frequency counting for rpm

[missouridave]

I want to measure rpm by counting the freqency of a 12 pole alternator. The range I need is 0-1000 rpm. Portabilty would be a plus. Any help would be appreciated.
Thanks,
Missouridave

[Chris Smith]

A 4017 can count, divide by ####7, but you need to state more before anyone can help you??

[abelk]

Are you trying to measure RPM based on the ripple in the output voltage? Are you wanting alternator RPM or are you trying to measure engine RPM (or whatever is driving the alternator)? How about something to mark the pulley and using an optical tach setup?

Anyway, give specifics of what you are trying to do. This will make it easier to come up with a solution.

- Kent

[missouridave]

I am trying to measure the engine speed of a railroad locomotive. The engine is directly coupled (shaft) to the alternator. There are a lot of these engines and the shaft is not easy to get to, so I don't really want to add any kind of pickup. That's why I thought measuring the frequency would be easier. This is a high voltage alternator that is rectified to DC. At 900 rpm it generates about 1000 volts. Idealy the tach would be hand held so I could carry it on the engine.
Dave

[missouridave]

Let me correct some of my numbers:
This is a ten pole alternator, not twelve and the frquency at 900 rpm is 120 cps.

[dyarker]

?????????
900RPM / 60 sec per min = 15RPS
wouldn't 15RPS * 10 poles be 150Hz ??????

[philba]

yes, 15 rps. 150 hz.

so the issue is going to be measuring the ripple. I think that you could capacitively couple, rectify (half wave), amplify if needed, then use a schmitt trigger to turn it into a pulse train. from there, you can measure the time between pulses to get 1/10th (?) of the rotational period. From that, you can calculate the frequency. If rotational frequency doesn't change very fast, you could simply count the number of pulses in a fixed period to get the frequency.

I'd be concerned that the alternator will be putting out a lot of noise. not sure how to handle that. maybe low pass filter after the rectifier.

[jwax]

Given that there may be noise, or other causes for inaccurate readings, is there any way to verify the accuracy of whatever you build? Is there a mechanical tachometer?