Evaluation Tool

[caisunjunhe]

Posted on January 02, 2014 at 16:47

In the BEMF compensation,  How can we evaluate the speed by ourselves?

#evaluation-tool

[Enrico Poli]

Posted on January 21, 2014 at 16:09

Hi,

It is the maximum motor speed which is required by the application. The paramter is used as a check: according to the application and motor characteristics the phase maximum current decrease with speed. If the target current is above the maximum current, a warning message is shown. Regards Enrico

[sakaguchi]

Posted on May 28, 2014 at 04:23

Hello,

I have a question about Evaluation Tool(using L6470), too.

I set to Holding current and Accelaration(or Deceleration, Running) current to same value,

but Evaluated value by Evaluation Tool is a little different by KVAL_HOLD and KVAL_ACC (or KVAL_DEC, KVAL_RUN).

According to the ''First dimensioning'' section of an application note(AN4144),

all the formulas that calculate it are the same.

These values are why different?

(Please see an attached file)

Thanks

Hi,

It is the maximum motor speedwhich is required bythe application. The paramter is used as a check: according to the application and motor characteristics the phase maximum current decrease with speed. If the target current is above the maximum current, a warning message is shown. Regards Enrico ________________ Attachments : KVAL.png : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006Hz11&d=%2Fa%2F0X0000000bJN%2FXdO38L2jf14kRmfyhuQPnMWn_shPcnfvwpyi7DwIJg4&asPdf=false

[Enrico Poli]

Posted on June 04, 2014 at 14:33

Hello,

the difference depends on the fact that holding current is DC and the other currents are AC (sinusoidal). In the first case (DC) the rms value of the current is equal to the peak value, in the second one the peak is about 1.4 times the rms value. KVAL value must be evaluated according to the peak. However the holding current usually should be significantly lower tha the running current. Please, check if you really need to keep the holding current high. It could cause thermal issues. Regards Enrico

[sakaguchi]

Posted on June 06, 2014 at 02:31

Hello Enrico,

Thanks for reply.

I have another question about the formula of INT_SPEED register value in the Evaluation Software.

According to L6470 Programing manual p43,

Equation9)   intersectSpeed[step/s] = INT_SPEED * 2**(-24)/tick

On the other hand, from AN4144 Application note p15,

Table3)      INT_SPEED = intersectSpeed * 2**(26) * tick

                         where intersectSpeed = 4*Rm/(2*Pi*Lm) [step/s]

Although I think that these are the same pupose formulas,

the indices of the exponentiation of 2 differ.

In the Evaluation Software, it seems that the formula of Table3  is used.

But in the dSPIN_Firmware_Library_STM32 C language program, Equation9 is used. for example, this macro below.

/* Intersect Speed conversion, range 0 to 3906 steps/s */

#define IntSpd_Steps_to_Par(steps) ((uint16_t)(((steps)*4.194304)+0.5))

Which formula should I use?

Thanks

Hello,

the difference depends on the fact that holding current is DC and the other currents are AC (sinusoidal). In the first case (DC) the rms value of the current is equal to the peak value, in the second one the peak is about 1.4 times the rms value. KVAL value must be evaluated according to the peak. However the holding current usually should be significantly lower tha the running current. Please, check if you really need to keep the holding current high. It could cause thermal issues. Regards Enrico

[Enrico Poli]

Posted on June 25, 2014 at 16:50

The correct formula is the one indicated in Table 3, in fact in the current revision of the datasheet the Equation 9 (now Equation 10) had been fixed.

As a consequence, the formula of the STM32 library is wrong. Thanks for your valuable feedback! It will be fixed in the next release. Regards Enrico

[sakaguchi]

Posted on July 02, 2014 at 04:08

Thanks Enrico,

It clarified very much.

With the occasion, I have a worrisome part in AN4144 Application note p.17

in Example2

 Lm=4mH; Rm=10Ohm  4*Rm/Lm = 10000 step/s

This is 4*Rm/(2*Pi*Lm) = 1,592 step/s, isn't it?

And more, since maximum compensation rate is 0.04%,

if starting Kval=0, I think the rotation speed at 100% compensation is 2,500(=100/0.04) step/s. (In Application note, it is written as 250 step/s.)

Any way, I am now using L6470 where five devices are daisy-chained.

It works very well.

Thank you for making nice device!

[Enrico Poli]

Posted on July 07, 2014 at 09:57

Many thanks for your feedback on the application note!

The errors will be fixed in the next revision of the document. Kind Regards Enrico