I have some old, real old aircraft instruments that have air core movements, I'd like to be able to drive these from a 'duino.
I've used micrcontrollers for a while with assembler, I have recently moved towards arduino and C programming.
There isnt much infor on the net for air core movements, surprisingly as nearly all vehicles from the 80's-90's used them for their instruments.
I have just completed a messy program that uses a bldc motor as a servo, this is going to be real similar, my air core radar guages are 2 pole making things simpler, they will position right off with the correct waveforms applied.
For analog meters you use PWM. As they have a coil use the same circuit as the one for controlling a motors speed but not its direction.
Try it out with an old fashioned AVO meter set to read voltage in the range 0 to 6 volts read the voltage on pin 13 and run the fade example. In this case no other hard ware is required as your already setup for 5v.
The code this guy has used is diffrent to the idea I had by that he's calculating the pwm's using trigonometery, I was thinking a lookup table or array.
Might try that.
He's saying in the first post that the movement he's playing with gets notchy at certain points as it rotates, I think I know why that is, pwm isnt good enough to control the coil current, something like this really needs a current source, such as a pwm to voltage to current circuit.
I have no experience with an air core motor, but it seems that it basically nothing more than a couple coils.
you have one on, the other off and the needle is solid at say 90 degrees. switch that off, the other one and is presto, 0 degrees,
both have half power, it would be at 45 degrees....
it would seem to me that the electronics that are in the car or plane, would have what you need and all you need to feed it would be the sensor input such as a pulse from the speedometer ?
if it does work that way, then the power needed to hold it at 89 degrees would be the issue you have for it 'sticking'.
what comes to mind is that if figure you need 1 unit of power at 90, and half from each coil at 45, follow the sine/cosine to derive power needed at the points to get relatively linear output, you would come to something of a nearly full power coil at 90 and almost no power at the adjoining coil. this would have you charging one coil nearly full, then discharging it. this takes time.
since you are near the coil, you could have the coil charge only 1/4, or less, and then have the second coil pull it off center to show the 89....89.1... etc. since you are near the coil (alignment of stator and rotor), you do not need to charge the coil fully. just think of the power reaching the rotor at the 45 mark. you need even less.
I have not read the data sheet on the MLX10407 but it might be that this is incorporated in the chip that the rotor gets a relatively even power at all 360 degrees of rotation?
Agreed, mostly.
I think the issue with pwm is the edges, the coils have a degree of inductance, current control applies full volts till the programmed current is reached improving things.
I'm going to try the link I posted earlier.
