Arduino Temperature Project

My Arduino temperature probe is about done. A small Arduino Uno uses two TMP36 temperature sensors to log temperature. The temperatures are retrieved by the newly written ACCycle.exe written in VB.NET. The previous version was an HTA, but Microsoft has kindly removed access to COM ports (MSCOMM.OCX) in Windows 8 so I needed to rewrite everything.


The above image from the Adafruit site article summarizes the connection to the TMP36 sensor. Very easy.

I wired two 1/8 inch stereo jacks for the sensors. One sensor is on a very short (4″) wire, the other is on a longer (about 6′) cable. The plugs go into two jacks mounted on the side of the project box.


Wiring of the 1/8 Stereo plug and jack. The sensor is inserted into the TO-92 socket; it can be pulled out if needed (maybe it will fry itself or something?).

The project box was $1 from Walmart, found in the Business items section (pens, stationary, tape, etc). I cut a hole for the USB port and drilled holes for the jacks and Arduino mounting screws.

The project box with Arduino and jacks. The two probes are shown plugged in.

The project box with Arduino and jacks. The two probes are shown plugged in.

The Arduino checks every second for any characters transmitted on the Serial line (USB). As seen in the image, the small yellow LED lights when the serial input is checked. If it receives a “T” character it reads the two sensors and returns a string of the form

196.07 192.01 F 91.15 88.90 C 1.41 1.39 V

The first two fields are Port 1&2 Fahrenheit temperatures; the next two fields are the Centigrade values, and the last two are the actual voltages read.

Main screen of ACCycle, showing 3 operational modes: Temperature Control, Timer Control, and Manual Control.

Main screen of ACCycle, showing 3 operational modes: Temperature Control, Timer Control, and Manual Control.

Options screen for ACCycle

Options screen for ACCycle

The ACCycle program displays the two probe temperatures, and optionally uses one of the probes to turn the Air Conditioner on/off.

The temperature data is logged into two files:

  1. A Boltwood one-line file. I feed this to Weather Display so I can track the temperature of the controlling probe.
  2. A log file of the temperatures in csv format so I can plot the temperature over time in Excel.

If someone is interested in the source to either the Arduino sketch or the ACCycle vb.NET program, drop me a line at eridanibrew – AT – gmailDOTcom.



Hallelujah, Unguided Tak Images

I am kind of afraid to log this, it may jinx the system…

The Tak is successfully running 15 minute exposures unguided!

I can’t really say why it is now working. I tightened down the clamshell holding the Tak a bit, otherwise nothing unusual. After completing the insulation and drywall project, I ran a new Tpoint model and tried things.

First I wanted to guide using the Edge 11; this has worked in the past. However, it wasn’t working – the guiding seems to be working well, the guide star stays in the tracking box, reasonably small guide errors. BUT, there seems to be a bit of flexure between the Tak and the Edge? Stars are slightly trailed in 10 minute exposures. ProTrack is off so it doesn’t interfere with the guiding.

Just for fun, I tried a 10 minute exposure with ProTrack. Say What? It actually worked! Even 15 minute exposures are working well. Longer than that is not necessary, my camera theoretically works efficiently with 10-15 minute exposures.

It has been running great for about a week now. I am nervous about switching systems between the Tak and the Edge. I have a separate model for the Edge, but maybe switching the systems will mess things up again. We will see.

 Tak Model

The Tak model has 206 points (some are thrown out, of course). The final Sky RMS = 12.7 arcseconds; excellent, but no different than values I had previously.

Tak 206 point Scatterplot

Tak 206 point Scatterplot. Click for larger image

Tak SuperModel terms

Tak SuperModel terms


Tak SuperModel graphs. Click for larger image


Tak Orthographic errors. Click for larger image.






Cooling the Observatory

I am starting to get more concerned about the temperature in the Dome during the day. It gets crazy HOT here in AZ.

Step 1 – installed a tarp over the western wall where the sun is hottest. The tarp attaches to the top of the wall (a few inches above the top edge, leaving room for the dome lower shutter to go over), then slants out a couple of feet to the ground. Attached the bottom to the ground with stakes, with big bricks on top to keep the wind from pulling up the stakes. The tarp appears to drop the wall temperature 5-6 degrees, and blocks the UV.

Step 2 – I decided to insulate and drywall the observatory. I should have done this in the very beginning, of course, but didn’t realize just how hot it was going to be. I had expected the window air conditioner to be able to handle the heat:(

Removed all the equipment except the telescopes. Put giant plastic garbage bags ove the telescopes and the mount. Pulled the pegboard off the walls, installed insulation both above and below the floor. Did this part at night so it was somewhat cooler – only 90 degrees:) Then put drywall on the walls, both above and below floor.

I installed a Vernier sensor device (LabPro) and the associated software on the observatory computer so I could monitor the temperature. I am measuring the temperature on top of the electronics cabinet, about a foot below the mount. There is clearly a strong temperature gradient from the floor to the dome. The window air conditioner is situated just above the floor, so the cooling mostly wants to sit on the floor and sink below it. I have a small fan below the air conditioner to help blow the cool air up into the room. This helps a bit.

Here is a temperature graph from before the insulation project:


You can see the overnight temperature of 73 rises quickly at sunrise. A little after 10AM the air conditioner was turned on, with the temperature about 92. The temperature drops to 87, about 5 degrees. It the continues to climb, although at a slower rate. The temperature peaked at 4 PM; the AC was turned off about 10:30 PM. Notice the squiggles aroun 9-10 PM; this is when the AC “cycles”, where it just blows for awhile, then actually cools awhile, then goes back to just running the fan. You might think this is oscillating around the AC setpoint, which is 72. However, the temperature does not seem to be close to that setpoint. I see this pattern show up every so often.

Her is today’s data, with the insulation installed.


At 10 AM the AC cut in and dropped the temperature a couple of degrees, after which the temperature continues to climb. The cycling of the AC is very evident.

At 1:30 PM, I went out and moved the computer cabinet and repositioned the small fan. I had cleverly positioned the cabinet so it “blocked” part of the AC vent, forcing cool air to flow through the cabinet (the cabinet has screen doors for air flow). I moved the cabinet so the AC vent was unobstructed. I also repositioned the small fan so it was blowing more upward from below the AC; it had gotten knocked to the side a bit.

These changes caused a) an immediate drop in temperature of about 7 degrees, and b) cessation of the cycling of the AC. I wonder if the cycling is removed by the small fan more efficiently moving the cool air away from whatever sensor the AC uses to measure the temperature.

Clearly the insulation project has not resulted in a room that can be controlled at a steady temperature by the AC. The dome is a major heat source which the AC cannot overcome.

Wall Temperatures

I took my little Infrared temperature measuring gun and checked the various wall temperatures, inside and out.


This was done about 1:30PM, so the East wall has started cooling from its high and the West wall has started heating in direct sun.

The interior wall gradient is pronounced, with temperatures going from 93 near the upper floor up to 112 on the interior dome surface.

The tarp is doing some good, lowering the temperature several degrees (and blocking UV). The metal door is absorbing a lot of heat.

The Celestron scope measured 101, while the Paramount measured 93.