What is the PWM frequency of the Allegro A4989

[Lord_grezington]

Hello all

I have been looking at a new project where I am hoping to use the Allegro A4989 http://www.allegromicro.com/~/media/Files/Datasheets/A4989-Datasheet.ashx, it seems like a very simple to use chip, however I am having a little issue.

What is the PWM frequency? It says that it uses a fixes off time, however the on time is hard to calculate as it completely depends on the inductance of the motor phase. The motor I intend to use has a phase current of 8A and an inductance of 12mH (running from a 48V supply). Using the inductance formula I get an on time of 2ms. However this will only happen once per step, then the off time comes into play when the current drops out the coil. From the data sheet I get 1.5us blank time, 21.75us fixed off time and 750ns Dead time. (I am using the 4Mhz clock frequency). This gives me a total of 22.5us total off period. Now if the voltage is 0 as PWM is off, what is the current level in the motor coil after 21.75us??? I need to know this so that I can use di to find out how long its going to take to reach the next trip level. The formula I used is the standard V = L di/dt.

I need to work out what the PWM frequency is to firstly calculate the Bootstrap capacitor value, and secondly calculate the Creg value.

Thank you for all your help in advance

 

[Lord_grezington]

Anyone managed to have any luck with the above question?
Let me know if you need more information and I will see what I can do to get it.

 

[john monks]

From what I gather from the datasheet your PWM is far less than 14kHz and therefore the bootstrap capacitor should be 20 times the total mosfet gate capacitance and cREG should be 40 times the value of the bootstrap capacitor. From what I gather these are minimum values so they can be much greater.

 

[Lord_grezington]

Hi John

Thanks for your reply

Where on the data sheet does it say that the PWM frequency is far less than 14Khz?

Thanks

 

[john monks]

It doesn't. Page 12. says "Fixed Off-Time The internal PWM current control
circuitry uses the master oscillator to control the length of
time the power MOSFETs remain off. The off-time, tOFF ,
is nominally 87 cycles of the master oscillator (21.75 μs at
4 MHz)" And the period of 14kHz is about 7.14us. Therefore your PWM must be much lower than 14kHz.

 

[Lord_grezington]

Hi John

Is 14Khz not 71.4us?

This means that my On time need to exceed 49.65us to get the frequency below 14Khz. I am not sure how to work this on time out as it is dependant on the inductance in the motor coils.

 

[Lord_grezington]

Will this not depend on the decay as well? A higher decay will result on a higher frequency.

 

[Lord_grezington]

Thinking about it, surely if I assume that the rate of change of current going into the motor is the same as the rate of change coming out? (as long as the decay is at the correct level).

This would create a 50% duty cycle, so off time of 21.75us, on time of 21.75us and then the blank time (1.5us) and the dead time (0.750us) gives a total period of 45.75us. So a total frequency of 21.8Khz. This sounds right as at least this is above the audible frequency range so we don't get the ringing.

Does this make sense?

 

[john monks]

I accept your correction. The period of 14kHz is 71.4us. I'm having trouble PWM from the datasheet. But you can still figure out what the bootstrap capacitor should be. And in any case, unless I am missing something, it seems to me the the value of these capacitors are minimum values. So why not simply use 0.1uF capacitors and be done with it?

 

[Lord_grezington]

Hi John

I have done so, total gate charge is 10nF, so I have used bootstraps of 0.33uF and a Creg of 15uF.

This also works with 22Khz

Thanks for your help