauge, bigger scopes might be 16-20 gauge. Note that having too short of a run also means you will be putting all of that heat in to a smaller area, risking damage by making a miniature toaster coil for your precious equipment. Get the right wire for the job. It will make your life easier.

I also didn't like the idea that the commercial heaters only heated the very front of the scope: the whole scope gets soaked here, which I don't want it dripping on to other electronics, rusting, mildewing, or just generally being gross to carry back in to the house. There's also some general nonsense about trying to radiate heat back across the glass from in front, which sounds nice except for the body of the scope (and therefore the back of the glass) having a big thermal gradient as a result. Keep the body of the scope, inside and out, at air temperature and things will be much happier:

Kapton tape is cool stuff: extremely heat tolerant (we use this stuff on the heater part of our 3d printer, which spends hours at 260+C), durable and thin. Once I had figured out about how many winds of wire would go around the scope I marked out the appropriate distance between them and laid down a couple layers to electrically insulate the wire from the tube. I wrapped the coil and a top layer of kapton together leaving just a couple of ends exposed for connection. You can see that by having some 14 feet of wire to provide the appropriate wattage I had a lot more flexibility than "just the tip" heaters like astrozap, kendrick, dewbuster, etc. Plus my cost was $10 for the controller and 'it was collecting dust' for the wire, say $10-20 if you had to buy it (you will have a lot leftover if you do). A roll of kapton tape is only a few dollars, and you should have some just for the fun of it. They sent that stuff to the moon, you know.

The test was pretty simple: I set the scope up outside and let it cool down until dew formed on the front glass (less than 10 minutes). This was a rare night in which outside was colder than inside (everyone else talks about letting the scope cool down after being inside...I get that for a couple months each year here). I turned on the heater and 5 minutes later the controller shut off: the glass was clear. Yeehaw. I left it outside for a few hours and the heater would click on and off every few minutes, scope never showed a hint of dew again.

The upside is that it does the trick, the downside is that I have to manually set a temperature. Next step is making a real controller to automate an accurate temperature setting.

No comments:

Post a Comment