You are here

Why is the speed lower in FOC mode than BLDC mode ?

9 posts / 0 new
Last post
stone_1706
stone_1706's picture
Offline
Last seen: 3 years 2 months ago
Joined: 2019-01-17 05:37
Posts: 11
Why is the speed lower in FOC mode than BLDC mode ?

Hi,

 I am a new guy , I just get a VESC4.0 board and try to test it .

what I found is  that speed  is different  using FOC mode and  BLDC mode .

BLDC motor :  400KV,  Power:25V.   MAX no-load speed should be 10000rpm @25V. (400 * 25 = 10000 rpm)

In BLDC mode  I got  9300rpm( it's not bad), but in FOC mode it just ony 8500 rpm,too slow ...

I think two modes(FOC,BLDC) could be different a little bit.  But  9300rpm : 8500rpm ,it's too slow to use FOC mode .

I tried to adjust  "FLUX_LINKAGE" value to smaller value , but the speed is not  speed up enough . if this value too small ,motor run failed . 

What I supposed to do  to get speed as same as possible between BLDC and FOC mode ?

Best regards

Stone.

 

 

stone_1706
stone_1706's picture
Offline
Last seen: 3 years 2 months ago
Joined: 2019-01-17 05:37
Posts: 11

When I change the PWM Freq from 30KHz to 40KHz , the max speed is different too. Freq is higher ,the speed also goes higher .

  If you want to change the speed ,you should modify the duty setting,not PWM freq ,right ? But why the speed are different  in different PWM freq setting? 

what the PWM Freq I should to use ?  20k? 30k? 40k?

Best regards

Stone.

 

 

 

sam.vanratt
Offline
Last seen: 3 years 1 month ago
Joined: 2018-11-18 14:09
Posts: 83

Hello Stone

in BC mode (called BLDC here) you could use a 100% Duty Cycle. In FOC the VESC4 needs a windows (DC max. 95%) to look into the current and flow. This means it's a bit slower then. The VESC6 allows 1000% DC, so BC and FOC should be the same. My big 12kg MagicPie Motor faces the same feature with a VESC4.12. (with former ESC in BC mode ~200RPM, now in FOC about 185RPM)

When you set up a higher switching cycle you get nearer the real 95% value as more cycles falls into the windows. Higher switching means higher losses when charging/discharging the FETs so 20kHz is standard. If you cool your VESC or the FETs I'm sure you can go higher and use the 40kHz, but the energy loss is still the price. Setting the input voltage higher would surely get an easier result.

Cheers

Sam

Pedelec usage in combination with a GoldenMotor MagicPie3 BLDC

stone_1706
stone_1706's picture
Offline
Last seen: 3 years 2 months ago
Joined: 2019-01-17 05:37
Posts: 11

Thanks for your reply.

I understood it,  right now PWM freq is 40kHz in FOC mode.

Thanks again.

Stone.

 

 

 

 

stone_1706
stone_1706's picture
Offline
Last seen: 3 years 2 months ago
Joined: 2019-01-17 05:37
Posts: 11

Hi Sam,

One more thing, VESC 4.0 use 2 shunt 0.001mÖ, VESC 6 has 3 shunt .

What is the big different between 2 and 3 shunts ?

Do we have to use 3 shunts to get best performance ?

Best regards

Stone.

linsus
Offline
Last seen: 2 days 20 hours ago
VESC Free
Joined: 2017-10-25 09:21
Posts: 15

hi,

with two shunts the 3d phase current is calculated as an approximation of the 1st and 2nd phase current. 
In the three shunt design you get a real measured value of what current accutally exsist in all phases at the same time. The measured value has higher accuracy then the approximation. 
Depending on the application, you could see an increase in performance on the three shunt design.

sam.vanratt
Offline
Last seen: 3 years 1 month ago
Joined: 2018-11-18 14:09
Posts: 83

Hi Stone

as mentioned it the three shunt version give a max. of 100% duty cycle (so higher eRPM) and everything is measured not calculated (measurement window remember?).

Cheers

Sam

Pedelec usage in combination with a GoldenMotor MagicPie3 BLDC

stone_1706
stone_1706's picture
Offline
Last seen: 3 years 2 months ago
Joined: 2019-01-17 05:37
Posts: 11

Thanks a lot !!!   got it .

:-) 

Stone.

speed3r
Offline
Last seen: 2 years 9 months ago
Joined: 2018-11-15 15:55
Posts: 11

In the three shunt design you get a real measured value of what current accutally exsist in all phases at the same time.

I'd like to use this quote to point out that in Vesc-Tool via sample the third current stil gets calculated. So if anybody uses this feature to locate faults one should be aware of this. Took me some minutes to understand that it is not my CSA that is defective.