Engines and Electronics

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Joined
Mar 24, 2011
Messages
1,136
Reaction score
409
Location
UK, Warwickshire
Having got my E T Westbury designed Seagull engine running, I am working on an Arduino based system for various functions including a tachometer and radiator fan control based on coolant temperature measurement. First, the current state of play, and then a couple of questions.

As expected, interference is an issue. Everything is all still in a development phase and, for example the Arduino has yet to be put in an aluminium box but it is wired fairly tidily with short runs and twisted pairs where possible.

The first noise problem I solved was with the (brushed) starter motor. Even with a schottky diode across the motor teminals, the tiddly little capacitors supplied with the motor were inadequate. After reading round, a 1000µF capacitor, nearly as big as the motor, has fixed that one.

The main problem is the ignition, obviously. I found In order to get a spark I needed a big dedicated battery pack just for the ignition. This allowed me to separate the ignition/engine ground from the rest of the system, and that was a big step forward.

The system now works, with occasional minor upsets. In the last couple of days I have tried the radiator fan and its control system. This is a 4-wire computer fan which uses a PWM speed control signal at 25kHz. I put together a very rough initial trial setup (with a potentiometer instead of the temperature sensor) and long wires trailing all over the place. To my surprise this works fine with the ignition on and the engine being turned by the starter.

The remaining problem is the display. I am trying to use a nice sharp, bright 240 x 320 TFT graphic display with an SPI interface. With the coil disconnected, it is fine. Trouble is, with the ignition functioning it works for as much as a second after the starter button is pressed then goes blank.

So, how do I go about trying to make it work? Or, am I wasting my time with this type of display? The wires between display and Arduino are no more that 2" long. The 5V & ground between the two are twisted together and I have tried various sizes of capacitor across the two. I have tried the signal wires looped once through little ferrite rings. I have tried wrapping everything except the actual screen in grounded foil.

More generally, how do I systematically go about EMI protection? My knowledge of electronics is rudimentary in the extreme. I don't have an oscilloscope and in even if I did, trying to look at that while the engine is running would be a long, long way off.
 
I have seen folks use the Arduino-based systems to control home scrapping furnaces, with controlled fuel heater, two modulated fuel flow valves for the burner nozzles, temperature feedback, etc., and it works very well.

Learning how to configure Ardruino-based systems would be fun, and I understand the concepts of PLC control systems well, but I have too many hobbies going. I have a foundry, a wood pattern shop, a machine shop, and a 3D printer for pattern making.

I don't think my wife would be happy if I started an electronics shop.

I have seen some cores installed on wires, and I assume they are iron-based?
They are clip-on things, and I see then on signal wires, such as connecting a camera to a computer.

.
 
Might help to know what sort of ignition you are using. Points and coil or CDI tripped by hall sensor or simple contacts?
 
Its a Minimag MIC1A transistor ignition, with a Minimag Novus coil, triggered by a hall sensor. The hall sensor also provides the tacho signal to the Arduino, via an opto-isolator to keep the ignition and Arduino grounds separate. It is a wasted spark arrangement: both coil HT leads are attached to plugs.

You can see the ignition setup, without any of the electronics, in the video here:

http://www.charleslamont.me.uk/Seagull/testing.html
 
Last edited:
I found In order to get a spark I needed a big dedicated battery pack just for the ignition.
Normally, ignition battery packs do not need to be enormous.

First move in isolating ignition noise would be to use completely separate battery supply on the monitoring gear.

Metal enclosure for the spark generating circuitry and shielded cabling on the HT lead with resistor type spark plugs would be the other.

Many of these are just "starting point" as all installs are different in one way or another.

You may want to look also at snubber modules or clip on ferrite beads on the both power supply lines.
 
Last edited:
Are all the ground points tied together at only one end. if you have multiple ground points you can have current running through wiring in the opposite direction of the intended signal this can generate package drops in SPI Bluejets is rite about the ignition systems. My next venture with the Arduino is an independent monitor form the computer. I'm the guy that GreenTwin was referring to about operating my scrap processing furnace with an Arduino. I'm also in the process of putting an electronic lead screw on my lathe.

Art b
 
I have seen folks use the Arduino-based systems to control home scrapping furnaces, with controlled fuel heater, two modulated fuel flow valves for the burner nozzles, temperature feedback, etc., and it works very well.

Learning how to configure Ardruino-based systems would be fun, and I understand the concepts of PLC control systems well, but I have too many hobbies going. I have a foundry, a wood pattern shop, a machine shop, and a 3D printer for pattern making.

I don't think my wife would be happy if I started an electronics shop.

I have seen some cores installed on wires, and I assume they are iron-based?
They are clip-on things, and I see then on signal wires, such as connecting a camera to a computer.

.

Grin - - - somehow - - - - I just start things - - - - the wife - - - - well - - - - her 'going nova' yarn/pattern/fabric stash means that I'm just not worried about things. As I'm making stuff that gets used around here - - - well - - - I'm not worried about adding more things to do.

(I don't have the foundry - - - yet!!! (beeg grin) but then I have a few other things going - - - LOL)
 
Are all the ground points tied together at only one end. if you have multiple ground points you can have current running through wiring in the opposite direction of the intended signal this can generate package drops in SPI
At the moment I have a 10" long, 3/16" square, brass grounding bar with holes every 1/4". Everything except the ignition is grounded to the bar by a fairly direct wire. Here is a detailed scale circuit layout, as it is at present, without the radiator fan.
(To see detail it needs to be zoomed to 150 or 200%)
 

Attachments

  • Electronic Layout 3.pdf
    123.1 KB
Back
Top