Air core meter movements.

I have some instruments that have air core movements I'd like to use, originally the pointer on the instrument only travelled 90 degrees, and I've a suspicion that the movement across the scale was no linear as I'll explain.

I have written some code that generates 2 sines waves, sine and cosine, the waves are generated by pwm, and to get proper sine I use a lookup table.

What I found is that the movements work but are way non linear, they start slow then sweep fast towards the centre of each quadrant, I metered the resistance of each coil sine and cosine, ones 160 ohm, t'other is 260 ohm, I think this counts for the non linearity, I was wondering if its a good idea to pull all the wire off and rewind both windings with the same no of turns to see if that makes them linear.
What I want to do is remove the stops and get them to work over 359 degrees sweep.

Decrease the amout of voltage proportionally to the 260 ohm coil. i.e. assume it has more turns so you have A* (sin(x) and 160/260 cos(x)); "A" is optional.

Maybe its too late at night for me.
I see what you say about the higher imp coil, not an option with a l298 driving it though, the rail is commoned.
I added resistance to the 160 ohm coil to make it 260 ohm, so the current through it would be similar to that of the 260 ohm.
That didnt work, the current might be similar but as your statement implies turns are not the same so flux will be well out.

I have 2 ideas, one to do away with the H bridge and drive the coils with 4 transistors on a split supply with the centre to ground (how they are connected as is) and make the 'ground' adjustable voltage so I can balance out the turns diffrence at 1/2 delfection.
The other is to filter the pwm to dc and use that to drive a voltage controlled current source for each coil.

I might have to pull all of the copper off to get to the end stops in which case I might as well wind the whole thing from scratch.

Ok so I tried powering the 2 coils from an independently adjustable split supply, the idea was to try and get the pointer to setlle at half way point and see what the voltages were.
Its very sensitive, I cant get the pointer to settle, it just swings from one side to the other at a certain voltage, probably right that being driven from dc, I have -10v on one coil and +5v on the other when it just starts to swing from fsd to lsd, kinda makes sense.
So I want to rig something up where I can independently adjust the voltage to each coil in the prototype circuit, I'm thinking an adjustable 'ground'.

Mechanical meters generally are uA or mA meters, and you add a shunt in parallel to make it read amps, or a multiplier in series to make it read volts.

So try feeding it through a series resistor, see if that calms it down - try 10K for a start.

Hi Nige, I tried something on those lines.
Air core meters have 2 coils and a moving magnet, the magnet orientates itself to the mean angle of the magnetic flux produced by 2 sine waves.
I think I'm trying the impossible again, a 360 degree air core needs both windings the same, this movement is designed to run of one phase an only travel 90 degrees.