Cascaded SMPS's have fighting feedback loops

[cupoftea]

Hi,
The following is a 150W 24Vin to 12Vout DCDC SMPS

VHK150W-Q24-S12 datasheet

https://www.cui.com/product/resource/vhk150w.pdf

In the datasheet it tells of the maximum capacitance load allowable, so that's good.
Also, Such power supplies are very likely to exist "cascaded" with another different SMPS upstream, and another different SMPS downstream.
As such, there is scope for fighting between their feedback loops.
Why was the feedback loop bandwidth not stated in the datasheet so users can take any necessary steps to avoid fighting of the feedback loops?.

 

[dick_freebird]

The negative input resistance of a DC-DC is going to
freak out what's driving it, without plenty of ballast.

 

[cupoftea]

Thanks, i agree, and wonder why for this reason the datasheet doesnt give the details of the power supply necessary to assure no fighting of feedback loops.

 

[FvM]

I've never seen dynamic specifications (e.g. controller bandwidth, output impedance) with a general purpose power supply. Some specify maximal capacitive load.

Need to evaluate respective parameters and conditions for stable operations on your own.

In practice I didn't see problems unless upstream switcher is too fast.

 

[cupoftea]

Thanks, as you know, the best case is when you have like it is in PFC----> Downstream switcher.
The PFC 's bandwidth is well below the downstream switchers.
So as you know, the BW's need to be different to each other. However, with decent sizing of output and input caps, i dont think its much of an issue.

 

[cupoftea]

The negative input resistance of a DC-DC is going to
freak out what's driving it, without plenty of ballast.

Thanks, this is very true...and when Murata (or whoever) come to do a general purpose PSU for the multiple markets, they are
in a bit of bother..because how can they know how much input capacitance to add in order to not upset the upstream SMPS?
(i mean each customer may connect to it with some unknown SMPS)

If they use too little input capacitance, then they may go unstable due to feedback loop fighting, between the PSU and the upstream PSU.....but if they use too much
input filter capacitance.then it could then make the upstream PSU go unstable due to having too much capacitance attached to its output.

So there is a balance to be had here, and in truth, a certain percentage of customers will buy the PSU and find it doesnt work in their system due to the instability.

Most DCDC power supplies can have about 1mF of extra added output capacitance and not go unstable so its probably best to keep the Cin below 1mF.
Would you concur?