Arduino controlled BLDC motor sensing

Thread Starter

JulesP

Joined Dec 7, 2018
383
Hi all,

This is a project thread that, due to its evolving nature, has continued on from 'Using a transformer in reverse' in 'General Electronics Chat'.

The next stage of this Arduino controlled motor project involves directing the 400V DC (or other more manageable voltage) to the sensorless BLDC as directed by the Arduino based controller and using the back EMF for commutation.

What has been observed by a colleague, who is using a 36 pole motor, is that the back EMF is laced with high voltage induction spikes that reset the Arduino and stop the motor.

One option is to convert the whole system to using Hall sensors but that is a major undertaking. A simpler option I imagine would be to filter the spikes out and wondered if suitable capacitors (say 10uF 600V) across each of the windings would do that effectively?

The interface and sensing feedback circuit, that sits between the Arduino and the motor, is shown below and includes my cap suggestion (in purple).

I would value any thoughts on whether this will work.

Thank you,

Jules

Motor Controller Circuit (with mods).jpeg
 

Thread Starter

JulesP

Joined Dec 7, 2018
383
What say 1uF, 600V? Does the peak transient voltage and rise time make much difference to the filtering?
 

Thread Starter

JulesP

Joined Dec 7, 2018
383
What is the nature of a BLDC motor that has 36 poles?
Max.
It’s a Samsung washing machine motor as per this link: https://www.espares.co.uk/product/es1578438/washing-machine- motor?pageNumber=2&PartTypeId=1752&ManufacturerId=596
This has three sets of 12 coils in parallel but apart from that I don’t know much more. I assume it is rated for 230V. The idea being explored is that pulsing with 400V DC will improve the efficiency. Regarding the pulse frequency that is also being explored but probably in the range 10-50kHz.
 

dendad

Joined Feb 20, 2016
4,451
Are you sure the motor does not have Hall sensors?
the F&P one has.
Hall sensors are pretty cheap and I would add some if they are not fitted.
 

Thread Starter

JulesP

Joined Dec 7, 2018
383
Are you sure the motor does not have Hall sensors?
the F&P one has.
Hall sensors are pretty cheap and I would add some if they are not fitted.
I don’t know at this point but if it doesn’t then fitting them can’t be straightforward since they would need to be embedded into the stator somehow. Even if it does I think one can chose not to connect them and use back emf sensing instead.
 

Thread Starter

JulesP

Joined Dec 7, 2018
383

MaxHeadRoom

Joined Jul 18, 2013
28,617
I suspect 4 pole motor if a direct copy of the F.P. as they typically run up to 1500-1700rpm.
The PDF's on the motors state RPF, rotor position feedback Optional, normally sensorless vector.
Max.
 

Thread Starter

JulesP

Joined Dec 7, 2018
383
I suspect 4 pole motor if a direct copy of the F.P. as they typically run up to 1500-1700rpm.
The PDF's on the motors state RPF, rotor position feedback Optional, normally sensorless vector.
Max.
Ok so normally back EMF sensing but Hall sensors installed. If 4 pole at any one moment how might they be energised? 2+ and 2- or perhaps 1+, 1- and two floating?
 

MaxHeadRoom

Joined Jul 18, 2013
28,617
The commutation is only two sets of the three winding's are energized at any one time.
This shows the commutation when sensors are used, the motor is back driven to produce the 3 sine waves.
Max.
 

Attachments

Thread Starter

JulesP

Joined Dec 7, 2018
383
The commutation is only two sets of the three winding's are energized at any one time.
This shows the commutation when sensors are used, the motor is back driven to produce the 3 sine waves.
Max.
Ok, so it’s not unreasonable for the motor shown on the link I sent to be set up as three sets of 12 windings with two of them at any one moment being energised (36 windings total). That is what others working on this motor have shared with me. If that is the case then would it be referred to as a 12 pole motor?
 

Thread Starter

JulesP

Joined Dec 7, 2018
383
To start testing my circuit I am planning to begin with a small 12 pole BLDC motor, such as the one in the pic, and slowly ramp up the voltage to its max and observe any transient spikes - also this will let me check the Arduino sketch. This particular motor has Hall sensors but I think I can just ignore those and use the back EMF from the three windings. Am I right in thinking that Phase 1 /2/3 in my circuit corresponds to the White/Red/Black on the motor? Apparently it is damaged if they are reversed.
 

Attachments

shortbus

Joined Sep 30, 2009
10,045
I don't want to rain on your parade, but do you think a Arduino has enough computing power to handle what your asking it to do? If they do have the capability why are most sensorless BLDC controllers DSP based?

Some reading for you, pick one or more; https://www.google.com/search?clien...8......0....1..gws-wiz.......0i71.bomppDA7IBU

Am I right in thinking that Phase 1 /2/3 in my circuit corresponds to the White/Red/Black on the motor? Apparently it is damaged if they are reversed.
Isn't reversing the sequence of the stator coils how a BLDC rotation is reversed?
 

Thread Starter

JulesP

Joined Dec 7, 2018
383
I think the Arduino is a flexible system that will work with sensorless but there may well be newer systems that are better in some regards. I will read up. The small group I am working with are all using Arduinos so I will focus on that for the time being. Bigger issues are the spikes in the back emf sensing feedback.

Yes but this particular motor it seems is designed to rotate one way. It’s not a stepper motor that would need to go back and forth.
 
Top