Hi!
I'm working on a project that will ignite a propane burner.
The pilotburner is ignited by a coil. The pilotburner has an integrated thermocoupler. I use the signal from the thermocoupler to decide if the pilotburner is burning or not. If it's not burning, the arduino activates the coil so the pilotburner ignites, or re- ignite(If the temperature of the thermocoupler falls below defined temperature).
The arduino communicates with another arduino 20meters away via RS485.
Reading of temperature from the Thermocoupler is done via a shield designed for this task.
Everything works, but when the coil ignites I believe there is severe disturbance from the coil and the arduino stops working. When the coil stops igniting, arduino is running normal again. I inserted an relay that cuts the signal from the thermocoupler while igniting, and that helped a lot(without the relay, the arduino was damaged).
Maybe someone has experience with this problem and give me some hints to what I can do for improve the design.
I attach a connection schematic(sorry that I only have Microsoft paint to draw in), but the drawing is correct according how the elements are actually connected. I also attach the sourcecode.
I ran into this problem where the igniter for a propane heater was resetting the CPU. In that case the signal was being coupled wirelessly and we were able to avoid it by rerouting cabling.
Looking at your diagram it seems that in your case it may be coming through the thermocouple wires and back to the arduino, especially if the arduino is being damaged from the event. So decoupling them with a relay is probably the best solution. Wasn't the thermocouple shield also damaged by this?
Thank you for your reply and tip. How did you transfer the tc signal wireless?
It was luckily only the usb /comport that was broken on the arduino, so I just replaced it with a new one after adding the relay for decoupling the thermocoupler signal wires when the coil is activated. After this modification, the arduino seems to work as it should even when it "resets" when the coil ignites. Even so, this is not a solution which I can accept since the arduino needs to continuously communicate with the maincontrol via rs485 since this controls solenoid valves etc.
I believe that the thermocoupler shield is well protected since it's still working without problems.
I guess I didn't explain that very well I meant that the noise from the igniter was coupling through the air to the CPU: there was no direct connection between the two.
That said, rerouting the connections was the immediate solution to get a prototype working. I think we also added rf chokes on the incoming power lines to the controller since the thought was that the rf was probably radiating and coupling into the power so we suppressed it there. I don't recall having any trouble with it after that.
If that's the case my guess would be that's where the high voltage was entering the system. MOV's on the USB lines may be helpful, but at the same time, they may affect the USB transmission... don't know.
I'm not sure how helpful a flyback diode will be here since the entire point of the ignition coil is to generate a high voltage via flyback in order to create a spark.
The matter of "cable dress" - how you arrange the wiring - is an almost universal and totally critical factor in all of these "Why won't my project with wires all over the place work?" problems.
As a boy in the 1950's living in rural Appalachia, we instantly knew an automobile was headed toward our house because the automobile's ignition could be heard on mom's AM radio. The radio would also warn of storms rolling into the piedmont as the dark clouds rolled across the mountain range.
Automobile ignitions systems have been improved as well as noise elimination on AM radios; these days, AM radio is essentially interference free.
I believe your issues to be entirely electrical noise related. Putting a "fix" on any single problem is extraordinary difficult as high voltage RF shielding and analog/digital low-noise techniques is somewhat of a lost art these days. Amature Radio publications once were great sources for tips on these black-arts. ARRL paper project books were invaluable in offering options for stubborn problems.
I'm not going to try to fix your problem, but I will refer you to a couple of short reads to provide a vocabulary and working concepts:
The goal is to contain the RF such that you do not "swamp" the analog/digital electronics with induced voltages and partially raised ground potentials.
If you know an Amature Radio builder/operator, I suggest a couple of quality hours discussing the interference subject to be well spent.
That's a fascinating idea! How exactly would you perform "noise elimination" on an AM radio?
On current engines, the ignition coil itself is mounted directly over the spark plug (or sometimes immediately between two) so it is effectively shielded, there are no longer any wires to a distributor.
HI!
Thanks for the tip. Should the shield of the cable be grounded in each end? I ask, because I think I remember something I read an article once discussing this matter(the article was related to RS485/serial communications).
The bypass capacitors: Do they have to be calculated for each application/circuit, or is there a value/size/type that is most likely to be adequate for my application?
Thank you for your comment. With «cable dress», do you mean holding high voltage wires away from low voltage/signal wires ?
Will it be enough to hold them apart from each other with some centimeters or do they require own âshieldedâ cable canals? My enclosure box which all components is mounted in is rather small, so the distance between the high volt and low volt wires isnât more than maybe 5-10 cm. The coil is mounted inside an own small steel enclosure, maybe not more than 2-3 cm from the Arduino.
Do you think my problems will be smaller if I put the coil in a separate steel enclosure some distance (20-30 cms) away from the enclosure for the Arduino?
I will of course always need to have the signal cable from the thermocouple to the Arduino. But with the relay that decouples the TC wires while the coil is working it may be enough?
As an aside, here in Australia when the ABC, the National broadcaster was being setup, the Melbourne AM station, 3LO, was given more transmitting power than the commercial stations.
One of the "stories" is that they were given the extra power so the wirelesses that were in the CBD, and usually played into the street, could be heard over the interference from the electric street trams.
Today, 3LO is 50kW and the most powerful commercial station is 10kW, and Melbourne still has electric trams.
No idea man, you will have to try about the grounding. You can start adding small caps around the thermocouple board. If you can get an oscilloscope you will be able to see the effect easily.
I meant that the noise from the igniter was coupling through the air to the CPU: there was no direct connection between the two.
