Hey guys, had a chance to lurk today and thought I'd comment on this.
First of nice work looking very good. There seems to be a bit of talk about power draw and MOSFETs so I thought I'd add some advice to the selection:
Firstly you're measuring humidity to calculate dewpoint right at the case of the electronics. If you get hot enough to need a heatsink then this is a really bad MOSFET choice. One-Two degrees of heat passing through the case to your DHT22 will throw out your calculations, and self heating prevents the DHT22 from reading 100% humidity so your dewpoint will follow the temperature and things just don't work (found this out the hard way).
So MOSFET selection. This SHOULD be a non-issue, but the IRF520 is not too good of a choice here. For a dew heater the duty cycles are very slow so a MOSFET will not spend any time in the active region, it will either be on or off. The active region becomes a problem when you're switching in the 100kHz kind of region, but for this application one of the most important numbers is R_ds(on) for the MOSFET.
The IRF520 has a Resistance Drain - Source in On state [R_ds(on)] of 0.27ohm. What this means is that when the power is being applied to the heater the mosfet looks like a 0.27ohm resistor. Drawing 2A through your heater? That's over half a watt for the MOSFET. Thermal resistance Junction to Ambient [R_thJA] for the IRF520 is 62 C/W so the junction of the MOSFET will be 30degrees higher than the ambient temperature. This isn't an issue for the MOSFET, but if you enclose it into a small black box, maybe add a second one for another dew strap you'll quickly turn your box into a heatsource right next to the temperature measurement.
So for ideal component selection rather than looking at the Amps D-S rating, instead look for a low R_ds(on) rating.
If you feel inclined to experiment more you could look at the part I used in my little dew heater controller which was an IR3315 High-Side Switch. This has a couple of awesome advantages over using an N-channel MOSFET:
- It has an R_ds(on) of 0.02ohm so an order of magnitude better than the IRF520
- It is high-side switch so it switches off the VCC line rather than switching off the Ground. Using an N-channel FET as a low-side switch is not an issue for a dewstrap ... unless you have a short circuit. On a low side switch there's always 12V on the output, so a short circuit to the telescope itself (most likely place, and usually grounded) is a short circuit even when your circuit is off.
- Speaking of short circuits using a single resistor the IR3315 can be configured for current limiting, so a short circuit won't damage anything even when you crank the dewheater to 11
Just some free tips
. It's a very nice and simple build all the same so thanks for sharing it with everyone.
(edit: Poo! I just forgot one very important detail. The IR3315 being a high side device needs 12V to turn it on. I installed a MOSFET driver before it, a TC4426. That complicates things.)
Also David for resistor wattage it's a simple calculation:
Volts (V) = Amps (I) * Resistance (R).
Power (P) = Volts (V) * Amps (I)
Put them together:
P = V^2 / R.
P = (5*5) / 4700
P = 0.005W so a resistor rated to 0.25W is more than adequate, and as a general rule of thumb typically every time you see a resistor in any kind of digital signal (e.g. 5V output of an Arduino) vs power (+12V side) service it's almost universally okay to select the lowest power rating.