Laptop brick running off DC 300V?

[elektrinis]

I want to ask if anyone has investigated this - can any laptop brick, and any modern PSU for that, run off DC 300V source instead of AC 230V? Since most of modern PSUs have an initial PFC stage, which is just a rectifier with boost converter, it kinda should? But is there any kind of protection, checking for AC? Anyone know of a "DC-ready" study, where they check how many modern devices will actually run off DC?

[f4eru]

Most will, but there could be some road block :
- a rare X cap active discharge circuit could make problems
- PFCs could make problems.

Also, note that this is not a working mode for which it was safety tested, so, it's at your own risk.

[NiHaoMike]

That has been looked at for data centers, it turns out most devices with switching power supplies just work.

[station240]

I want to ask if anyone has investigated this - can any laptop brick, and any modern PSU for that, run off DC 300V source instead of AC 230V? Since most of modern PSUs have an initial PFC stage, which is just a rectifier with boost converter, it kinda should? But is there any kind of protection, checking for AC? Anyone know of a "DC-ready" study, where they check how many modern devices will actually run off DC?

I've done this experiment with a small 100W PC power supply that has PFC.
Only changes I made (mostly for efficiency) were:
1) Remove the bridge rectifier (-1.2V from 240VAC = 0.5% losses).
2) Remove the AC input common mode chokes (clearly designed for AC mains)
3) Remove the X2 filter capacitor and parallel discharge resistor (again designed for AC frequencies)
4) Add a fat 400V capacitor to the input. (for noise/peak current reasons)
Worked fine on as little as 60V DC.

Most PFC controller chips show internal block diagrams, it's just dumb logic gates and opamps.
To actually check for AC would require some form of memory, to calculate changes in input voltage over time.
Plus that would leave PSUs to shutdown in areas with poor/noisy mains, eg industrial, rural.

I've got a few spare laptop PSUs lying around, newest is probably 5 years old. Not sure if any are PFC without cracking them open.

[David Hess]

The ones I have tested all just worked, but I have read that some active PFC input stages will not correctly operate with a DC input.

[guenthert]

[..]
2) Remove the AC input common mode chokes (clearly designed for AC mains)
3) Remove the X2 filter capacitor and parallel discharge resistor (again designed for AC frequencies)
[..]

     Uh, aren't those filters meant to reduce RFI?  I suppose in a data center it won't matter, but in an electronics lab you won't make friends if you remove those from a SMPSU.

[mikeselectricstuff]

The filtering components will work just as well on DC. Safety-wise the only potential issue is the internal fuse, which may not be rated for DC operation, and could arc over in a fault condition.

[f4eru]

     Uh, aren't those filters meant to reduce RFI?  I suppose in a data center it won't matter, but in an electronics lab you won't make friends if you remove those from a SMPSU.

EMC is important everywhere, not only in a lab.

[SeanB]

Yes most sub 100W are not going to have PFC, so will run fine on DC. Just use an external DC rated fuse on the power lead, to provide failure protection for the cable. For most 100W power supplies a 3.15A DC rated ceramic fuse will work, not going to fail on power on, but will at least open when the semiconductors go short circuit. Even with PFC most of the time the PFC will just turn into a boost converter to make an internal 400VDC rail out of the 300V, or will simply not start up as the output is above the switch on threshold. All PFC senses output voltage along with input, so one or the other loop will do regulation. you can even leave the input bridge in place, as likely losses in it will be low with no current pulses being impressed on it. EMC capacitors will survive fine as well, just act as extra filtering for noise coming in or out. Would suggest having a fire resistant casing for the units though, just in case of some failure and the external fuse not going fast enough.

[wraper]

2) Remove the AC input common mode chokes (clearly designed for AC mains)
3) Remove the X2 filter capacitor and parallel discharge resistor (again designed for AC frequencies)

Clearly designed to block high frequency noise. And in case of PFC, X2 capacitor may also act as input smoothing capacitor. You've done unnecessary and counterproductive thing.

[ocset]

What is the name of the PFC controller on there?
But anyway, it should  work fine....it will just think that its an extremely low frequency mains!!
No PFC controller  "checks for AC"...unless its controlled by a custom micro that does  check for this.

I guess you could increase the speed of the feedback loop of the PFC to make it work a bit better....then again, no point if you dont have highly transient loads.

[ejeffrey]

I want to ask if anyone has investigated this - can any laptop brick, and any modern PSU for that, run off DC 300V source instead of AC 230V? Since most of modern PSUs have an initial PFC stage, which is just a rectifier with boost converter, it kinda should? But is there any kind of protection, checking for AC? Anyone know of a "DC-ready" study, where they check how many modern devices will actually run off DC?

I've done this experiment with a small 100W PC power supply that has PFC.
Only changes I made (mostly for efficiency) were:
1) Remove the bridge rectifier (-1.2V from 240VAC = 0.5% losses).
2) Remove the AC input common mode chokes (clearly designed for AC mains)
3) Remove the X2 filter capacitor and parallel discharge resistor (again designed for AC frequencies)
4) Add a fat 400V capacitor to the input. (for noise/peak current reasons)
Worked fine on as little as 60V DC.

Note that when you unplug this the input jack will be live with 300V and a big capacitor to back it up.  Leave the diode in unless you are hard wiring it or have a connector with no exposed metal.

[David Hess]

Note that when you unplug this the input jack will be live with 300V and a big capacitor to back it up.  Leave the diode in unless you are hard wiring it or have a connector with no exposed metal.

And the PFC circuit already has a suitably sized input capacitor which will work fine in a DC application.