ST7FMC Starter kit

gernot · ‎2004-11-21

Posted on November 22, 2004 at 07:09

ST7FMC Starter kit

gaetano · ‎2011-05-17

Posted on May 17, 2011 at 10:17

Hello,

my feeling is that your PI parameter Kp is too high (=255).

Reduce it at very low value (10 or 20 at max) and try again.

Regards,

Tanio

gernot · ‎2011-05-17

Posted on May 17, 2011 at 10:17

Hello,

i'm running the ST7FMC Starter Kit by Softec Microsystems and have some troubles with GUI-crashes. These crashes seem to be depending on the Parametervalues choosen.

I have two motors which i want to drive.

First one is working well. With parameters, approbiate for the second motor the GUI crashes immediately after the programm and parameters are flashed. This error is reproduceable. Are there any limitations, which are not checked by the GUI?

Thanks for your help.

gernot · ‎2011-05-17

Posted on May 17, 2011 at 10:17

Hello,

thanks for your tip, but it is still not working. I set kp to 20 but the PC-SW is still crashing. Then i tried to run the motor in sensorlessmode, see item 2. Sometimes it still crashes, but not that often. I see no difference on different values for kp. It seems to me, that the choosen ''Detection Mode'' is the important point. I did some other tests, here are the results:

1) AMETEK Motor in sensorless mode. Result: working well, but it seems to me, that the rotor of the fan is running in the false direction. When i look through the topcover of the fan the rotor is running clockwise, while the tangential air-output of the fan suggests a counter clockwise rotation. So i checked the wiring of the motorphases. red-J11.1, yellow-J11.3 and black-J11.5. This wiring is chown on the poster supplied with the eva-board. When i change the red one with the yellow one, the airflow across the inverterstage is much faster.

2) I tried to run the motor discussed in the previous posting in sensorless mode, but i failed. the motor is accelerated to a speed of 600 rpm with a BEMF of about 1.7 Vpp, i think this should be enough. When motor is freewheeling after failed starting, i measure a offset of 7.5 V. This offset seems to be generated by the +5V, +15V supply. To which level should i set the zero crossing threshold voltage? Do i have to consider the offset voltage?

One more question:

When using a BLDC-Motor with 120Â° sensors, how is the phase angle between motorphases and Hall-IC. How is polarity of the Hall-IC-Signal? E.g.: Hall-IC1 is switching from +5V to gnd, when Motorphase 1 is zerocrossing from + to -.

Thanks.

gaetano · ‎2011-05-17

Posted on May 17, 2011 at 10:17

Hi,

the detection mode, theoretically, shouldn't affect the behavior of the system in term of PC crashes. Obviously, the use or not of the hall sensors can induce differences in the working mode that, at the end, cause the problem.

May you specify what kind of motor are you using (high/low voltage and the rated current). In sensorless mode how big is the startup current when the motor start successfully?

Concerning you comments:

1) Probable you are right: from the fan profiles it seems that the motor should run in counterclockwise. In spite of that, the air flux comes out in both direction, and this is the important thing.

2) It is strange! Repeat the same test in the same configuration by using the Ametek motor and let me know about your voltage when free wheeling. When in sensorless, don't forget to change the position of the W14, 15, 16 jumpers in high position!

The threshold is fixed as much lower as possible (theoretically 0V) your sistem allows to do. In no any case, you should take care of an offset voltage.

3) I am attaching a technical note to clarify about the sensors arrangement when in 120 degrees and in 60 degrees.

Keep in touch,

Tanio

gernot · ‎2011-05-17

Posted on May 17, 2011 at 10:17

Hello,

the motor i am using is a 70 V motor, the power of the motor is about 200W.

I'm doing some more tests in sensormode to evaluate the current for successfull starting.

Offsetvoltage:

the Offsetvoltage seems to be of no importance. As you see in the attached file it is there, before the motor is started, when it's freewheeling and when it is stopped. Picture was taken using the AMETEK-motor in sensorless mode.

I'll come back, when finished.

gaetano · ‎2011-05-17

Posted on May 17, 2011 at 10:17

Hi,

is the changing from 20 to 1 gradual or abrupt?

During the initial phase, you should decrease gradually from one coeffecient to the next one: this avoid bad working conditions for your system.

Let me do an example:

if your motor is not compatible to have delay coefficients equals to 1, as you are trying to do, the motor starts to lose synchronism causing a rotor blocked conditions at the end.

This event comes with a big increase in the motor current: if your power supply has a limitation in the current capability, this causes a drop in the effective voltage applied to the system, causing a system crash because the Starter kit is not more supplied with a right level of voltage and the ICC communication is abnormally interrupted.

Hence, for avoiding to cause all these troubles, it is a good rule to change a little bit the delay coefficients and see what happen in the system, monitoring, for instance, the shape of the motor currents and try to keep in synchronism the phase current with the related BEMF.

Keep in touch,

Tanio

gernot · ‎2011-05-17

Posted on May 17, 2011 at 10:17

Hi,

i checked the current in sensormode: With the parameters in the attached file, the motor is running. But if i set the delay coefficients to 1 the GUI crahses.

The shape of the current isn't what i expected. but i am not abel to tune the current shape due to the crashes of the GUI. Could you please check why the GUI is crashing. It isn't possible to work with, if you have to restart it every few minutes. I'm adressing this point to Softec, too.

What about the software for sinusoidal currentshape in combination with sensors. I read somewhere in the forum, that it should be relaesed in Q3/2004, but i haven't found it yet. Where can i get the software and are there hardwarechanges on the EVA-Board necessary.

How is the currentmeasurement done in this case. Do i have to supply a shuntresistor in each leg of the inverter, or is one shunt, as built on the EVA-Board sufficient?

Thanks

gernot · ‎2011-05-17

Posted on May 17, 2011 at 10:17

Sorry,

but i forgot to attach the mentioned file. So here we are...