I solved this problem! I am using bldc servo amplifier with two sinusoidal 
input, the third one is computed by the amplifier itself. A feed the amplifier 
with bldc.A-value and bldc.B.value, than i changed this to bldc.A-value and and 
bldc.C-value and success :) now the motor moves as suppose to (i have to tune 
the pid parameters).

I have one more problem. Now when i start the bldc.init the motor moves the 
opposite direction, away from index marker. And i can not change this (i can 
not twist the phases because of previous problem). Is there a way to somehow 
change the home direction or should i change the bldc.comp?

Regards
Klemen





________________________________
 From: Klemen Dovrtel <klemen_dovr...@yahoo.com>
To: andy pugh <bodge...@gmail.com>; Enhanced Machine Controller (EMC) 
<emc-users@lists.sourceforge.net> 
Cc: Peter C. Wallace <p...@mesanet.com> 
Sent: Sunday, February 24, 2013 9:45 AM
Subject: Re: [Emc-users] hal - bldc - linear motor
 
Thank you Andy, now i see this, i looked at the source code...

Now i set up a clarke.comp and feed it to with sin/cos signal. I did one 
electrical revolution (2pi) and measured the motor linear move - tt was 60000 
encoder counts = 30 mm. Then I indexed the motor and measured the encoder 
offset, it was -0.7 rd.

Then i set up bldc.com again. I set the bldc.scale = 60000 and bldc.poles = 2 
and tested it. I indexed the motor and then moved it manually. I can see the 
rotor-angle is 0 at x=0 mm and again at x = 30 mm. So this should be fine now.

Then i set the encoder-offset to -0.7/2pi *60000 = -6687 encoder counts. I 
tested the set up again and it seems ok, the rotor-angle is 0.11 at x=0 mm and 
again at x = 30 mm.

Ten i feed the bldc.value with some value. The motor just moved a bit and 
stayed still. But if I understand this correctly the motor should accelerate? 


I don't get this. I than randomly tried different  encoder-offset values and 
the result was always the same, the motor just moved a bit and stayed still 
again.

Any idea what could be wrong?

Regards,
Klemen




________________________________
From: andy pugh <bodge...@gmail.com>
To: Klemen Dovrtel <klemen_dovr...@yahoo.com>; Enhanced Machine Controller 
(EMC) <emc-users@lists.sourceforge.net> 
Cc: Peter C. Wallace <p...@mesanet.com> 
Sent: Sunday, February 24, 2013 1:12 AM
Subject: Re: [Emc-users] hal - bldc - linear motor

On 23 February 2013 16:30, Klemen Dovrtel <klemen_dovr...@yahoo.com> wrote:

> I noticed that at the index the rotor angle is set to 0. Is it possible that 
> there should be some offset regarding to real motor angle? I did not find any 
> rotor angle offset input in bldc.component. Is this the way to solve this?

The rotor angle should be the actual rotor position. With a linear
motor this is a slightly different concept, but ought to be a measure
of where the motor is electrically. If the index is not at the motor
electrical zero then you can use the encoder-offset parameter to
offset it.

I think a linear motor will always be declared as 2-pole to bldc (you
_might_ want to choose a different number if the encoder counts per
electrical cycle are not an integer, but I can't think of any other
reason).

If you need changes to bldc to suit a linear motor, I am open to suggestions.

-- 
atp
If you can't fix it, you don't own it.
http://www.ifixit.com/Manifesto
------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_d2d_feb
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users
------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_d2d_feb
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to