LM5069: calculation of correct amount of capacitance on the Live Backplane

Hello Rakesh,

thx for your quick reply. My customer came back w/ some follow on questions/remarks:

Customer is wondering why TI would put the statement regarding “adding the backplane capacitance” in the data sheet if power supply capacitance is usually enough?

In their application, they have one 24VDC (i.e. 24V @ 5A) source providing power for up to twelve module bases. The bases, when connected in series, form one long Copper bus that allows distribution of the 24V power. Their modules are then plugged into these bases.

They do implement a 33V TVS component across the input to the LM5069.


Thank You,

Marcus

Hi Marcus,

TVS right at the input of LM5069 is the key requirement for transient protection, any additional capacitance on the live plane would help reduce the stress on the TVS diode.

It is a generic statement made in the datasheet -> as the amount of capacitance depends on the path impedance and the load current the hot swap circuit shuts off. As a general thumb rule (1uF/1W), I would recommend 120uF for this 120W source.

Best Regards,
Rakesh

Hello Rakesh,

again thanks for your reply. I got one more add on question if there is a formula for capacitance that includes path impedance and hot-swap load current? (with reference to your general rule of thumb (1u/1W).)

Thanks,
Marcus

Hi Marcus,

Whenever the hot swap circuit shuts off the load current, all the energy stored in the path inductance (parasitics) should be absorbed by the capacitor. Using the energy balance equation 1/2L*i*i = 1/2C*v*v and assuming 100nH at the source side of live backplane, 100A of interruption current by the hot-swap.
The amount of capacitance needed to restrict voltage rise to 10V => C=L*i*i/(v*v) = 100nH*100*100/(10*10) = 10uF.

Remember, this is on the live backplane. The Hot-swap still needs TVS right at its input.

Best Regards,
Rakesh

Hi Marcus,

Does this clarify you? Any followup questions from the customer?

Best Regards,
Rakesh