Wanted,Temperature operated cowl flaps

[John McNamara]

I have ordered a servo tester and will have a play with it, I have also ordered an Arduino as a backup plan.
I suspect that the mention of Arduino witchcraft might sound horribly complicated, it's not but gives an easy way of setting the fully open and fully closed temperatures. With the simple servo tester modification the closed position of the servo will vary between summer and winter due to the ambient temperature.
I am guessing that fully closed at 20c and fully open at 100c would be a good starting point.
Both methods look like costing about a fiver to make

Hi Steve,
Are those figure realistic or are they a "guesstimate"? On a summers day with the sun on the cowl I imagine the in cowl temparature may be 25 degrees plus. Then again, the draft through the cowl may bring that down to the 20 degree you propose. I don't know myself, I am just speculating. Without a means of mesuring and logging the in cowl tempaprature vs the cylinder head temparature, I think we will be guessing what the numbers should be. Its not a problem for me to do this with Jeti, and the software can drive the servo to open and close the cowl given sensible temparature parameters, but how will you get the data from the ardino set up? Can that be done?
John

[Stuart Atwill]

Sounds great but way beyond me, tweeking other peoples codes; I'm lost!
I am about to retire so may have more time to 'bone up' on electronics and Arduino.
I forgot to mention my crude adjustment of flap opening sensitivity. I just moved the temperature sensor closer to or further away from the cylinder head, simpels!

[Steve Perry]

I am keeping it simple and not converting things to actual F or C temperatures, the sensor will go in the airflow somewhere near the engine with an adjustment for how hot is fully open, an override switch that can be activated by throttle position or any other method to activate it
There will be a manual override switch for use when setting it up to send the servo to the fully open position, also a servo reverse jumper, pull it out if the servo moves the wrong way.
I am going to work on 90 degrees of servo travel with adjusters to give up to an extra 10 degrees in both directions to make it simple to set up.
The last parts to test it and allow work on the software should arrive tomorrow.

Any other needed features I have missed then please suggest them

I could lash it up by just soldering wires to an arduino nano ( aprox 40mm X 20mm ), but solder wire connections are prone to breaking due to vibration
I have designed a PCB to take an arduino chip that uses standard plug in RX servo connectors to avoid solder wire vibration issues
It might sound expensive to have a pcb made, not so, but there is also the postage from China.

JLCPCB will make 5 100mm x 100mm PCB's for $2, my PCB is 33mm X 50mm ( I can make the smaller but am working with what is cheap and easy ), so I can get 30 PCB's for the $2, 33mm x 3 and 50mm x 2

Due to the current virus situation normal postage from China ( about 15 quid ) is erratic to say the least as airmail goes in the hold of passenger flights that are nearly non existent DHL on the other hand have their own cargo aircraft but costs an extra tenner but I can order on sunday and it will be delivered on thursday !
I am certain that more than one person will eventually want one of these so the development cost can be spread over the 30 PCB's, estimated cost from me is a tenner

I will be talking to a friend with a 3D printer about making cases for them as an add on

For the type of adjusters just google "bourns 3296", nothing more complicated than a small screwdriver to set it up

I suspect there are Arduino experts reading this that could do in seconds what it will take me a few days to do

[Steve Perry]

Looking at the example code https://www.allaboutcircuits.com/projects/servo-motor-control-with-an-arduino/
We can see that first we tell what is connected to the pins such as " int potentio = A0; // initialize the A0analog pin for potentiometer", the adjustable output from the pot is connected to the A0 pin

The most complicated line is " angle = map(angle, 0, 1023, 0, 179); // scaling the potentiometer value to angle value for servo between 0 and 180) ", this takes the voltage on pin A0 that is measured on a scale of 0 to 1023 and converts it to 0 to 179 degrees

Using the same instructions I can connect the upper and lower travel potentiometers to other pins and scale to something like 0 - 10 and add or subtract them from the 180, I can subtract 90 first to give 90 plus and minus the 10 up and down from the adjusters

It's a different language but easy to understand and probably years to master

[Steve Perry]

I have just realised that the temperature adjuster needs to be not on the PCB but mountable somewhere where it can be adjusted with a screwdriver wings on and engine running

[Stuart Atwill]

I had a look at the website, looks promising but beyond me!
On my current setup I just moved the temperature sensor closer to or further away from one exhaust header to "calibrate" the cowl flaps but servo travel adjustment would be useful.

[Steve Perry]

I thought an adjustment for temperature would be easier than removing the prop, removing the cowl, move the sensor, refit cowl, and refit the prop

[Stuart Atwill]

Indeed it would make it much easier, very sensible.

[Steve Perry]

I have been looking at this from the wrong direction by trying to control a servo and adjust it to the temperature
It becomes easier if it measures the temperature and adjusts the servo, if lower that the set temperature move the servo a notch in the closed direction, if higher than the set temperature move the servo a notch in the open direction
Fully closed / open set by the adjusters, just keep the servo within those limits
Temperature involves some nasty maths, but other folk have already done it for me, such as
https://www.electronicwings.com/arduino/thermistor-interfacing-with-arduino-uno ( Don't even try to understand it ), but it gives an output in centigrade
The sensor that I am using is rated up to 100C so I can just scale the output from the adjuster to give 20 to 100 and check it against the actual temperature of the sensor
This gives the possible option in a later version of being able to plug in a £5 display on the ground to show the temperature that the sensor is reading.

Another issue that I have to deal with is that an arduino wants no more that 5 volts but has to be fed from a typical flight battery that vary from 4.8 for 4 x nmih and upwards.

Much tinkering, testing, looking things up on the internet, and head scratching is going on here

I am certainly not going to make my fortune with these, so once I have sold enough to get my development costs back I might make the pcb design and software available for personal use should anyone want to make one beyond the point that I stop making them

There may be someone in China reading this and thinking that they can knock out a thousand of them at £1 each inc postage, in that case I will also want £1 each from them so they would have to sell them for £2

[Glenn Masters]

I think it will be the cowl gill hinging kit and operating mechanism that will cost the money - not to mention the hours of model making to make it happen. A good project for its own sake though ? After all that work I think I would want an overide by-pass control to amaze my friends with because, unless the model idles on the ground for a good while in hot weather , all the good stuff will happen in the air some distance away ?

[Steve Perry]

As previously mentioned there will be an option to plug in an external switch that will switch between auto and fully open

[andy Meade]

Not sure if it'll help, but I saw this recently when researching the CARF Jug
https://www.youtube.com/watch?v=mC9z6QMkgFo

[Steve Perry]

I have almost got the code finished, of course when I test it there will be something wrong in several places lol
At the moment the code will slowly open the grill / flaps to match the temperature of the sensor until they are fully open, the other option is Bang Bang where they stay closed until the set temperature is reached and then switch to full open.
Bang Bang will cause issues at lower throttle settings where they suddenly close and then open again once the reduced airflow has caused the engine to heat up again, then cool it until they close again, etc.
I can add a jumper to switch between the 2 modes if required.

[Stuart Atwill]

Wow, I didn't think this would create so much interest, all I can say Steve is that I am totally lost with the tech bits. Personally I would not be interested in a bang-bang system, sounds too much like my original radio gear before proportional become common. First was free flight glider then Diesel power, at last we had radio gear. The Tx had a red button on it, push once and hold for left, twice and hold for right, push three times and hold to move the Diesel engine carb one of three positions! Boy, we have come a long way from there.
As Glenn says, an override switch would be great. The cowl flaps take time but highly detailed models take lots of time as I always build from scratch but they cost only a few pence extra, just a strip of Mylar and odd bits of scrap.
Can't wait to see the final product, thanks so much Steve for your enthusiasm.

[Steve Perry]

I will forget about the bang bang option
I am currently trying to make it fail safe, if the temperature sensor breaks or shorts out then the flaps need to open fully and not close
Since I will get quite a few circuit boards for $2 plus postage I want to get all the needed features built into it before ordering them.

A better description of how it works is that the flaps will open just enough to keep the sensor at the preset (adjustable 20 to 100 degrees) temperature, if the temperature cannot be maintained then the flaps will stay at fully open until the temperature drops.

[Stuart Atwill]

You seem to be covering all the bases!
I spoke to Mike at Model Radio Workshop as I have a suspect expensive servo. I mentioned the cowl project to him and he told me that a couple of years ago he was approached by one of the power boat boys. They would love to be able to control the flow of water in the cooling system to quickly bring the engine up to optimum temperature and keep it there, could be a market there to help cover costs.
Any ideas why my Futaba BLS156HV wing flap servo started jittering one time when I turned my radio on? It worked perfectly after that but I replaced it anyway and have been rather reluctant to fly the plane since.

[Steve Perry]

Servo jitter can be caused by the potentiometer in the servo.

[Steve Perry]

I have added 2 self resetting fuses, one on the thermistor temperature sensor so that it will disconnect if the connection to it shorts out and trigger the flaps to open fully, the other is on the supply to the unit so that if something goes wrong it will not flatten the flight battery, the flaps will stay where they are rather than going fully open if this happens.

[Glenn Masters]

Lets hear some more about the actual gills and how they will operate ? It will be tricky to create the mechanism for reliable use thats not over complex. I know the fullsize F4U uses a cable that tightens to close against sprung open gills. . . this would result in a lot of moving parts within a stressful, high temperature / vibration prone area . . . whats your ideas guys ?

[John McNamara]

Hi Guys,
For those of you tempted to consder the Jeti route, here is a video from my good friend Harry Curzon showing how a servo controlled cowl flaps can be controlled using a temparature sensor inside the cowl, or on the cylinder head as you wish. Nothing else is required in the model. You will be able to measure temperatures in real time as well as log them in the transmitter for analysis later. You can then decide what numbers are sensible to set in the transmitter for opening and closing the cowl flaps. The method used here can be applied in all Jeti transmitters as the software is common to all models, only the number of software options changed with the Transmitter type, from DS 12 up to DC & DS 24
https://www.youtube.com/watch?v=uxdis3hDw9w

www.jettstreamuk.co.uk
John