LTC1871 boost converter output capacitor selection

[tomshirvo]

Hi All,

Working on a boost converter with the LTC1871 to go from 15Vin(min) to 36Vout with max output current of 5A.
With all of my calculations I have come up with needing the following output capacitor value;
ESR less than 0.023ohms
Capacitance more or equal to 46uF
RMS ripple current of 5.9A

I am finding it really hard to find capacitors that will fit into this range. I understand that I can use multiple capacitors in parallel and a combination of ceramic and electrolytic. Ceramic capacitors don't seem to have a ripple current rating on them and electrolytic caps seem to only have large ripple current with high capacitance.

Is 5.9A ripple current a lot?
How do I satisfy such a large ripple current?

It also seems that capacitor ripple current ratings only go upto 100kHz, I am switching at 300kHz. Does this mean everything above 100kHz is the same value?


Other details of design;
300kHz switching
4.8uH inductor

This boost converter will be powering brushed DC motors using an H-bridge.
Power source will be milwaukee power tool batteries.


Thanks in advance!

[nano]

5.9A RMS appears to be a lot. Where do you get that value from?

To calculate capacitor RMS current first calculate the diode RMS current (see attached picture).
Then substract the output current (remember we have RMS values):
Icaprms = sqrt(Idiodrms^2 - Iout^2)

[tomshirvo]

Hi nano,

The datasheet shows me how to calculate the ripple current on page 17 or in the image added below taken from page 17. http://cds.linear.com/docs/en/datasheet/1871fe.pdf

so something like this;

Io(max) = 5A, Vo = 36v, Vin(min) = 15v

5.9A = 5A*SQRT((36-15)/15)


I found the following capacitor, http://au.element14.com/search?st=EEU-EE2C101

MPN: EEUEE2C101
http://www.farnell.com/datasheets/2078714.pdf?_ga=2.59243502.2138889683.1496621153-1216399895.1480304478

They are 100uF 160v with a ripple current of 1660 mA@100kHz

Firstly they don't say what the ESR is so thats really helpful. What does that mean??

Secondly I would need 4 of them to have a minimum of 5.9A ripple current ability.

From my calculations showing that I only need 47uF output capacitance, is there any problem with having 400uF?

[jbb]

5.9A ripple is within reason. It'll depend on the duty cycle and inductor current ripple.

Moving to an interleaved design would reduce the ripple quite a bit, but needs more parts.

There are 4 capacitor options:
1) electrolytic. Has ageing and operating temperature concerns.
2) many ceramics in parallel.  Beware the fact that capacitanc drops a lot with DC voltage.
3) film capacitors can be excellent but are normally through hole
4) tantalum caps can be very effective but need to be de-rated in voltage so they don't catch fire or worse.

Be aware that current can flow from the battery, through the inductor, through the upper diode and into a short-circuited output. The control IC can't stop this, so protection (maybe a fuse) should perhaps be deployed.

[tomshirvo]

Thanks jbb!

I have got a fuse on the battery size of the boost converter. I also have "Smart High-Side Power Switch" (BTS723GW) on the output side of the boost converter that then goes to the motor H-bridge IC. These have short-circuit protection and over current protection features.

[nano]

The math in the datasheet checks out: 6A cap. RMS value is right.
Most electrolytics can handle higher rms current without problem. Just make sure to derate them. Details can be found here (page 9):
https://www.vishay.com/docs/28356/alucapsintroduction.pdf

From my calculations showing that I only need 47uF output capacitance, is there any problem with having 400uF?

More capacitance is no problem as long as you make sure you regulation loop is still ok.

[David Hess]

This is a common problem with aluminum electrolytic capacitors.  The output capacitance and voltage rating has to be increased to achieved the desired ripple current.  Using many small capacitors in parallel in place of a single capacitor also helps to increase the ripple current rating because of increased surface area to volume.  Polymer aluminum electrolytic and polymer tantalum capacitors would be suitable but derating will still require multiple parts in parallel.

At such low values of ESR, the printed circuit layout becomes important to share current between the capacitors.

The ESR dominates the frequency compensation so increasing the value of capacitance to achieve the required ripple current rating should not cause problems with frequency compensation; this is especially the case with aluminum electrolytic and solid tantalum capacitors.  Having too low an ESR like with film or ceramic capacitors and sometimes polymer capacitors is more likely to cause problems but this should be taken into account when implementing the frequency compensation.

So using aluminum electrolytic or solid tantalum capacitors in parallel to achieve the required ripple current rating and ESR yields a higher capacitance than required but this is not a problem.  Using polymer aluminum electrolytic or polymer solid tantalum capacitors in parallel to achieve the required capacitance will likely yield a lower ESR than required complicating the frequency compensation but this can be handled.

[jbb]

I have got a fuse on the battery size of the boost converter.

Perfect.  I suggest you make it's current rating quite generous to a) keep resistive losses low and b) reduce the number of nuisance trips (there will be a current pulse every time you plug the battery in).  Also make sure you use a good fuse-holder.  I hear that some of them add quite a lot of resistance.