DCDC module input capacitor is getting over-ripple-currented?

Dear Engineers,
A contractor has done the attached prototype PCB for us using the JTK3048S05 DCDC module from XP Power.com.
(spec is 48vin, 5V, 5A output.

Would you agree that at the input, the 2u2 and 82u caps should swap places with each other?

(it seems difficult to say, because there’s no way of knowing if there is already capacitance inside the DCDC module? If there’s no internal caps, then if it’s a BCM buck in the module, then the 2u2 cap will be heavily over-ripple-currented.)

-----------------------------------*****************---------------------------------------------------------
Datasheets...

MSS1038 (12uH) datasheet

A759KK225M2EAAE685 cap (2u2) datasheet
**broken link removed**

A759SZ826M2EAAE150 cap (82u) datasheet
**broken link removed**

JTK3048S05 DCDC module datasheet 48VIN, 5V, 6A output

I don't think that there's enough information from the data sheets to know what the ripple will be. You could measure it and find out. You just need a small resistance, say 100 mOhm, in series with the 2.2 μF capacitor and you use an oscilloscope to measure the voltage across that resistor when it's running.

I would tend to agree that there could to be too much ripple in the 2.2 μF capacitor. However, it's an isolated converter, which implies that it has a transformer inside it, so the peak input current is likely to be far less than the output current. Also the specifications quote quite low values of ripple with some filtering, and the EMI filter circuit doesn't show any input capacitor after the inductor, and quite modest capacitance values before the inductor.

The capacitors suggested in the module's data sheet are ceramic. They have lower ESRs than electrolytic capacitors, so they are more effective at reducing ripple and they don't get hot if there is a lot of ripple current, so they could well be a much more suitable.

and the EMI filter circuit doesn't show any input capacitor after the inductor

Click to expand...

...Yes thankyou...i noticed that......it seemed odd to me too...but as you imply, i guess there must be some internal input capacitance....its just a case of how much?

zenerbjt said:

...Yes thankyou...i noticed that......it seemed odd to me too...but as you imply, i guess there must be some internal input capacitance....its just a case of how much?

Click to expand...

Well the frequency, current waveform and internal capacitance are all unknown, so you original question isn't going to get a definitive answer.

also watch out for the ringing on the input voltage (at the input to the module). The inductor you are using has very little series resistance so it will form a tuned circuit with the capacitance and possibly give you large ringing spikes at the VIN pin of the module. Putting a resistor in parallel with the inductance will add some damping and remove this.... or use a ferrite bead that has a higher series resistance and much more damping