PIC24: Advice on PWM use for BLDC and boost-converter

Thread Starter

Robin66

Joined Jan 5, 2016
266
I have a PIC18F2620 circuit (v2) which drives a 3-phase BLDC motor and incorporates a boost-converter. At low speeds the motor drive is PWMed, and as speed increases there's a transition point where the PWM duty cycle = 100% and the boost-converter turns on. I did all of this manually using 16bit CCP modules and interrupts.

I'm now designing v3 and am looking at using a PIC24. I see there are chips which have 3-phase BLDC drive in mind and so have 6 ch PWM. However for v2 I found driving the motor to be the easy part; driving the boost converter burnt far more instructions. This makes me think I should drive the motor as before and so I'd only need 2 PWM modules (for the L and H sides of the boost converter). I don't see any PIC24 chips with 8 PWMs.

Am I missing a trick ignoring PICs dedicated motor control offerings? Or is there a better way to accommodate a boost converter and motor control using a PIC24? Any advice greatly appreciated.
 

Papabravo

Joined Feb 24, 2006
15,085
I'm going out on a limb here, but if the boost converter is resource intensive and you might get a better bang for the buck using the integrated motor control, I would look at using a dedicated controller chip for the boost converter. It seems like there might be an advantageous compromise in there somewhere. As in: "With all the HS around, there must be a pony in there someplace".

What are the specifications on the Boost converter?
 

Thread Starter

Robin66

Joined Jan 5, 2016
266
Yes that's an idea I thought about briefly. The ON time is ~5-10us and the free-wheeling time can be anywhere in the range 10-100 us. So at max speed the interrupt would need to service the boost converter every ~10us, cf. motor period of 500us at max speed.
 

Papabravo

Joined Feb 24, 2006
15,085
Yes that's an idea I thought about briefly. The ON time is ~5-10us and the free-wheeling time can be anywhere in the range 10-100 us. So at max speed the interrupt would need to service the boost converter every ~10us, cf. motor period of 500us at max speed.
There are at least three companies that are pushing the bleeding edge of technology in DC-DC converters:
  1. TI -- Texas Instruments
  2. LTI -- Linear Technology
  3. Maxim
I think you would be well advised to survey some available parts and maybe even sketch out a paper design; just for grins. You might be impressed with some of the new features like very low quiescent current consumption and soft startup. This might give you a considerable break on average current consumption, if the DSP doesn't have to flat out all the time.

I'd still like to understand the boost converter specs in terms of voltage, current, and power in and out.
 

Thread Starter

Robin66

Joined Jan 5, 2016
266
tbh I've been glossing over quiescent current considerations because the motor is 250W max rated so I'm only worried about wastage in the driving MOSFETs. If the PIC dissipates 0.1W doesn't concern me.

The boost convert will have the following specs:
  • 20V input
  • Up to 40V output
  • Up to 400W
So at full power the input will be sourcing avg 20A and the output will be sinking avg 10A@40V
 

Papabravo

Joined Feb 24, 2006
15,085
tbh I've been glossing over quiescent current considerations because the motor is 250W max rated so I'm only worried about wastage in the driving MOSFETs. If the PIC dissipates 0.1W doesn't concern me.

The boost convert will have the following specs:
  • 20V input
  • Up to 40V output
  • Up to 400W
So at full power the input will be sourcing avg 20A and the output will be sinking avg 10A@40V
That power level is a bit outside my comfort range. Do it any way you want that works and is safe.
 
Top