Creating RF to test circuits

[Tobias]

I build electronics for motorsports. As you can imagine electrical noise is a problem. The ignition systems are bad ass.

I have an ignition system on my test bench. I have a little electrical motor that spins the crank trigger. The crank trigger has four magnets that are used to trigger an inductive sensor. The sensor then sends the signal to the ignition system to fire. The ignition system is connected to a coil and test spark plug. It isn't a spark plug that you would literally use in a car though, it doesn't have a ground strap so the spark jumps around 0.250" to the exterior part of the test plug. I also use a race car battery to power the system.

So I can use this to test my circuitry for RF problems. At least I thought so. Some circuitry works great on the test bench yet on the race car it has problems. Talking to a friend of mine he said that there are two types of RF noise generated, one from high voltage and the other from high current. He suggested making a pressure chamber to fire the plug to really simulate a car. He also suggested instead of using a spark plug in open space use a bolt that is adjustable in relation to the ground strap. Then I can change spark gap.

Are these two ideas good? Are there any other ideas that I can implement on my test bench?

[Melanie]

Screening your device (ie enclosing it in a GROUNDED sealed box) will eliminate a lot of ambient noise floating about... but that is only half the battle...

The other half comes down the wiring into your device. Whatever you are building usually doesn't sit sealed-up in it's own box. It's connected to the outside world. It's THAT external wiring that introduces noise into your device.

That wiring could be your SUPPLY from the vehicle's Battery, inputs from Sensors and Outputs to Motors, Solenoids, relays, Aerials etc.

The SUPPLY is usually the biggest culprit. But if you run Sensors (eg a temperature Sensor) and their cables pass close to areas of high noise generation (ignition circuit or even near high-current carrying supply lines), then you will get noise induced into the wiring. There are ways of minimising such noise. Screen Cables is an obvious one, but also designing LOW IMPEDANCE circuits (that will pick up less noise), and adding filters on all I/O's also may be required. In many instances you might have to do more than one thing in order to make your device noise immune.

For real critical applications I've got an ICOM-7000 Transceiver which I load into an antenna right next to my device and it's cabling/sensors. If I can key say 5W into the Antennas at a spread of frequencies in the HF, VHF and UHF bands without my device dying, I know I'm pretty sure it's bomb-proof. It is an extreme test! You can pick up an ICOM-706-IIIG quite cheaply and it is only capable of 10W anyway, so it's just perfect for generating RF noise (which is the equivallent of what most of the young Radio Hams do anyway) for testing.

[Tobias]

Thanks for the input. The boards are enclosed in an extruded aluminum housing. I don't have ground connected to the housing though, I will get that done.

One thing I did on the new board is use a DC/DC converter and optically isolated all the inputs.

I will look around on the internet for a ICOM-706-IIIG

Thanks again