DRV8306: Selecting the ideal decoupling capacitor

Hello Tejas,

Thanks for your patience, our experts will be getting back to you tomorrow!

Thanks,

Matt

Hi Aaron,

thanks for such a detailed response. Just a follow-up question: L is the phase to phase resistance, right?

I mean phase to phase inductance. Also, another question, if we take our max current to be 5A, motor inductance is 480uH and the allowed transient is,aboutt 100mV we get a required capacitance value of about 3000uF. That is way above 220uF that is recommended. Did I misunderstand something?

To arrive at this formula stated above It seems you equated the energy stored in the cap to the energy stored in the motor coils, right? can you explain your assumptions behind this? Just so that I can better understand its limitations.

Hi Tejas,

Yes, that is how we arrived at the formula, assuming that all energy dissipated from the motor comes from the stored energy in the bulk capacitor and we are disregarding energy losses such as thermal and heat. But since this is an estimation based on a more ideal system, we would recommended testing to ensure that your value is achieving the 100mV transient.

It is more common for BLDC systems to see 1.2x to 1.5x the nominal DC voltage due to inductive spiking from the motor inductance, and in order to achieve small transients at the supply, more bypass capacitance is required. You can also reduce inductive spiking in the powerstage by reducing gate drive, optimizing the board layout to have small current paths and wide traces, or adding bypass capacitors and/or RC snubbers.  

If you are in need of saving space in terms of height (since you mentioned that the 220uF capacitor is the tallest), you could try adding smaller capacitors in parallel. 

Have you measured what transient the 220uF bypass capacitor gives you at the supply?

Hi Aaron,

Aaron Barrera said:

Yes, that is how we arrived at the formula, assuming that all energy dissipated from the motor comes from the stored energy in the bulk capacitor and we are disregarding energy losses such as thermal and heat. But since this is an estimation based on a more ideal system, we would recommended testing to ensure that your value is achieving the 100mV transient.

Understood. 

Aaron Barrera said:

If you are in need of saving space in terms of height (since you mentioned that the 220uF capacitor is the tallest), you could try adding smaller capacitors in parallel.

I thought of that. But we are also constrained by cost. Our target cost is very aggressive so I was trying to find the best solution. But if nothing works I am planning to use this option. 

Aaron Barrera said:

Have you measured what transient the 220uF bypass capacitor gives you at the supply?

Yes, we have but due to some limitations, we were only able to take readings at no load. Where the motor only consumes about 120-140mA. And here with the 220uF as bulk cap, we got voltage variations of about 0.2V at the supply. With a cap of 10uF (tested to see the limits) we got a voltage variation of about 1V.

Hi Tejas,

I will get you a reply tomorrow by the end of the day - investigating with the team internally.

Hi Aaron,

Sorry I made a mistake. We measure 0.02V or 20mV with the 220uF at no load.

Sounds good thanks.

Hi Tejas,

Could you clarify about the transient you measured? Now you are 20mV (not 200mV) and the customer's original specification is to be within 100mV, correct? 

If so, can you experiment and use smaller capacitor values and see how that changes your transients at the supply? This will be a trade-off between your bulk capacitance size and transient allowed at the supply.

Thanks,
Aaron

Yes but that was at no-load speed. At load, it will be higher. But As you said I'll have to do trial and error and find out the optimal value. Thanks for your help.

Ok got it. Will take a look. Thanks.

Hi Tejas,

I'm going to close this thread for now. Please ask a related question if you have any other questions for us!

Thanks,

Matt