Electronic Tachometer Circuit

[C. Smith]

I wish to monitor the rotational speed of a 2" shaft in the 10 to 400 RPM range. I could attach a magnet to the shaft and position a pickup coil to generate pulses. What I am looking for is a simple, very basic circuit that will process these pulses and drive a 3 digit LED display showing RPM. The circuit should be able to be built on a Radio Shack project board out of easily obtainable parts.

 

[Tim Wescott]

I wish to monitor the rotational speed of a 2" shaft in the 10 to 400
RPM range. I could attach a magnet to the shaft and position a pickup
coil to generate pulses. What I am looking for is a simple, very
basic circuit that will process these pulses and drive a 3 digit LED
display showing RPM. The circuit should be able to be built on a
Radio Shack project board out of easily obtainable parts.

at 400 RPM (7Hz) you should be able to do this with a Basic STAMP, as
long as the PIC thereon has a timer.

You should also consider a pickup that's better for low-speed use than a
coil-magnet, and one that gives you more than one pulse per revolution.
I'd consider a hall-effect sensor or an optical encoder. If the
design gods smile upon you it may even be practical to print up a paper
tag and use a reflective optical pickup.

 

[Joerg]

Tim is right, for very slow speeds a magnet may be a problem. But mind the application here. If things can get dirty then the magnet is the better solution. Especially if you really don't need to measure below 10 rpm. Just make sure you clamp the sense circuitry appropriately so the signal doesn't fry any devices.

Regards, Joerg.

 

[Spehro Pefhany]

Tim is right, for very slow speeds a magnet may be a problem. But mind the application here.
If things can get dirty then the magnet is the better solution. Especially if you really
don't need to measure below 10 rpm. Just make sure you clamp the sense circuitry appropriately
so the signal doesn't fry any devices.

There are magnetic (hall) devices that work similarly to the variable
reluctance sensors but respond right down to DC. I've got samples to
try with our tacho module but haven't gotten around tuit.

Best regards,
Spehro Pefhany

 

[MikeM]

I wish to monitor the rotational speed of a 2" shaft in the 10 to 400 RPM range. I could attach a magnet to the shaft and position a pickup coil to generate pulses. What I am looking for is a simple, very basic circuit that will process these pulses and drive a 3 digit LED display showing RPM. The circuit should be able to be built on a Radio Shack project board out of easily obtainable parts.

Basic tachometer can be built by having the shaft pulse trigger
a one shot. The average voltage at the output of the one shot
is reasonably proportional to shaft speed. This voltage could
be displayed on a Digital Panel Meter module.

Biggest problem is that when the shaft is turning slow,
the pulse rate at the output of the one shot could be as slow
as 0.3Hz. It is very hard to low pass filter the one shot's
output to get rid of the ripple. It would take too long for
the filtered output to settle to a new value after an RPM change.

In the real world, RPM measurement is usually done by timing
the period bewteen two successive pulses, a simple job for
a PIC style microcontoller chip. To display the RPM, it is
necessary to divide a constant by the Period, convert to
BCD and then drive a display.

For example, if your shaft is turning 600RPM, that is
10 pulses per second, or a Period of 100ms.

Suppose that you use the Pulses to gate a fixed 10KHz oscillator
into an upcounter (Timer register). After 100msec, you will have
accumulated 10000 x 0.1 or 1000 cycles.

What constant divided by 1000 yields 600? Answer is 600000.
So when each pulse arrives, the microcontroller has to divide
600000 by the value in the timer register, then zero the
register for the next cycle.

It then has to convert the dividend to decimal, and drive
3 digits worth of seven-segement displays...


Or you can buy this:
http://www2.towerhobbies.com/cgi-bin/wti0001p?&I=LXPT31&P=FR

 

[Spehro Pefhany]

Basic tachometer can be built by having the shaft pulse trigger
a one shot. The average voltage at the output of the one shot
is reasonably proportional to shaft speed. This voltage could
be displayed on a Digital Panel Meter module.

Biggest problem is that when the shaft is turning slow,
the pulse rate at the output of the one shot could be as slow
as 0.3Hz. It is very hard to low pass filter the one shot's
output to get rid of the ripple. It would take too long for
the filtered output to settle to a new value after an RPM change.

In the real world, RPM measurement is usually done by timing
the period bewteen two successive pulses, a simple job for
a PIC style microcontoller chip. To display the RPM, it is
necessary to divide a constant by the Period, convert to
BCD and then drive a display.

For example, if your shaft is turning 600RPM, that is
10 pulses per second, or a Period of 100ms.

Suppose that you use the Pulses to gate a fixed 10KHz oscillator
into an upcounter (Timer register). After 100msec, you will have
accumulated 10000 x 0.1 or 1000 cycles.

What constant divided by 1000 yields 600? Answer is 600000.
So when each pulse arrives, the microcontroller has to divide
600000 by the value in the timer register, then zero the
register for the next cycle.

It then has to convert the dividend to decimal, and drive
3 digits worth of seven-segement displays...


Or you can buy this:
http://www2.towerhobbies.com/cgi-bin/wti0001p?&I=LXPT31&P=FR

Or this:
http://www.trexon.com/pdfs/tachulator_module_revB.pdf

(but I'm biased on the matter)

Best regards,
Spehro Pefhany

 

[MikeM]

Or you can buy this:
http://www2.towerhobbies.com/cgi-bin/wti0001p?&I=LXPT31&P=FR

Forgot to mention. The Tower tach is designed to count
two optical pulses per shaft rotation (two bladed propellor)

 

[Rich Grise]

I wish to monitor the rotational speed of a 2" shaft in the 10 to 400 RPM

range. I could attach a magnet to the shaft and position a pickup coil to
generate pulses. What I am looking for is a simple, very basic circuit that
will process these pulses and drive a 3 digit LED display showing RPM. The
circuit should be able to be built on a Radio Shack project board out of
easily obtainable parts.

If you have access to the mechanical stuff (which I presume you do, since
you're talking about putting a magnet and sensor) you might try what I did
once:
Made a slotted disk with 60 slots, and used an optical interruptor (LED/PHD)
go get 60 pulses per rev. Then just count how many pulses in a one-second
gating interval, and there's your direct RPM. At the time, a colleague of
accomplished the same thing with only one pulse per rev, but this wacko
counted milliseconds, used that to preload a divide-by-n counter, which then
clocked the RPM counter. Or something like that. To this day, I don't
understand how he did it.

Good Luck!
Rich