Magneto interference with arduino

Hi guys, I'm trying to use a freetronics eleven (UNO) with a LCD screen shield on an alternator based generator I have built. The board will be used to read volts, amps and RPM of the 2HP two stroke motor. It will also interface with the field coil on the alternator to control max load on motor as well as the voltage and current outputs. I believe all is well with the interface between the UNO and the alternator as I have reduced all the inputs to 4v via resistors to ground and the RPM input is taken from one of the three phases of the alternator winding before the bridge rectifier to avoid feedback from the motors magneto.
The problem I have is the magneto on the motor is causing the LCD screen to read out some kind of ancient Egyptian language and the board eventually shuts down. The unit works perfectly if you remove the magneto from the motor and spin it all by hand so I'm led to believe this is not a code issue. I have used double shielded cable all round and mounted the boards in an aluminum enclosure to try and reduce noise. It seems now the interference is finding its way to the unit via the ground wire and I have no idea how to stop it as I need to have the arduino grounded to the generator as it is its power source.
Any help will be much appreciated but, if possible, try to dumb it down for me as I'm not a guru with this stuff (yet).

You ned to add lots of capacitors to reduce the interference with seriese inductors. Look up decoupling techniques.

Thankyou, I shall, would you know if I would have to do it to the earth only or to all the inputs, the other inputs read the alternator so am I able to presume they are electrically isolated from the earth noise?

Justscary:
Thankyou, I shall, would you know if I would have to do it to the earth only or to all the inputs, the other inputs read the alternator so am I able to presume they are electrically isolated from the earth noise?

Sorry I think your English fell apart there so I am not sure what you said.
You need everything to share a common ground, read:-
http://www.thebox.myzen.co.uk/Tutorial/De-coupling.html
and
http://www.thebox.myzen.co.uk/Tutorial/Power_Supplies.html

Sorry, I was asking if I should decouple the power supply on the arduino as well as try to somehow decouple the magneto as well ( it has a low volt magneto and a seperate coil), or just the arduino supply would suffice? This decoupling stuff seems pretty in depth, am I going to hurt anything if I just use the trial and error method?

am I going to hurt anything if I just use the trial and error method?

No.

Basically you apply more and more decoupling until it woks correctly. The more spread around it is the better.

Thanks heaps for your help, I'm going to give it a go. I'll post here to let you know on progress, but it might take a while as I don't get much time to play with it.

Good luck, I know this is tricky. Also try and keep wiring neat and run away from sources of interference.

Thanks, I will keep that in mind, would it be prudent to replace the magneto wires with shielded cable as well (before the coil) and earth the shield on one end, or do you think it would not make much difference. I believe that most of the noise is from the ht lead after the coil. Do you have any ideas on shielding that lead as I can't replace the cable, it is a sealed unit on the coil. Once again Mike, I really appreciate the help, and I'm learning a lot in the process......

I believe that most of the noise ....

Electronics is a matter of what is true not what you believe. To track things down you need test equipment, like a scope.
Interference has two ways of getting into your circuit.

  1. By electromagnetic propagation here shielding can help. You could wrap the lead in copper foil the type they sell for stained glass work and ground one end only.
  2. By conduction through the wires either signal wires or most probably through the power supply wires. Here capacitors help but you can't use too much capacitance on the signal wires otherwise it affects the signal. See this for some things you can do with signal wires.
    Protection

Awesome, thanks again, it is certainly a lot more in depth than I imagined! I'll give it all a go in the next few days and let you know how I get on. This project might be the perfect excuse to buy that scope I've been putting off, he he he.

Neat wiring may not necessarily be the best solution at times. Depending on the freq of the interference, you may have to reroute the offending cable/s.
Also, though you say its entering via gnd, you need to examine the inputs to the arduino very carefully, because a signal needs a potential difference. Have you tried ferrite beads on the non-ground leads to the arduino/display box ?

No, I haven't tried that, my biggest problem is room as I'm trying to make it as small as possible. This means that everything is close together and I do know that when I diagnose on cars and things, the ht lead can interfere with my multimeter, and it is not a cheap one either. Given that knowledge, and the fact that the whole generator is around a 12 inch cube, I'm led to believe that it is the ht lead that is causing it. As mike said, it is a matter of what is true, and I don't have the correct test equipment yet, I have to go on my gut feeling. I am having a small bit of success with the decoupling capacitors but I haven't had a lot of time to play with it. I will keep you posted.....

Make sure your engine has a "resistor spark plug". They should be standard on any modern engines but it's worth a doublecheck.

http://www.ngksparkplugs.com/techinfo/spark_plugs/faq/faqresistor.asp

Chagrin, I saw your post and thought "don't be stupid, everything has one now" but I did check and I guess the Chinese don't use them unless they have to. Swapped it to a Bosch "R" type plug and everything is fine. Peeved me off because I normally do the simple things first, but if it weren't for you, I'd still be off trying other things. I have found that I still need decoupling on the power supply, which is a big help but now with both, no errors at all. A big thanks to both of you for your help!