Asymmetrical voltage in Puma. Why?

The7

10 kW
Joined
Jul 11, 2007
Messages
761
Location
Vancouver
http://endless-sphere.com/forums/viewtopic.php?f=2&t=4528

In the picture of AL1020, the applied voltage is symmetrical because the applied voltage is in phase with the BEMF.
Seems this is the common goal to be achieved by designers.

In the picture for Puma (Fechter), the applied voltage is about 22 deg leading the motor BEMF.

A) Why this happens?

B) Can it be configed to achieve the symmetry?
 

Attachments

  • Crystalyte Controller copy.jpg
    Crystalyte Controller copy.jpg
    36.1 KB · Views: 739
I think the diffrence is the Delta configuration of the puma. Since you have to overcome some bleed over form the other phases, you have to keep power on the coils longer. Delta wound aren't that efficent, but more power dense. since the puma spins up to higher RPM, the loss is negligable. compared to the power it produces.
 
A) Why this happens?

Most 6-step controllers (e.g. Crystalyte, Ananda and Shenzhen) will output one phase voltage which leads one Hall signal by 30 deg.

Crystalyte 4000 could be configed for the winding BEMF to have -90, -30, 30 or 90 deg with respect to one Hall signal.
It will use the 30 deg config so that the applied phase voltage is in phase with the phase BEMF.

Puma (as well as MAC-BMC) could be configed for the winding BEMF to have -80, -40, -20, 20, 40 or 80 deg with respect to one Hall signal.
It is delta-connected. So the equivalent phase BEMF becomes -120, -70, -50, -10, 10 or 50 deg with respect to one Hall signal.
The best choice is the 50 deg config, then there will be an 20 deg advance.
It is not adviceable to use the config that gives "retard timing:.
Seems that this is choice by Puma as seen in Fecther's picture.

B) Can it be configed to achieve the symmetry?
If it is re-connected into Y-connection, then the best choice is the 40 deg config and it gives an 10 deg advance.
It is still not symmetrical but seems to be a better choice.
The only drawback is have a higher battery voltage (1.73 times).

Fecther did suggest that Puma could relocate the Hall sensors. But this also means that all sensors will sit on the stator poles and the performance in forward and reverse will not be the same.

Still unable to find a good reason why Puma uses delta-connection.
 
The7 said:
B) Can it be configed to achieve the symmetry?
If it is re-connected into Y-connection, then the best choice is the 40 deg config and it gives an 10 deg advance.
It is still not symmetrical but seems to be a better choice.
The only drawback is have a higher battery voltage (1.73 times).

If it such option exits for the Puma, I will use this modification because the current to motor will be reduced to 1/1.73 ( or 0.58 ) and the motor cable will be much happier.
 
Back
Top