Self accelerating reed switch magnet spinner.

[MileHigh]

TK:

Well, if you're happy then I'm happy!  It's amazing the sensitivity of the op-amp inputs.  I just like CONTROL and that setup does indeed give you control.

I just had an interesting idea for next year's Pulse Motor Build-off.  Everybody has to do it based on a VCR helical tape head spindle and the target is the least amount of input power for a given high RPM, say 10,000 RPM.  Or perhaps a hard drive spindle would be even better and they are more readily available.  I was going to say "maximum RPM" but I would hate to see anybody get a shard of metal flying right through their cheapo protective goggles and into their eye.  Also, you are disqualified if you don't purchase some kind of transparent tube of protective plastic to put over the pulse motor.  You could define some standard plastic tube spec that anybody could buy at a Big Box hardware store.  And you have to clearly demonstrate how you make your input power measurement.  You get bonus points if you don't rely on your digital multimeter to heavily.  Can you get creative and use some analog tricks to measure the power input?  Somehow work in a Wheatstone bridge?  Same thing if you have a power output measurement to do.

That might be pushing the envelope too much but you never know.  You can always completely change gears and have an egg dropping competition instead....

MileHigh

[MileHigh]

Well, I am going to try to finish off my "cheapo watt meter" design.  I am sorry to say in Windows 7 I don't think there is a built-in object oriented graphic program for dummies.  I went to the Cnet Downloads page, which used to be my favourite place for safe free software, but it's been corrupted as far as I am concerned and I don't want to struggle with them refusing toolbars at every step and the whole nine yards.  Way back in the 90s there was a great freebie object oriented image editing software program but I forget what it's called.

Okay, here goes.

Here is the project:  Let's assume that you have two fresh 12.6-volt batteries in series powering your pulse motor.  Let's assume that the average power consumption of the motor is seven watts.  So we want to have an op-amp configuration that computes the average of the current pulses coming out of the 25.2 battery voltage source and outputs that average current value as seven volts DC to your cheapo multimeter.  Makes sense?  So the multimeter display is showing seven volts DC which actually means seven watts of average output power.  So effectively you will be using op-amps to construct an analog computer to accomplish this task.

We will power the op-amps with +/- 9 volts using two 9-volt batteries.  That makes for a compact setup.  You could use 12-volt batteries for more headroom but they are large.  The op-amps have to have +9 volts and -9 volts, and you want to have a ground handy (junction of the two batteries) because you will need it.  I wall call that Gnd9V to avoid possible confusion.

Let's look at the logic behind the design.  If the pulse motor is consuming seven watts, and the voltage source is 25.2 volts, then the AVERAGE current flow (not the pulsed current flow) is (7/25.2) = 278 milliamperes.  So the op-amp analog computer just has to sense the current pulses, and then average them to a DC value, and with that average current computation output seven volts DC and that represents seven watts of power consumption by the pulse motor.  (In this design the averaging and the requirement to output seven volts will be combined into one step.)

The design is very basic and I am sure there are better ways to do it - but this should work.

IMPORTANT:  This design is so "dumb" that it will only work for a fixed source voltage.  Of course it can be adapted to work with different fixed voltages, but it's still pretty dumb.

Note also that it will work for measuring the power INPUT to the pulse motor by the source battery, AND it can just as easily work for the power OUTPUT by the pulse motor into the charging battery.

[MileHigh]

Okay, with great trepidation I will try to describe the circuit.  We are going to try using a 0.1 ohm current sense resistor.  But, everything could work just as easily with a one-ohm current sense resistor but certain other values would naturally have to change.

Here is the pipeline:

[pulsing output from source battery] -> [current sense resistor] -> [unity gain voltage follower] -> [inverting amplifier with variable amplification] -> [RC averaging filter] -> [multimeter on DC voltage with inverted probes]

[unity gain voltage follower]
The input for the unity gain voltage follower is on the battery side of the current sensing resistor.
The Gnd9V is connected to the other side of the current sensing resistor.
The only thing this section does is buffer the voltage waveform from the current sensing resistor without loading it.

[inverting amplifier with variable amplification]
The inverting amplifier is where we need to figure out how much amplification we need to generate the seven volts average output.  Let's assume that we have a 1 Kohm input resistor, and a 20 Kohm 10-turn trimpot as the feedback resistor.  So that means the gain for this amplifier can vary between zero and -20.
The inverting amplifier is amplifying the current sensing resistor waveform so that the average voltage of the pulse waveform is negative seven volts.

[RC averaging filter]
The input to the RC averaging filter comes from the output of the inverting amplifier.
The RC averaging filter consists of a 1 Kohm resistor that connects to a 1000 uF capacitor which connects to Gnd9V.
The output of the inverting amplifier connects to the 1 Kohm resistor.
This filter converts the amplified current sensing resistor waveform to a steady negative seven volts DC.

[multimeter on DC voltage with inverted probes]
The red probe of the digital multimeter connects to Gnd9V.  The black probe of the multimeter connects to the junction of the 1 Kohm resistor and the 1000 uF capacitor.
This is your wattage output display -> positive seven volts.

Okay, one more post to crunch some numbers to see if everything will work.

[MileHigh]

Number crunch time.   We are designing this analog computer based on a 25.2-volt voltage source, i.e.; two fresh 12-volt batteries in series.  We will crunch the numbers assuming a seven watt average power output by the batteries into the pulse motor.

The hypothetical average voltage across the 0.1 ohm resistor is 0.278 amps x 0.1 ohms = 0.0278 volts.

That means the op-amp gain to get 7 volts out is (7/.0278) = 252.

I don't like this, the sense voltage is too low for comfort and the amp gain is too high.  Let's change the 0.1 ohm current sensing resistor for a 1-ohm current sensing resistor.

Wih a 1-ohm current sensing resistor note the hypothetical average voltage across this resistor is 0.278 volts and the amplification that we need is now 25.2.   That makes a lot more sense.

So that means that the inverting amplifier can still have a 1 Kohm input resistor.  However, we will change the feedback resistor to a 30 Kohm 10-turn trimpot.   So now our amplification can vary between zero and 30, just what we need.

I am also having second thoughts about the 9-volt batteries.  I am concerned about any possible clipping of the amplified current sense waveform because of limited voltage headroom.  So I would check that on a scope and if there is clipping, then the power for the op-amp would have to be changed to +/- 12 volts.

So in summary:  1-ohm current sensing resistor, 1 Kohm input resistor, 30 Kohm 10-turn feedback variable resistor, 1 Kohm RC filter resistor, 1000 uF RC filter capacitor.

Note that in real life, some of this may need tweaking.  I am really concerned about the amplitude of the current pulses causing clipping at the output of the inverting amplifier.  You could lower the amplification of the inverting amplifier for example by half.  So that means that your desired negative seven volt output representing seven watts would be reduced to negative 3 1/2 volts.  By doing that you give the op-amp twice the headroom for doing the averaging function in concert with the RC filter connected to the op-amp output.  The down side is that you have to add a third op-amp as a 2X inverting amplifier to double the voltage output so that the multimeter reads seven volts again.

[MileHigh]

Now the fun part after you do all that work and all that number crunching.....  How hard or how easy is it to adjust this analog computer to make it do what you want it to do?

The good news is that it's pretty darn easy.   I am going to assume that you can set up a variable current source of some form or other, but even that isn't necessary.

Let's work with the assumption that six watts of average power is supplied by the 25.2-volt source battery for starters.

So the current that you have to put through the current sensing resistor is (6/25.2) = 238 milliamperes.

Therefore, you put your digital multimeter on current measurement and put that in series with the 1-ohm current sensing resistor.  Put some kind of variable voltage source across these two components (throw in a resistor in series with the voltage source if that makes it easier) and adjust the current so that the digital multimeter shows 238 miliamperes.  Then, adjust the 30 Kohm 10-turn trim pot so the output from the analog computer reads six volts.  That's it, the analog computer is now ready to make average power measurements of the pulsing current coming out of your pair of 12.6-volt batteries in series.

Another example...

Now let's work with the assumption that eight watts of average power is supplied by a single 12.6-volt source battery.

So the current that you have to put through the current sensing resistor is (8/12.6) = 635 milliamperes.

Therefore, you put your digital multimeter on current measurement and put that in series with the 1-ohm current sensing resistor.  Put some kind of variable voltage source across these two components (throw in a resistor in series with the voltage source if that makes it easier) and adjust the current so that the digital multimeter shows 635 miliamperes.  Then, adjust the 30 Kohm 10-turn trim pot so the output from the analog computer reads eight volts.  That's it, the analog computer is now ready to make average power measurements of the pulsing current coming out of a single 12.6-volt battery.

MileHigh