RCTL Technical Details Section
UPDATES (newest first) and TECHNICAL FINDS/NOTES:
Download the RCTL v.1.0b manual in .pdf format.
- 1/29/04 - While the original RCTL is still being used with great results,
Project Chinook is seeking to expand on the RCTL idea and gain even more useful information about the helicopter!
- 7/2/03 - IMPORTANT RESEARCH RESULTS: the location of the sensor on the cylindar head has a HUGE effect on the temperature reading!!! In the raptor, I have had the sensor mounted on the exhaust side of the cylinder for all the tests until about 6 flights ago, when I moved it 180 degrees to the opposite side of the exhaust (I had heard that the Venom temperature sensor should be placed here, so I thought I might give it a try). The temperatures were about 40 degrees F less! No wonder I was lacking in power - I was rich, since I was tuning based on the temperatures on the exhaust side of the cylinder. I know I shouldn't be tuning based on temperatures, but I really didn't want to hurt the engine (based on what others say is proper engine temps). I have since re-tuned (main needle out about 2 7/8ths) for more power, at an ave temp of 215 on cylinder, opposite exhaust side. It should be noted that the exhaust side of the cyinder was initially chosen based on the theory that its temperature would change more rapidly (nearer to the exhaust gasses. This was true, as the other side takes longer to change its temperature. SECONDLY: the temperature has been observed to rise about 15 degrees after the engine has been shut off after a flight. It then peaks, and drops quickly. I hope to have a screenshot of this phenomenon. Keep in mind, that if using a Venom temperature monitor, it might record this "peaking" as the highest temperature of the flight, when in actuality, the flight condition of the engine is cooler. The RCTL manual has been changed to reflect this better mounting location.
- 6/9/03 - Added link on the Sales page to the RCTL v.1.0 manual in .pdf format.
- 5/9/03 - Updated tech details with pictures: my own heavily used prototype mounted on the Raptor, plus a photo showing a proposed mounting method. See below.
- 4/26/03 - Flight today. Continued testing of new settings. Last flight shown below. It had GREAT power. I feel I don't want to go higher in temps than this, since I've got the power I want. Oh, and the new plot style (no boxes) is just a setting on the TI-89. I like it better this way.
- 4/26/03 - (flight date sometime early in April).A major re-adjust to the engine. It has over a gallon through it, so I wanted to set it up for good power (the normal settings). For break in, the low/high needles were at 2 1/4 and 4 1/4. This offered cool temps, but there was NO power when climbing!! Simply leaning the top caused the temp to go up rapidly so I couldn't get it to the recommended 3 turns... I reset the low needle to 3 turns, re-adjusted the throttle idle settings (it was pouring fuel in idle!), and set high needle to 3 turns. Much better performance,a nd fairly cool temperatures!! Finally, I have an engine with more power than my old OS32, and at lower temps, too (I may lean a bit more). Here's a temp log. Note the increase in temps from 200 to 220 over this last 5 minute interval. It may have been more of a function of me stressing the engine more in the later part of the flight, as opposed to tank lean-out. Other graphs at the same outing (not shown) do show a leaning due to fuel level, but it is not worrisome-high (10 deg, or so, IIRC).
- 3/28/03 - More flights. Re-covered sensor w/ heat shrink (it had split). Excellent results. Temps were same as measured w/ Raytec gun. Average around 200.
- 3/23/03 - Prototype test in helicopter successful! Graph screenshot shown below. Flight was full tank, with hovering and fast forward flight mixed in with climbouts, etc.. Engine temperatures show a consistant CHT of around 170 deg F which seems good for a rich break-in period. It is interesting to note that when the engine was shut down, the temperature appears to have risen, seen in the "spike". The y-max and y-min of the graph are 270 and 150 with y-scale of 10.
- 3/22/03 - prototype test in helicopter succussful! Test conditions: Raptor 30, TT39 engine with break-in settings, windy (approx 50 deg. ambient temp) and rainy. Sensor mounted between head and exhaust port, on the cylindar. RCTL operation flawless. Below is a graph screenshot of a short test flight with an adjustment stop in the middle (dip down in temperature). The y-max and y-min of the graph are 270 and 150 with y-scale of 10, FYI.
Components:
- Microcontroller - This was chosen to provide on-board ADC, use FLASH memory for future firmware upgrades, be cheap, and as small a package as possible for easy assembly. The Atmel Tiny15L fit the bill. It is used in an 8-pin DIP package, and has a 10-bit ADC onboard.
- Voltage Regulator - Aa +5V regulator is used to provide a very constant supply voltage to the microcontroller and the resistor bridge feeding the ADC.
- Thermistor - The temperature sensor is a thermistor. The current prototype uses a 100K. It provides a workable range, about 0.5 deg resolution from 170 to 250 deg F (the calibration equation uses this range).
- Other components - various resistors (1% tol. for bridge resistors), 9V battery connector, 2.5 mm stereo jack for link cable, 2-pin connector for sensor, micro-pushbutton for mode select, power LED, mode LED.
Theory of Operation:
Upon powering on, the RCTL selects which mode to go into depending on the state of the "mode select" button. If pushed while powering on, recording starts. If not, it waits for a button push to start transmitting data.
In record mode, every 5 seconds the microcontroller reads the voltage difference between a reference voltage set at 2.5 V and an input voltage that varies with temperature (the voltage at the thermistor/resistor junction). This value is stored in one of 60 EEPROM bytes inside the chip, and hence won't be lost when power is turned off. After 60 readings have been taken, the oldest reading is replaced by the latest. The result is that the last 5 minutes of recorded values are kept.
In transmit mode, the RCTL waits for a button press to start its transmission. The computer or calculator receive program should be put into the "waiting" state. When the key is pressed, the microcontroller sends each byte to the receiving device. The receiving device then converts these raw 8-bit values into temperatures based on a linear calibration formula. The user can then view the measurements as raw numbers or graphically.
Test Results
Below are some test results an other such data.
Temperature vs. ADC value line fit & measured data points:
This test was done in the following way: a sensor was mounted to a model engine. The engine was heated up to approx 270 deg in an oven. Temperatures were recored with a Raytec MT-4 infrared temperature gun, along with the ADC value from the microcontroller. An analysis in Excel gave the following: Intercpt = 134.7, slope = 0.657, R^2 = 0.9956, Std Error = 1.24 .. In summary, very good results are obtained. It is true that these values vary depending upon the exact mounting, thermistor variablility, etc.. but it will work for the R/C applications.
Calibration note: The proposed method of testing the RCTL's calibration is a "boiling water dunk test". This consists of dropping the sensor in a cup of boiling water (or had JUST been boiling) and recording the results. Preliminary testing suggests that the readings should give a temperature very close to the actual boiling point of water at the testing altitude.
Prototype #1 (my own):
Below are a few pics of the first prototype to be built off of a breadboard. The first scan is the latest model that is being flown in my helicopter with great results. The next few are of the same board before it was made flight-ready. The last two are the first (mostly finished) prototype that will be sent out to a beta tester. All the major components can be seen. Notice the small size and small number of components! The sensor mounting arrangment shown in the first picture has worked well during testing, but is a little too complex, still. Hopefully, an easier solution will be found that is just as reliable. A 9V NiMh battery is rubberbanded under the board for this prototype.
Calculator screenshots:
Here are some screenshots of the software that runs on the calculator.
Here are some new photos of the prototype mounted on the heli. The last photo attempts to show the proposed mounting method clearly. ! added May 9, 03 !
