Derive a tachometer signal from magneto

Thread Starter

geekoftheweek

Joined Oct 6, 2013
517
I have your standard twin cylinder lawnmower engine that I would like to build a tachometer for. I plan to use the wire you ground to shut the engine off as the input if possible. I don't know the values to expect is where the problem is as well as if it is even possible. If possible would something along the lines of megasquirt's circuit work? http://www.msextra.com/doc/general/ms1v3schems.html on the Tach Input drawing (third schematic from top) and referring to the opto isolator circuit.

I could also probably order up some hall effect sensors and try that instead, but I would kind of like to use up some things I've had laying around forever now.
 

drc_567

Joined Dec 29, 2008
881
According to one of two magneto reference articles, the flywheel on your lawnmower should have two or more large magnets on the side, so as to create dynamic magnetic flux in a wire coil and generate a high voltage to the spark plug.
One possible method of detecting the flywheel rotation would be to place a Hall effect sensor in such a manner that the flywheel magnet flux activates the Hall sensor producing a countable pulse.
A metal wall between the flywheel and the sensor could require opening a small aperture so that the magnet flux would not be attenuated by the metal casing.
... Describe what manner of readout is acceptable to you. A pulse counter would give an actual numeric result. However, it may be possible to fabricate a needle type analog meter.
... Just a basic idea at this point, will surely require some refinement.
...
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
517
@drc_567 I plan to connect it to a PIC... probably 16f (whatever I have laying around still). I actually am not planning a readout at this time, but when I get around to it.

@Dodgydave simple and effective. I remember seeing this before, but didn't think of it off the top of my head. I actually have a few handfuls of op amps in the box so this sounds like the way to go. I'll have to play around and see what happens.
 

drc_567

Joined Dec 29, 2008
881
... Final thought
A copper sense wire placed around the plug wire may eventually fail due to vibration fatigue.
A ferrous wire could last longer ... Something to watch for.
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
517
Thanks again. An actual number would be nice. Just for some background...

I planted a bunch of bushes too far from any water source. I'm using a 25 gallon tow behind sprayer as a water tank / drip irrigation system feeder. I want to be able to leave the pump run, monitor battery voltage, and when it drops enough start the engine, charge the battery, and shut down. I just need a way to figure out when the engine is turning fast enough to call it running and shut off the starter. It does have a pressurized oil system and I probably have a sensor around I could read with an ADC to get the pressure and do it that way, but I need to make sure the pressure rises fast enough to call it workable.

I can do anything programming wise so really all I need is a digital signal to time.
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
517
I was actually experimenting earlier and found with the test setup I could run the pump for 30 minutes with around a .05 volt change in battery voltage on a basic lawnmower battery. The pump was actually only running 20% of the time I would estimate. My guess calculations are around a 50 - 60% duty cycle on the pump for the final version and maybe a deep cycle battery that I charge when I'm done might be a better way to go.
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
517
That depends upon the type of magneto (points or solid-state) and what signal (if any) appears at that wire.
You'd need an oscilloscope to determine that.
Ahhh. Points type and no scope.

Also wondered a bit how a proximity switch would be affected by the magnets. Could do one of those fairly easily.
 

JMW

Joined Nov 21, 2011
127
I've worked on a bunch of these engines. Don't ever recall have points on a magneto. It just fires when ever the magnet passes the coil.
 

JMW

Joined Nov 21, 2011
127
Thanks again. An actual number would be nice. Just for some background...

I planted a bunch of bushes too far from any water source. I'm using a 25 gallon tow behind sprayer as a water tank / drip irrigation system feeder. I want to be able to leave the pump run, monitor battery voltage, and when it drops enough start the engine, charge the battery, and shut down. I just need a way to figure out when the engine is turning fast enough to call it running and shut off the starter. It does have a pressurized oil system and I probably have a sensor around I could read with an ADC to get the pressure and do it that way, but I need to make sure the pressure rises fast enough to call it workable.

I can do anything programming wise so really all I need is a digital signal to time.
I would buy suitable battery and use a solar panel, and an el cheapo charge controller. I know, I know you want to build something. I understand that, this is a classic case for solar.
 
Small engines with magneto ignition always have points and condenser, now replaced with solid-state SCR modules. The points are needed for the fast dI/dt change to make a spark. A slow moving magnet doesn't really do it.

Megasquirt 1's tach circuit I'm not a fan of, it has many issues. Most tach circuits on the Internet are crap because a coil primary dishes out a >400Vpk spike and you don't want to load that down at all and cause a weaker spark. It dumps 6mA ignition pulses into Vcc, uses a 1A transistor and has no debounce of the coil waveform. It's not a "true zero-crossing detector" either and those don't work with a coil anyway due to ringing causing false-triggers. I would not bother building it.

You can just use one or two transistors (monostable) to pick off the ignition pulses. You have to tie into the kill switch anyway, to turn the motor off, so adding a tach monitor there is easy. The trigger level is somewhat unique for each engine (magneto, DC power) and 2/4 stroke. V-twins misfire a lot so a tach must trigger with fire or misfire voltages.

For remote/auto start of an engine, there is also a crank-timer so you don't turn the starter too long, three attempts with timeout between tries, low battery detection etc. before going into lockout.
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
517
Megasquirt 1's tach circuit I'm not a fan of, it has many issues. Most tach circuits on the Internet are crap because a coil primary dishes out a >400Vpk spike and you don't want to load that down at all and cause a weaker spark. It dumps 6mA ignition pulses into Vcc, uses a 1A transistor and has no debounce of the coil waveform. It's not a "true zero-crossing detector" either and those don't work with a coil anyway due to ringing causing false-triggers. I would not bother building it.

For remote/auto start of an engine, there is also a crank-timer so you don't turn the starter too long, three attempts with timeout between tries, low battery detection etc. before going into lockout.
I do have one of the circuits built already I was trying out before. It worked fine for my truck, but when I connected it to an old points system on a tractor it wouldn't run right and didn't work... kind of what made me ask the question.

I thought about just doing a timer to crank the engine. Usually after the first start of the day it will fire right up pretty much just by bumping the key the rest of the day. Bump the starter, wait a second, check and see if the voltage is up to a running voltage, then either wait a couple seconds to retry, or let it run 10 minutes or so.

I would buy suitable battery and use a solar panel, and an el cheapo charge controller. I know, I know you want to build something. I understand that, this is a classic case for solar.
It would be fun to build something for sure! Part of me wants to do it just to say I did it.
Luckily I already have all but a solar panel. Solar would be smarter in the long run, but this is going to be a temporary fix to get me through the summer. If all goes according to plan next year will be burying water line for a faucet.
 

MisterBill2

Joined Jan 23, 2018
6,967
If you bare really ambitious you could wind a few turns of wire around the outside of the magnet coil and get your own tachometer pulses. AN LM2917 tachometer IC will drive your meter, but since you really only want to switch off the cranking motor you don'e need that.
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
517
Thanks @MisterBill2. The LM2917 and a comparator would be a very simple way to determine if the RPMs are high enough. Also the turns of wire around the coil sounds pretty good also. It wouldn't have the vibration issues of the plug wire and I could silicone it in place.
 

MisterBill2

Joined Jan 23, 2018
6,967
REally, if you could get enough turns around the magneto housing you might get enough power to operate a relay when the motor caught and ran up to normal idle speed. That could be a super-simple circuit that could be very reliable..
 

JMW

Joined Nov 21, 2011
127
Small engines with magneto ignition always have points and condenser, now replaced with solid-state SCR modules. The points are needed for the fast dI/dt change to make a spark. A slow moving magnet doesn't really do it.

Megasquirt 1's tach circuit I'm not a fan of, it has many issues. Most tach circuits on the Internet are crap because a coil primary dishes out a >400Vpk spike and you don't want to load that down at all and cause a weaker spark. It dumps 6mA ignition pulses into Vcc, uses a 1A transistor and has no debounce of the coil waveform. It's not a "true zero-crossing detector" either and those don't work with a coil anyway due to ringing causing false-triggers. I would not bother building it.

You can just use one or two transistors (monostable) to pick off the ignition pulses. You have to tie into the kill switch anyway, to turn the motor off, so adding a tach monitor there is easy. The trigger level is somewhat unique for each engine (magneto, DC power) and 2/4 stroke. V-twins misfire a lot so a tach must trigger with fire or misfire voltages.

For remote/auto start of an engine, there is also a crank-timer so you don't turn the starter too long, three attempts with timeout between tries, low battery detection etc. before going into lockout.
You must speaking about multi-cylinder engines. I have two Briggs and Stratton 6.5 HP engines. One is from 04 the other 08.. Trust me there are no condensors, points etc. Magneto coil connects directly to the spark plug. Plus several others from the great asian supplier. I seriously doubt an engine used as a battery charger is multi-cylinder.
 

Thread Starter

geekoftheweek

Joined Oct 6, 2013
517
You must speaking about multi-cylinder engines. I have two Briggs and Stratton 6.5 HP engines. One is from 04 the other 08.. Trust me there are no condensors, points etc. Magneto coil connects directly to the spark plug. Plus several others from the great asian supplier. I seriously doubt an engine used as a battery charger is multi-cylinder.
I have a recent Predator engine and after looking through the parts diagram it doesn't show points or anything connected to the magneto either. Maybe newer technology or greedy people trying to shave a couple bucks.

B&S used to have points under the flywheel back in the day. I tore apart plenty of perfectly good 3.5 horse push mower engines for no good reason as a younger one and "put them back together".

The engine I'm using is actually an 18 horse opposed twin Kohler. A little much for a battery charger, but it also is what came with the mower.
 
Top