next topic »

[on:]

Hi,

   So, i'm building a railway locomotive from scratch ( www.lms10000.org )  - Yes, I know it silly...

   We need to monitor our 247 Litre, 16cylinder diesel engine, and as it was built in the 1940s, it has no electronic systems at all on it. So i've built a system out of an arduino mega that can monitor things like  oil pressure, water temperature, air temp etc. etc.  For all these, there is a sufficiently robust and easily available sensor.

   What i'm having trouble with is finding the following sensors;
              Something to detect RPM (hall effect or similar)
              Measure exhaust temperature (upto 1200deg F)
              Measure air inlet pressure (it's a turbo, so 2.5bar max abs press)
              Measure individual cylinder fuel pressure (400 Bar+)

   I can find some sensors that cover these ranges, but they are usually not suited for the loud, noisy, oily constantly vibrating environment inside a locomotive. For example, while the "reading RPM" demo  (http://playground.arduino.cc/Main/ReadingRPM ) would work fine, the sensor would last about 2 minutes.


   I suspect that there are automotive sensors that would be useful, (especially for the RPM and inlet pressure) but I haven't been able to figure out which ones.  Does anyone here have experience of which parts have worked well in these applications, and can be easily interfaced to an arduino?


Mark B

[Reply #1 on:]

You can't use the hobby sensors we use mostly with the Arduino.
There are many industry strength sensors, they are for example used in power plants and oil refineries.

[Reply #2 on:]

For the RPM sensor, you could use a US1881 Hall sensor and 2 magnets, but you might have to encapsulate the sensor leads and the connections to them in epoxy if the vibration is severe.

For EGT measurement, thermocouple probes are standard. These are used in light aircraft engines where there is also a lot of vibration, and seem to last quite a long time. See http://learn.adafruit.com/thermocouple/ for information on interfacing thermocouples to microcontrollers.

[Reply #3 on:]

Thanks DC42;  I've found a different hall effect sensor in an automotive form factor, butI'm not entirely sure how i'll go about getting it to trigger my RPM counting interrupt as it doesn't output +5v logic levels.

Thanks to the pointer to thermocouples - I knew about these, but I hadn't spotted the amplifier breakout boards, so that looks like a workable solution (mounted in a small, sealed box in the "V" of the engine).


MB

[Reply #4 on:]

I've used variable reluctance pickups (like these: http://www.electronumerics.com/CRA/magnetic_sensors.htm) on actual diesel locomotive engines to measure RPM. They are easily conditioned into a square wave with an analog comparator circuit or you can get them with builtin electronics. They are very rugged.

You can measure the time of one revolution of the flywheel and calculate RPM from that or if you need a faster update, count gear teeth from some gear that is driven by the crankshaft.

Some of these pickups can survive for years immersed in hot oil but might be more expensive than you can tolerate.

[Reply #5 on:]

Thanks PapaG,

   They look like just the right sort of sensor, and rugged enough that I could mount it anywhere on or in the engine.  I'll look into the internally conditioned ones.  Many thanks.

MB

[Reply #6 on:]

Thanks PapaG,

   They look like just the right sort of sensor, and rugged enough that I could mount it anywhere on or in the engine.  I'll look into the internally conditioned ones.  Many thanks.

MB

You're welcome.

[Reply #7 on:]

check - http://www.temperatures.com/ - for a lot of info about temp sensors