How would I measure RPMs in a diesel engine?

[gregoryfenton]

Hi all

My current project is a good one (details will come later but it's something I am proud to be involved in) has several devices being monitored via a microcontroller - udoo http://www.udoo.org in this case, I opted for the quad core.

One of the tasks is to measure the RPM of a diesel engine. What would you guys suggest as to the best way to accomplish this?

My opinion is to go for a magnet mounted on the flywheel and a reed switch to count pulses, would this work?

Comments and suggestions are more than welcome

[the nerdling]

instead of the reed switch use a hall effect sensor, the reed switch would wear out and not be as accurate because of the movement

[Dave]

I wouldn't put anything on the axle itself. Instead, I would put a magnet on top of a hall effect sensor and fix that above the teeth of a spinning gear. The teeth moving in front of the magnet+sensor assembly would cause the magnetic field to fluctuate and you would see that on the output of the hall sensor.

[Toble_Miner]

Hi,

My opinion is to go for a magnet mounted on the flywheel and a reed switch to count pulses, would this work?

This would probably work but it would put some imbalance to your flywheel. You should at least use two magnets, one on each side of the flywheel. Also if you have a really high rotation speed it might be very difficult to balance the flywheel with the magnets.
In my opinion using some kind of a disk mounted onto the shaft of the engine with one or more slits in it and a light barrier shining through those slits once each turn of the engine is a much better idea.
Tobias
P.S. Please excuse my maybe quite rusty English. I'm from Germany and didn't write any Engish text for quite a while now.

[Mr Smiley]

Info on reed switches

http://www.biltek.tubitak.gov.tr/gelisim/elektronik/dosyalar/31/reed.pdf

How there made 



You may, at high speed have contact bounce problems.

Probably better off with a magnet and hall sensor or a slotted opto, if your going to connect to something that does say up to 5,000rpm then the lighter the item connected to the rotating shaft the better.

 

[Dave]

In my opinion using some kind of a disk mounted onto the shaft of the engine with one or more slits in it and a light barrier shining through those slits once each turn of the engine is a much better idea.

Unreliable. The engine bay is not a clean environment, the light sensor will get dirty and you will lose your signal.

[victor]

Optic Sensor

You can use a reflective adhesive tape and optic sensor so it does not add weight to the shaft.
You can do it using dark or light strips, you can do the threshold in software if your adc is fast enough or you can use discrete logic to do that and feed a digital pulse/clock to your micro controller.

Watch at 2:00


This will work to KHz range, but be aware of if it get dirt it may stop working, so you may need to enclosure it.
If you use infrared, be aware that hot metals may emit infrared it can affect your low threshold.

Hall-effect sensor

Or use hall effect sensor, computer fan use it as a tachometer, some fans goes as fast as 5k RPM and AFAIK this will cover the range of most diesel engines.
Steel is magnetic so you can predict some challenges you may encounter.

Brush

You can use kind of brush like used on DC motors, but it will have limited life, but it is easy to set up and cheap.

[NDT]

Hello Gregory!
   I'd rather go with a hall effect sensor counting the teeth on the flywheel or a magnetic pickup (coil magnet, essentialy the same). At least that's how everybody is doing it today (that doesn't mean it's the right way to do it, but you have a huge line of sensors to choose from). I've seen optical measurement on some Nissan engines too, but it's not really used. I've been in the EFI (electronic fuel injection) world for quite some time, and I've never seen a successful development using a method other than what I've just written. Besides, you get expandability, because you could measure TDC (top dead center) position if you make a notch or a recess on one of the teeth in such a way that the sensor doesn't react to that particular tooth. If you don't want such a huge resolution, you could attach the sensor to the injector pump, and with the crankshaft/pump pulley or gear ratio, it's easy to know the RPM.
   In a diesel engine, I see that the vibrations will be a major enemy of the reed switch setup you proposed. I don't know if it will affect the measurement, but no moving parts is always better!
   Some words about the two types of sensor I just described...
   Hall Effect: doesn't require heavy signal conditioning, it delivers a constant voltage pulse regardless of the RPM, it's very easy to use, but it relies on the Hall Effect, so you need magnetized teeth or a "trigger wheel", which is a disc with a number of teeth on the edge. It also needs a power source, usually +12V or +5V, but produces also a square +12V or +5V pulse.
   Magnetic Pickup: It's technically called a VR (variable reluctance sensor). It doesn't need magnetized teeth or an external power source, but the signal is AC (alternating current), and the output voltage also varies with the engine RPM, up to 50 volts or more! So you will need a much more complex circuit than the Hall Effect sensor. The VR sensor is cheaper but lager than the Hall Effect sensor.

[Toble_Miner]

In my opinion using some kind of a disk mounted onto the shaft of the engine with one or more slits in it and a light barrier shining through those slits once each turn of the engine is a much better idea.

Unreliable. The engine bay is not a clean environment, the light sensor will get dirty and you will lose your signal.

It this is the only problem you could just encapsulate the section of the shaft the whole optical measurement stuff is on in a small plastic case. Also I think it's quite hard to block the entire light passing through a light barrier. The emitter- and receiver-area are both quite big in most light barriers.

[hechen]

What engine are you trying to install it on?  Most modern engines have crank sensor with tone wheel and cam position sensor as well.  You could calculator the rpm with either of these signals.  I would not recommend installing your own tone wheel because it could "unbalance" the crankshaft.

[deth502]

most new engines have the fuel "distributor cap" electronically controlled, as opposed to the older engines that were mechanically controlled. (forgive my lack of knowledge on the correct terms. most of my work is done on gasoline engines, i rarely have to work on our diesel equipment). the diesel engine is run off of injections of fuel to each cylinder, much in the same way a gas engine is run by applying spark to the fuel in each cyl. this is analogous to your old distributer run off of your camshaft, while the newer ones are computer controlled through coil packs.

but i digress. the newer electronically controlled engines derive the tach signal from the computer. i would look into possibly tapping into that some how.

on an older mechanical engine, if it were me, id stray away from the flywheel and look to the balancer up front to get a signal from.

[Icchan]

Somehow I personally would use optic measurement... since there's need for quite fast speeds and high reliability. Shine an infra red LED trough the teeth of a gear and use photo detector (diode) with IR filter to measure the pulses. feed the signal trough an amplifier to micro-controller or some ready made frequency counter and you're done. Even if the duty cycle isn't 50% it should be able to calculate the edges. I would think you could get more than enough performance from this

[NDT]

most new engines have the fuel "distributor cap" electronically controlled, as opposed to the older engines that were mechanically controlled. (forgive my lack of knowledge on the correct terms. most of my work is done on gasoline engines, i rarely have to work on our diesel equipment). the diesel engine is run off of injections of fuel to each cylinder, much in the same way a gas engine is run by applying spark to the fuel in each cyl. this is analogous to your old distributer run off of your camshaft, while the newer ones are computer controlled through coil packs.

but i digress. the newer electronically controlled engines derive the tach signal from the computer. i would look into possibly tapping into that some how.

on an older mechanical engine, if it were me, id stray away from the flywheel and look to the balancer up front to get a signal from.

Diesel Engines have "heater" plugs, they are not for igniting the air/fuel mixture, just to get the engine started easier when it's cold. The air/fuel mixture is ignited because of the high temperature generated by the compression of the mixture inside the cylinder. So, diesel engines don't have a distributor, or spark plugs synchronized with the engine RPM neither in mechanical or electric or electronic form. They don't have coil packs either.

Of course if it's a diesel engine with an ECU of some sort, it's really easy to tap into the RPM output line and get the RPMs from there.

[Maximus64]

How about read signal from the fuel injector? It can be used to calculate the RPM I think.

[drake]

I agree with everyone else, don't add any weight to the crank shaft/flywheel. I work at a shop where we balance flywheels for diesel trucks and its precision work. If the engine is older and doesn't have electronic sensors to tap into on either the crank or cam, maybe you can pick up a signal from the harmonic balancer.

[NDT]

What engine are you trying to install it on?  Most modern engines have crank sensor with tone wheel and cam position sensor as well.  You could calculator the rpm with either of these signals.  I would not recommend installing your own tone wheel because it could "unbalance" the crankshaft.

If the trigger wheel is already balanced (as all commercial trigger wheels are) there's no problem fitting one.

[deth502]

most new engines have the fuel "distributor cap" electronically controlled, as opposed to the older engines that were mechanically controlled. (forgive my lack of knowledge on the correct terms. most of my work is done on gasoline engines, i rarely have to work on our diesel equipment). the diesel engine is run off of injections of fuel to each cylinder, much in the same way a gas engine is run by applying spark to the fuel in each cyl. this is analogous to your old distributer run off of your camshaft, while the newer ones are computer controlled through coil packs.

but i digress. the newer electronically controlled engines derive the tach signal from the computer. i would look into possibly tapping into that some how.

on an older mechanical engine, if it were me, id stray away from the flywheel and look to the balancer up front to get a signal from.

Diesel Engines have "heater" plugs, they are not for igniting the air/fuel mixture, just to get the engine started easier when it's cold. The air/fuel mixture is ignited because of the high temperature generated by the compression of the mixture inside the cylinder. So, diesel engines don't have a distributor, or spark plugs synchronized with the engine RPM neither in mechanical or electric or electronic form. They don't have coil packs either.

Of course if it's a diesel engine with an ECU of some sort, it's really easy to tap into the RPM output line and get the RPMs from there.

no, like i said, they have a fuel injector pump that work in that fashion. i made the distributor cap reference to describe how the fuel is metered in a diesel engine. im sorry if that was unclear.

[hechen]

How about read signal from the fuel injector? It can be used to calculate the RPM I think.

This method is not 100% accurate; when you let off the accelerator there will be a pause on the fuel injector.

[NDT]

One thing we all agree on: if we don't know what type of diesel engine we are talking, any solution could be adopted. If it has a mechanical fuel pump, or an electric one but with mechanically controlled injectors, the only way to go is a sensor in the crankshaft. Unless of course it has an ECU outputting an RPM signal.
Please Gregory, would you be kind and clarify this point?

[hechen]

What engine are you trying to install it on?  Most modern engines have crank sensor with tone wheel and cam position sensor as well.  You could calculator the rpm with either of these signals.  I would not recommend installing your own tone wheel because it could "unbalance" the crankshaft.

If the trigger wheel is already balanced (as all commercial trigger wheels are) there's no problem fitting one.

Its the method of mounting the trigger wheel/ tone wheel that will cause unbalancing.

[IanB]

So, diesel engines don't have a distributor, or spark plugs synchronized with the engine RPM neither in mechanical or electric or electronic form. They don't have coil packs either.

But diesel engines do have fuel injection, and they therefore have fuel injectors to inject fuel in a precisely controlled way at each cylinder. The fuel injection is absolutely synchronized with the engine RPM (how else would the engine run?) and in the case of electronic direct injection the injectors may look somewhat similar to a coil pack. I read nothing inaccurate in what deth502 wrote.

[dacrawf]

Modern electronically controlled diesel engines can have multiple pilot, a main, and multiple post injections per stroke for each cylinder. So measuring injection pulses is not a reliable way to go.  The flywheel should have notches around its circumference that the OEM crank shaft sensor is using, check to see if the flywheel housing has places to mount additional VR or hall effect sensors. As pointed out, another option would be to measure the speed on the front side of the engine via the balancer or some other component that is part of the belt drive. If the engine electronics are equipped with a open communications bus such as SAE J1939, you could read the data from that... Best of luck...

[Paul Moir]

Left field idea; if looking for a generic way to fit to any (hopefully multicylinder) engine:  a piezo electric sensor like a knock sensor.   Required engine installation would be therefore a simple bolt.

If going for a specific engine the magnetic pickup/gear method is very good.  You get lots of pulses/rev off say the flywheel ring gear.  Just use some some off-the-shelf automotive pickup that's magnetic.  That said, tachometers work well enough off 2 signals/rev for example with 4 cylinder, 4 stroke spark ignition engines.

[NDT]

What engine are you trying to install it on?  Most modern engines have crank sensor with tone wheel and cam position sensor as well.  You could calculator the rpm with either of these signals.  I would not recommend installing your own tone wheel because it could "unbalance" the crankshaft.

If the trigger wheel is already balanced (as all commercial trigger wheels are) there's no problem fitting one.

Its the method of mounting the trigger wheel/ tone wheel that will cause unbalancing.

Again, if it's a commercial trigger wheel it mounts with the same hardware as the crankshaft pulley, so you are not adding/substracting off-balance weight

[dacrawf]

So, diesel engines don't have a distributor, or spark plugs synchronized with the engine RPM neither in mechanical or electric or electronic form. They don't have coil packs either.

But diesel engines do have fuel injection, and they therefore have fuel injectors to inject fuel in a precisely controlled way at each cylinder. The fuel injection is absolutely synchronized with the engine RPM (how else would the engine run?) and in the case of electronic direct injection the injectors may look somewhat similar to a coil pack. I read nothing inaccurate in what deth502 wrote.

The problem is, you don't know how many injections are being done each stroke. Some diesel engines can do 5 or more injections per cylinder per compression/combustion stroke. This will very based on ambient conditions, engine load, engine operating mode, and many other things...