HELP ! DC link capacitor sizing

[Electro nS]

for H bridge motor controller
how do u design your DC link capacitors :
according to this article :

https://www.google.com/?gws_rd=ssl#q=is1.3.2.pdf

use maximum allowble voltage ripple to find C , and then the current ripple from L and f and Vbus , then we get capacitor current rating.

if i apply those values i come up with non sense numbers :

The controller maximum current is 50A , 25A continous , motor inductance is 100uH , voltage is 25v f =16khz .
Current ripple= 0.25*vbus/f*L=0.25*25/16k*100u= 3.9 Apeak.. = 1.4 Arms !!! very low number !!

and using allowble voltage ripple of 1% , capacitance is about : C=Vbus/(32*L*deltaV*f*f)= 120uF
is this type of calculation valid ??
i usually find >>1000uF in similar controllers operated from battery offcourse.

 

[schmitt trigger]

"and using allowble voltage ripple of 1% , capacitance is about : C=Vbus/(32*L*deltaV*f*f)= 120uF
is this type of calculation valid ?? "

Simple answer: no

For most capacitors in these applications, the limiting factor is the self-heating due to the ESR * I ripple^2.

Capacitor manufacturer's provide maximum ripple requirements.
Even then, good engineering practice dictates that one significantly de-rates this maximum level. This to have a long service life, specially if the drive is to be used on harsh environments.

Additional consideration for battery-power applications may dictate that one increases this value even further. Why? Because batteries can become expensive replacements, and again self heating due to ripple is a life-reducing contributor.

 

[Electro nS]

"and using allowble voltage ripple of 1% , capacitance is about : C=Vbus/(32*L*deltaV*f*f)= 120uF
is this type of calculation valid ?? "

Simple answer: no

For most capacitors in these applications, the limiting factor is the self-heating due to the ESR * I ripple^2.

Capacitor manufacturer's provide maximum ripple requirements.
Even then, good engineering practice dictates that one significantly de-rates this maximum level. This to have a long service life, specially if the drive is to be used on harsh environments.

Additional consideration for battery-power applications may dictate that one increases this value even further. Why? Because batteries can become expensive replacements, and again self heating due to ripple is a life-reducing contributor.

what about ripple current , how to know how to get the ripple current of my circuit ? is the formula correct ??

 

[mtwieg]

You haven't specified what kind of motor you're driving, or what its ratings are. Also any DC-DC converter feeding the DC link will also contribute to ripple current.

 

[D.A.(Tony)Stewart]

DC Link Caps supply the impulse currents created by the step load voltage.
The max Current is a result of the total Effective Series Resistance (ESR) including Cap, Bridge and load coil resistance.

Thus the %ripple on the DC link is just the ratio of the Cap. ESR to the total ESR for switching of V+.
Normally for low supply ripple of 1%, the Cap ESR =1% of the load ESR.
This assumes the ESR of the regulated supply is also <1% given by the step load response as this is part of the loop.

Considering most motors have a stall (Max) current = 8x RMS current.

If the controller can only handle 2x the continuous Amps, then a series resistance or a soft start by design ought to be considered.

Batteries can be modelled by a very large capacitance, with ESR, and some ESL

 

[Electro nS]

You haven't specified what kind of motor you're driving, or what its ratings are. Also any DC-DC converter feeding the DC link will also contribute to ripple current.

pmdc motor , 400w rating , i have specifed that the power is supplied from batteries in the last sentence of my post ..

still no commented on the formula that is used to detect the ripple current from the switching of the motor /// "Current ripple= 0.25*vbus/f*L " is it right ??