L6470 int_speed limitation - not compatible with my motor

ran · ‎2013-04-24

Posted on April 24, 2013 at 15:21

Hi,

I'm exploring the EVAL6470H with my stepper motor.

according to the motor electrical parameters, the INT_SPEED should be around 2000 steps/s. The problem is that the L6470 has a maximal limitation of 976 steps/s for this parameter.

So if I want my application to work at 6000 steps/s, I cannot get a steady current while accelerating to that point (I can, though, ''trick'' the parameters so I would get the wanted current at 6000 steps/s).

Why do you limit the INT_SPEED parameter so low, and do you have a workaround solution for me, or is it impossible to use this motion control with my motor?

Thank you.

#l6470

Enrico Poli · ‎2013-04-29

Posted on April 29, 2013 at 11:49

In this case you can set the INT_SPEED parameter to the maximum value.

If you need more details to the voltage mode, you can consult the specific application note

http://www.st.com/st-web-ui/static/active/en/resource/technical/document/application_note/DM00061093.pdf

.

Regards

Enrico

ran · ‎2013-04-29

Posted on April 29, 2013 at 12:24

That is what I do, but since this is not the real INT_SPEED of my motor, I can't get a steady current in the acceleration/deceleration phases, and I have to calculate myself a workaround ''final slope'' that will give me the desired current at the desired speed (for every wanted speed, a different ''final slope''!).

Enrico Poli · ‎2013-04-30

Posted on April 30, 2013 at 13:35

I see.

Can you post the parameters of your application and motor (target current, supply voltage, Rm, Lm, ke, etc)?

Enrico

ran · ‎2013-04-30

Posted on April 30, 2013 at 16:46

Sure.

target current - 1A

supply voltage - 24V

Rm - 6.2 Ohm

Lm - 1.88 mH

ke - 0.009 V/Hz

This gives an INT_SPEED of ~2100 step/s, but I'm limited to ~1000 step/s.

If I want to run at 6000 step/s for example, I can calculate a workaround final slope that will give the wanted current at that speed (by intersecting the graphs), and I get a steady current up to 1000 step/s, but from 1000 to 6000 step/s I get an overcurrent since the final slope starts too early.

Thank you.

Enrico Poli · ‎2013-05-06

Posted on May 06, 2013 at 11:03

Ok,

I made some calculations.

First of all also using the ''ideal'' compensation curve the device reaches the 100% of duty cycle before the 6000 step/s speed (at about 4600 step/s). That meas that above this speed the voltage sinewave applied to the motor is not further increased.

Try this configuration and let me know how it work:

kval_acc = kval_dec = kval_run = 0x42

st_slp = 0x06

fn_slp_acc = fn_slp_dec = 0x0b

int_spd = 0x3fff

Enrico

ran · ‎2013-08-29

Posted on August 29, 2013 at 11:33

Hi Enrico,

Sorry for the late response.

You are right, and I do encounter a decreasing phase current starting from around 5000 steps/sec due to lack of BEMF compensation (driver is maxed out).

I understand there's nothing I can do to resolve this issue, except for using higher motor voltage (which I don't want to do currently).

Will the L6472 be able to drive my motor with the same specifications? or would it also be unable to reach the target phase current at high speed?

Thanks,

Ran.

Enrico Poli · ‎2013-09-04

Posted on September 04, 2013 at 16:45

Hi Ran,

The different control mode implemented in the L6472 could give you a bit of extra current (and torque). However the limitation is due to the supply voltage, so it is still present.

Enrico

ran · ‎2013-09-12

Posted on September 12, 2013 at 11:00

Hi Enrico,

I tried to play a little more with the settings of the L6470 (using the SPIN eval software) to see if I can get better results at high speed, and I found out that if I change the Step Mode to Full Step I do achieve the desired phase current at high speed.

However, if I use anything other than Full Step while setting Full-Step Speed to a value much lower than the intended speed, I do not get the wanted phase current (although it does change to step mode).

This is a bit unexpected, as I would assume going to full step after passing the Full-Step Speed would behave the same as using Full Step mode from the start.

This finding also show that electronically it is possible to achieve the wanted phase current at my desired speed, it's just a driver issue.

I would appreciate your feedback on my finding.

Thanks,

Ran.

Enrico Poli · ‎2013-09-22

Posted on September 22, 2013 at 11:00

Hi Ran,

When the device switches between microstep and full-step using the Full Step Speed threshold the amplitude of the voltage squarewave (Full-step) il limited in order to avoid discontinuities (see respective section in the device datasheet).

So your finds are correct: the automatic full-step gives a lower voltage (and then a lower current) that the normal full-step.

Enrico