I'm trying to use a little DC motor as a speedometer to measure the crank speed on a bicycle. Initially, I connected it to a 2n2222 transistor, and found that it worked well for what I needed, but only if I spun it at very fast RPMs (1500+). I need it to measure RPMs in the 70-100RPM range. I looked at a mechanical solution and geared it up to 1000 RPM, but it was not enough. I decided to use a Darlington pair of 2n2222's to increase the sensitivity and add a 5v power supply. Here's the problem. The Darlington Pair is sensitive enough that if I touch the Base, it fully saturates the pair. But when I hook the motor up to the Base and Emitter, I have to spin it quite fast to saturate. What am I doing wrong? I've got a 470ohm resistor on the collector, and I've tried various resistances up to 100k on the Base, but all have the same effect. The end goal of the circuit is to vary the voltage C->E such that at 100 rpm it's fully saturated and below that is partial. At 100rpm the motor puts out around 30mV.

 

Normally a PM field DC motor, the output voltage is proportional to the RPM?
What voltage is the motor rated at?
If the motor is rated at 12v at 1000 rpm, then 12v is the voltage it will generate at 1krpm.
If the output is too high for your application, just use a suitable resistive divider?
Max.

 

Normally a PM field DC motor, the output voltage is proportional to the RPM?
What voltage is the motor rated at?
If the motor is rated at 12v at 1000 rpm, then 12v is the voltage it will generate at 1krpm.
If the output is too high for your application, just use a suitable resistive divider?
Max.

The motor is rated 1.5-3v and it has a maximum RPM of 8,300RPM. I can assume that's 3v. I suppose a motor rated at 12v at 1000 rpm would be perfect if I had it, because that's 1.2v at 100 rpm which should saturate the transistor.

My suspicion is that the output is too low for my application, that's why I tried it increase the sensitivity of the input.

 

Ideally you need a servo motor tach, these are 7 to 20v/1krpm.
This is an example ebay 260620985959
Max.

 

Hmm, that would work, but I'm bothered by my darlington pair not working for this, and I just hate an unsolved mystery.

 

All the bike 'computers' I've seen use a hall sensor and magnet to collect the crank and wheel data. Then use that for pedal count and distance.

 

The transistor only starts to turn on with around 0.6V at the base so when the motor is producing less than that, nothing much will happen. If you add a voltage divider to bring the base up to about 0.6V and add a capacitor in series with the motor to remove the DC component then you will see something happening. Have a look at the schematic near the bottom of this page:
http://www.allaboutcircuits.com/vol_3/chpt_4/9.html
The whole chapter, on the menu at the left, is pretty detailed.

 

All the bike 'computers' I've seen use a hall sensor and magnet to collect the crank and wheel data. Then use that for pedal count and distance.

+1
It's so much easier and precise to go digital right at the sensing step. Other than missing the pass of the magnet, I can hardly think of any condition that would ruin the accuracy. With an "analog" approach of spinning a motor, I can hardly think of anything that won't affect accuracy.

 

With an "analog" approach of spinning a motor, I can hardly think of anything that won't affect accuracy.

With Tach units such as Servo-tek that have been used for several decades in servo drive feedback in the velocity loop, the control is surprisingly accurate.
Max.

 

With Tach units such as Servo-tek that have been used for several decades in servo drive feedback in the velocity loop, the control is surprisingly accurate.
Max.

But that is an entirely different situation. A servo motor is a more stable rotation, not a variable one pedal stroke to the next.

From the Servo-tek site - "DC Tachometers with output signal between 1 and 10 volts / 1000 rpm " http://www.marshbellofram.com/servo-tek/types/servo-tek-sealed-dc-tachometer-generators/ The bike pedals won't give that kind of result. Maybe the bike wheels with some gearing for total speed though.

 

I was just pointing out that small DC motors such as servo-tek tach's or quality miniature motors such as Maxon, are capable of accuracy of output and go up to 21v/1krpm.
The servo motor mention was just an example of a analogue application.
Max.

 

Have you measured the output of your generator (motor) with a digital volt meter while rotating it at the expected speed? Digital panel meters that are 200mV full scale are fairly inexpensive and may be sensitive enough for your need.

I'm really confused as to the need or application of the darlington in your application. Can you post a schematic?

 

I was just pointing out that small DC motors such as servo-tek tach's or quality miniature motors such as Maxon, are capable of accuracy of output and go up to 21v/1krpm.

I accept that, but I guess I can't picture it in a cycling application where temperature can swing widely, the equipment may be briefly under water, covered with ice, snow, sand or mud, and so on.

I suppose you could solve these problems but my point was, why bother when there is such a good technology already in commercial use?

 

^^ For around $10 at Walmart.