Remote start petrol water pump

[wigman27]

Hi,

I have been asked to put together a kit for remote starting a petrol water pump. I'm going to use a micro at one end controlling two Xbees for wireless comms. It is about 500m away.

I can figure out most things but I am having a little trouble working out how to detect when the motor has started and when to disengage the start solenoid. Obviously if you're starting it at the pump then you just let go of the key when it fires.

One thought I had was to use a comparator and have one side connected to the battery voltage and the other side connected to the charge light circuit and somehow detect the voltage when the alternator is charging and therefor the motor is running.

Do you think that will work? Does anyone know of a better way to do this?

Thanks

Lee

[dr.diesel]

Do you think that will work? Does anyone know of a better way to do this?

I would use redundant hall effect RPM sensors.  Compare this against a range of known starting RPMs, then once some level above that consider the engine running.

[richard.cs]

Starting a petrol engine like this is actually quite a difficult problem. Detecting that's it's started could be done by rpm (hall sensor or alternator) or acoustically - a piezo mechanically coupled to the exhaust easily picks up the pressure pulses if the engine fires. Deciding what to try if it doesn't start first time is harder - a human uses all their senses for this (did it sound like it fired? is there black smoke? blue smoke? does it smell of petrol? etc).

Does the starter linkage mechanically disengage once the engine fires? If it does and you don't detect it then a conventional series-wound starter can rev to destruction. You'll may need to consider if it's worth having control of the choke as well, it depends on your climate really.

[AlfBaz]

I'd be inclined to do some simple testing and see how viable it would be to simply get a piece of wire and wrap it around the spark plug lead and "sense" the pulses.

You can start but removing the sparkplug from the engine and touching the base of it onto the chassis. This will allow the starter motor to crank the engine, fire the spark plug and not actually start the engine. I would also place a dummy plug in the engine whilst doing this so that you don't get the faster RPM due to lack of compression. This way you can determine the average rpm created by the starter motor.

Once that's done get an rpm profile of the engine actually starting and the point at which you would manually let go of the start button.

[wigman27]

The Hall effect sensor is a really good idea!! That will work well.

The only thing is I wanted to get away from using a mcu at the pump end if I can, do you know of a way to latch and output based in an RPM without using an MCU? I could send the counts through the Xbee but it's a bit slow for that kind of thing.

Also as far as it not starting correctly, I will probably set it for a max starting time.. Say 5 to 7 seconds, if it hadn't started by then I will show a pump error and they will have to walk down and check it out.

I'm not 100% certain if the start disengages the actual pump hasn't been bought yet. I just wanted to see if it was possible.

But latching something at the pump at a certain rpm would be good if it's possible.

Thanks again

Lee

[ConKbot]

since it has an alternator you could detect RPM by alternator whine in the DC power,  measure closer to the alternator rather than closer to the battery, obviously. A frequency to voltage converter (would work for both the hall effect sensor output, and alternator whine) and a low pass filter/integrator  and comparitor to geta "yeah looks like the engine is on" signal. 

[G7PSK]

I would use the oil pressure switch virtually all petrol engines have one these days or a tapping for one. 

[richard.cs]

I would use the oil pressure switch virtually all petrol engines have one these days or a tapping for one. 

I had assumed a small engine of a few hp, those are usually splash lubricated and have a switch on oil level rather than on pressure. On a bigger engine with pumped lubrication an oil pressure switch might work but on some engines you can get oil pressure just by cranking, on some others (such as my car) is appears only about 5 seconds after starting.

[BillyD]

Another option that springs to mind would be to use a current transformer to monitor the load on the starter motor on the basis that this would reduce once the engine fires up.

I have no idea how feasible this is though, it's just prompted by the fact that I've been playing around with current transformers in a different application recently.

[SeanB]

Just buy a pump with remote start capability and connect the single relay contact needed?

Simple methods used are a timed start period, a delay and a second timed start period bypassed if the engine RPM is above a minimum speed detected from pickup of spark plug magneto pulses or CDI RPM output. If second period gives no start then the startup has failed and it times out to give low RPM alarm.