LTC_1298 12 Bit ADC + Linear Motor = Dead ADC????

TJHJ

Ok Hi all,

I am driving a 1/4 HP linear motor w/ position feedback. It is basically an 8k pot that provides a voltage signal for position. I am driving the motor using the VNH3SP30 integrated H bridge Driver from ST. They are all on a common board, running out through an amp tyco weatherproof connector. The Feedback pot, has its own direct to ground pin and power being pulled from the 5v regulator on the board. I am only using 1 channel of the LTC1298. The other channel is tied to ground through a 1k resistor.

There is another LTC1298 driver on the board, providing a different function, it continues to work fine.

The same prop chip is running the motor driver, both ltc1298's and other functions correctly.

I am using the standard CD_LTC1298 object to get information from the ADC.

If I put a new LTC1298 chip in the board. and do not move the motor. It works fine. I get a reading that makes sense. For the voltage at the pin. Once I move the motor, I get nothing or very strange values. Like 2v at the pin and the adc is reading a value of 7.

I am at a loss here? Any ideas?

Back EMF feedback killing it? If so how would I filter/ protect for this?

Thank you,
TJ

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I owe everyone here a bunch, So thanks again for answering my dumb questions.
Projects. RG500 ECU system. PropCopter. Prop CanSat. Prop Paste Gun.
Suzuki RG500 in a RGV 250 frame.
Bimota V-Due (Running on the fuel injection system)
Aprilia RS250

Leon

Motors like that require very careful controller design to avoid problems with transients affecting the electronics.

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

Rayman

You might try adding a capacitor to ground on the input to the ADC. That will keep the voltage there steady...

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My Prop Info&Apps: ·http://www.rayslogic.com/propeller/propeller.htm

Toby Seckshund

1/4 horse !?!

what volts/current and from where? (You may have noticed that I have an interest in PSUs)

most of problems that arise from heavy current are because the independant ground, power and signal are not independant. Earth loops are my second best earner (after Smile PSUs).

They are usually a right (expletive) to sort out (ker-ching...). check that the ground return etc is totally isolated ie three inipendant wire not connected to chassis etc, all the way back to the prop-a/d and that mountings etc do not comprimise this, the "star" arrangement must be maintained, especially with the reaction speed of the prop.

If the fault is truely killing the a/d then negative ground volts may be more of a problem than positive ones, 1V neg is worse, tham 1V positive, for substrate reversal.

To check if the back EMF is the killer then put the motor into a known posision and do not move it, check the readings, try toput into another posision read again and then motor to another position and see if it kills another, sacroficial but may prove something

TJHJ

@ Toby,
DC 12v 20 amps Motion Industries Built the actuator it drives a 240 lbs rack around.

Thats the approach I took. Put one in, Check it, Move it... Dead. Put a new one in, Good, move it dead... Rinse and repeat... At 10 bucks a pop really starting to get agitating.

Anyone got any tricks to snub the emf noise at the motor?

@Rayman

Should it be in series? or ????

Thanks again all,
TJ

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I owe everyone here a bunch, So thanks again for answering my dumb questions.
Projects. RG500 ECU system. PropCopter. Prop CanSat. Prop Paste Gun.
Suzuki RG500 in a RGV 250 frame.
Bimota V-Due (Running on the fuel injection system)
Aprilia RS250

Tubular

Couple of things.

I was thinking increase the source resistance from the pot, and put a reverse shunt diode in to shunt any negative voltages safely away, but you can't do that because...

From the graphs on the data sheet you really should be driving it with a relatively low source impedance. The graphs only go up to 10k and there is already weird stuff happening by then. An 8k pot probably isn't low enough without introducing signifcant errors and nonlinearities. So, put a unity gain op amp (voltage follower/buffer) between your 8k pot and the ADC. Socket it so that you blow the op amp if anything, not the ADC. Still put a series resistor from the wiper, then add a shunt reverse diode and cap (as Rayman suggested) at the input to the op amp. You want a rail to rail op amp.

Due to dynamic effects the wiper of the pot could well be going open circuit for brief periods, capacitively picking up stray voltages, and thats whats blowing your ADC

tubular

Rayman

Try putting a 10k resistor in series from you pot to the ADC and then an 1n or 10nF cap from the ADC input to ground... If that doesn't work, you may need some opamp buffer like Tubular says...

You often need a cap at the input to an ADC because it draws a little current when it samples and you can't let that change the voltage there...·

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My Prop Info&Apps: ·http://www.rayslogic.com/propeller/propeller.htm

Post Edited (Rayman) : 6/14/2009 10:05:07 AM GMT

Toby Seckshund

With such huge volts and currents, in comparison to CMOS, the adc and everything else will have to be armour plated against everything on the big bad outside world. You have certain proof that you have a hostile enviroment.

I would try series resistors to all of the pot's wires, small incomparison to the 8K say about 470 Ohms ish and have clamping diodes to the rails after this, together with a good ESR cap to deck, small inductors may help too. The errors at the top and bottom pot reading can easily be compensated by software and I doubt that the pot variations are fast.

Make sure that all power feeds come from the source in a star so that there are no rail induced lumps and bumps, perhaps separate the thinking bit's supply from the grunt's supply. Inductors and caps on the motor will have to be of biblical ratings!

Thinking about the speed isssue, do you actually require such a good adc, is the resolution of the position pot that good or could sigma delta be good enough. You would still have to do all the protection tricks though as the prop is them in the firing line. At least the prop is cheaper than the ADC

Post Edited (Toby Seckshund) : 6/14/2009 10:37:30 AM GMT