Tach Sender on a BMC 1.8

[Jen-in-Wellies]

This is on a friends boat, not my own. The engine and instruments are of around 1979 vintage, a BMC 1.8 engine and Smiths instruments. We are developing a digital instrument set for the boat, so I am looking at interfacing this with the various senders on the engine. The sender for the tachometer has two terminals and goes in to the side of the engine block underneath the exhaust manifold near the rear of the engine. The associated tachometer has no power supply, or ground, except that to the illumination bulb, so all power to move the needle comes from the output of the sender. The two wires from the sender go direct to the tacho. A picture of the sender is below.

 

Does anyone recognise this sender type? How does it work? What sort of output does it produce? AC proportional to speed, pulses? What sort of voltage? What sort of frequency compared with the engine speed? It can probably be worked out eventually from measuring the output, but I don't have an oscilloscope easily to hand and any hints and short cuts would be a big help.

 

The alternator on the engine doesn't have a tacho output we could use as an alternative. Fitting an inductive sensor to work off a crank pulley would require a lot of work making brackets etc that I would like to avoid.

 

Thanks,  Jen

 

 

[Tony Brooks]

It looks very much like those fitted to Jags in the 60s'70s. If it helps I think it is a simple spinning magnet generator and I would expect an AC output with the frequency dependant upon rotor speed. May not be correct but that was the impression I got from playing with them.

 

Those jags had Smiths instruments.

 

If I am right then a diode in series to give half wave rectification and it might drive an ordinary "alternator" sensed Tacho providing the adjustment allows.

[Jen-in-Wellies]

Thanks Tony,

I'll do some experiments based around that set of assumptions. Putting a multimeter on it set to AC Volts should hopefully show something when the engine runs. If correct, a diode will give a set of half wave pulses as you describe, then a bit of electronics to give a 5V square wave triggered off say a rising edge, which should feed in to the PLC nicely for doing the sums, then output to the liquid crystal display. Easy said!

 

Jen

[Chewbacka]

If getting a new tacho why not get a diesel tacho and let it pick up it’s signal from the alternator or is the alternator too old?

[Jen-in-Wellies]

1 minute ago, Chewbacka said:

If getting a new tacho why not get a diesel tacho and let it pick up it’s signal from the alternator or is the alternator too old?

Because we are trying to make a set of digital display instruments while spending the absolute minimum amount of money! ?

The alternator is an old style Lucas one, with just an output to the warning light bulb. Hence we are trying to use the existing tacho sender. Will also be using the existing temperature and oil pressure senders if at all possible as we are tight wads. All of this is going to feed in to an Arduino PLC and the results will go to a liquid crystal display and various warning LED's and buzzer.

 

Jen

[Jen-in-Wellies]

A quick experiment with the multimeter set to AC volts seems to back up what Tony suggested. At idle it shows around 1.8V, rising to around 8V as the revs increase. I am guessing frequency is also proportional to engine speed, but can't measure that.

 

Jen

[Chewbacka]

1 minute ago, Jen-in-Wellies said:

A quick experiment with the multimeter set to AC volts seems to back up what Tony suggested. At idle it shows around 1.8V, rising to around 8V as the revs increase. I am guessing frequency is also proportional to engine speed, but can't measure that.

 

Jen

Some multimeters have a frequency measurement range, I guess yours doesn’t.  Just a thought, if you have an old crystal earpiece or  a piezo speaker - both high impedance, if you connect it across the transducer output you should be able to hear the frequency change and you can get free phone apps that measure audio frequencies, so acoustically  couple to your phone, give you an idea of the frequency range you are working with.

[Tony Brooks]

If it has that sender then it just might have bimetallic instruments and senders. If so both will produce some kind of square waveform output. If the temperature and oil pressure gauges rise slowly at turn on this is likely to be the case. If not the temperature sender should be easy enough being a thermistor. The oil pressure sender I am not so sure about because I understand modern ones use a piezo (spelling?) sender and I am not sure how they get a continuous varying output.

 

Sounds an interesting project.

[Jen-in-Wellies]

8 minutes ago, Tony Brooks said:

If it has that sender then it just might have bimetallic instruments and senders. If so both will produce some kind of square waveform output. If the temperature and oil pressure gauges rise slowly at turn on this is likely to be the case. If not the temperature sender should be easy enough being a thermistor. The oil pressure sender I am not so sure about because I understand modern ones use a piezo (spelling?) sender and I am not sure how they get a continuous varying output.

 

Sounds an interesting project.

The temp and oil pressure gauges are just old style bimetallic strip ones, so will likely be off resistance senders. The temperature sender will almost certainly be a thermistor, which are very non linear with resistance against temperature, however the relationship can be coded in to the PLC.

 

If the project progresses as we hope I'll start a separate topic for it.

 

11 minutes ago, Chewbacka said:

Some multimeters have a frequency measurement range, I guess yours doesn’t.  Just a thought, if you have an old crystal earpiece or  a piezo speaker - both high impedance, if you connect it across the transducer output you should be able to hear the frequency change and you can get free phone apps that measure audio frequencies, so acoustically  couple to your phone, give you an idea of the frequency range you are working with.

Good suggestion with the earpiece. Another way for audio range frequencies is to use the line in to a laptop, with some protection circuitry and a suitable software app to make a cheapo oscilloscope. I did make one years ago, but it has long gone along with the PC.

As voltage from the tacho is also proportional to engine speed, what I could do is rectify it and just use analogue voltage measurement in to the Arduino for engine speed. This is probably the simplest solution, rather than try to measure frequency. The existing tacho probably works the same way, with the moving coil just acting as an AC voltmeter.

 

Jen

[Chewbacka]

7 minutes ago, Jen-in-Wellies said:

The temp and oil pressure gauges are just old style bimetallic strip ones, so will likely be off resistance senders. The temperature sender will almost certainly be a thermistor, which are very non linear with resistance against temperature, however the relationship can be coded in to the PLC.

 

If the project progresses as we hope I'll start a separate topic for it.

 

Good suggestion with the earpiece. Another way for audio range frequencies is to use the line in to a laptop, with some protection circuitry and a suitable software app to make a cheapo oscilloscope. I did make one years ago, but it has long gone along with the PC.

As voltage from the tacho is also proportional to engine speed, what I could do is rectify it and just use analogue voltage measurement in to the Arduino for engine speed. This is probably the simplest solution, rather than try to measure frequency. The existing tacho probably works the same way, with the moving coil just acting as an AC voltmeter.

 

Jen

Output frequency to engine speed is probably a linear relationship, but voltage may not be, which might entail a bit of mapping, depending on the accuracy you want, but probably simpler and cheaper, otherwise you are looking at a frequency to voltage ic.  

Added or an AtoD chip to feed the Arduino

Edited June 8, 2018 by Chewbacka

[Tony Brooks]

8 minutes ago, Jen-in-Wellies said:

The temp and oil pressure gauges are just old style bimetallic strip ones, so will likely be off resistance senders.

If the oil pressure sender is a metal can about 1/2" thick and about 3" diameter then its a bimetallic sender and it will produce  a square wave output.

From memory It has a bimetallic strip in it with a heater coil around it. The heater coil earths via contacts on the end of the bimetallic strip. A diaphragm in the sender pre-loads the strip so the contacts are pushed together more tightly.

 

When the gauge sends current to the sender the bimetallic strip gets heated until it bends up and breaks the contact. the strip then cools and remakes the contact. The time the contacts are made depends upon the preload on the strip so the mark space ratio depends upon the preload on the strip. The strip in the gauge takes time to cool and heat up so takes up a position that depends upon the length of pulses and the position is transmitted to the needle. If I am correct there is usually a voltage stabiliser module involved that operates the instruments at around 5 volts AVERAGE  but again most were PWM modulating bimetallic units. It might be cheaper to use a modern sender than develop a 12v to 5V converter (later Fords used a solid state one I think but have a feeling they used a different voltage) and mess about with averaging the PWM signal.

[Jen-in-Wellies]

Thanks again Tony,

More very useful information.

 

Jen

[Tony Brooks]

I should have said the PWM signals were at an exceptionally low frequency when compared with modern stuff.  You could get them to flash small lamp bulbs.