Forums

series input parallel output SMPS

Started by bitrex March 29, 2018
I have a bunch of small 12 volt input ATX SMPS power modules each rated 
for 120 watts. about 14 bux each, surplus. Fanless 200 watt modules are 
harder to come by and significantly more expensive.

I'd like to slave two together to get ~200 watts out, with the inputs in 
series. 24 volt 8 amp power bricks are pretty common.

I found this whitepaper on voltage balancing SMPS inputs in series but 
it seems more applicable to from-scratch designs.

<http://www.deltartp.com/dpel/dpelconferencepapers/apec11_yj.pdf>

Got any suggestions?
Are they even isolated ground?

SMPS certainly can't be stacked in that way without an extra controller.

Tim

-- 
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Contract Design
Website: https://www.seventransistorlabs.com/

"bitrex" <bitrex@de.lete.earthlink.net> wrote in message 
news:xK5vC.38628$tY3.172@fx21.iad...
>I have a bunch of small 12 volt input ATX SMPS power modules each rated for >120 watts. about 14 bux each, surplus. Fanless 200 watt modules are harder >to come by and significantly more expensive. > > I'd like to slave two together to get ~200 watts out, with the inputs in > series. 24 volt 8 amp power bricks are pretty common. > > I found this whitepaper on voltage balancing SMPS inputs in series but it > seems more applicable to from-scratch designs. > > <http://www.deltartp.com/dpel/dpelconferencepapers/apec11_yj.pdf> > > Got any suggestions?
On 03/29/2018 09:30 AM, Tim Williams wrote:
> Are they even isolated ground? > > SMPS certainly can't be stacked in that way without an extra controller. > > Tim >
Good question, if the ground is tied on both sides that's gonna be a problem. They look to be strictly buck converters on the board, not flyback.
On 03/29/2018 09:44 AM, bitrex wrote:
> On 03/29/2018 09:30 AM, Tim Williams wrote: >> Are they even isolated ground? >> >> SMPS certainly can't be stacked in that way without an extra controller. >> >> Tim >> > > Good question, if the ground is tied on both sides that's gonna be a > problem. They look to be strictly buck converters on the board, not > flyback.
Yup the continuity tester indicating that the input jack negative and ATX output common are both joined to the same ground plane.
On 03/29/2018 09:20 AM, bitrex wrote:
> I have a bunch of small 12 volt input ATX SMPS power modules each rated > for 120 watts. about 14 bux each, surplus. Fanless 200 watt modules are > harder to come by and significantly more expensive. > > I'd like to slave two together to get ~200 watts out, with the inputs in > series. 24 volt 8 amp power bricks are pretty common. > > I found this whitepaper on voltage balancing SMPS inputs in series but > it seems more applicable to from-scratch designs. > > <http://www.deltartp.com/dpel/dpelconferencepapers/apec11_yj.pdf> > > Got any suggestions?
For "historical interest" the main semiconductor components on the board are two RT8105 synchronous buck controllers, one MC64063, one AP1538 18V/3A buck controller, one mystery 8 pin SOIC the manufacturer attempted to conceal with paint but that magnified/palette invert cell phone camera picked up fine, labeled what looks like "WT751002 S 429 ZSAX G", one 4825P P-channel MOSFET in an 8 pin package and four IRFS3006 N-channels in DPAK
On Thursday, March 29, 2018 at 9:20:36 AM UTC-4, bitrex wrote:
> I have a bunch of small 12 volt input ATX SMPS power modules each rated > for 120 watts. about 14 bux each, surplus. Fanless 200 watt modules are > harder to come by and significantly more expensive. > > I'd like to slave two together to get ~200 watts out, with the inputs in > series. 24 volt 8 amp power bricks are pretty common. > > I found this whitepaper on voltage balancing SMPS inputs in series but > it seems more applicable to from-scratch designs. > > <http://www.deltartp.com/dpel/dpelconferencepapers/apec11_yj.pdf> > > Got any suggestions?
I've done that with bench supplies and Meanwells. How about just buying a bench power supply. 30V @10A is about $150 from volteq. http://www.volteq.com/switching-power-supply.html What are you doing with all that current? George H.
On Thu, 29 Mar 2018 09:20:28 -0400, bitrex
<bitrex@de.lete.earthlink.net> wrote:

>I have a bunch of small 12 volt input ATX SMPS power modules each rated >for 120 watts. about 14 bux each, surplus. Fanless 200 watt modules are >harder to come by and significantly more expensive. > >I'd like to slave two together to get ~200 watts out, with the inputs in >series. 24 volt 8 amp power bricks are pretty common. > >I found this whitepaper on voltage balancing SMPS inputs in series but >it seems more applicable to from-scratch designs. > ><http://www.deltartp.com/dpel/dpelconferencepapers/apec11_yj.pdf> > >Got any suggestions?
Switchers generally have a negative input impedance, so a series connection will be unstable. -- John Larkin Highland Technology, Inc lunatic fringe electronics
On 03/29/2018 10:30 AM, bitrex wrote:
> On 03/29/2018 09:20 AM, bitrex wrote: >> I have a bunch of small 12 volt input ATX SMPS power modules each >> rated for 120 watts. about 14 bux each, surplus. Fanless 200 watt >> modules are harder to come by and significantly more expensive. >> >> I'd like to slave two together to get ~200 watts out, with the inputs >> in series. 24 volt 8 amp power bricks are pretty common. >> >> I found this whitepaper on voltage balancing SMPS inputs in series but >> it seems more applicable to from-scratch designs. >> >> <http://www.deltartp.com/dpel/dpelconferencepapers/apec11_yj.pdf> >> >> Got any suggestions? > > For "historical interest" the main semiconductor components on the board > are two RT8105 synchronous buck controllers, one MC64063, one AP1538 > 18V/3A buck controller, one mystery 8 pin SOIC the manufacturer > attempted to conceal with paint but that magnified/palette invert cell > phone camera picked up fine, labeled what looks like "WT751002 S 429 > ZSAX G", one 4825P P-channel MOSFET in an 8 pin package and four > IRFS3006 N-channels in DPAK
Oh the "mystery IC" is probably this" <http://www.hardwaresecrets.com/datasheets/WT751002.pdf>
On 03/29/2018 10:32 AM, John Larkin wrote:
> On Thu, 29 Mar 2018 09:20:28 -0400, bitrex > <bitrex@de.lete.earthlink.net> wrote: > >> I have a bunch of small 12 volt input ATX SMPS power modules each rated >> for 120 watts. about 14 bux each, surplus. Fanless 200 watt modules are >> harder to come by and significantly more expensive. >> >> I'd like to slave two together to get ~200 watts out, with the inputs in >> series. 24 volt 8 amp power bricks are pretty common. >> >> I found this whitepaper on voltage balancing SMPS inputs in series but >> it seems more applicable to from-scratch designs. >> >> <http://www.deltartp.com/dpel/dpelconferencepapers/apec11_yj.pdf> >> >> Got any suggestions? > > Switchers generally have a negative input impedance, so a series > connection will be unstable. > >
Yeah they would need the input voltages actively balanced, but even so stacking these in series and taking the output in parallel is gonna be a problem due to the input and output sides sharing a ground.
On Thu, 29 Mar 2018 10:35:21 -0400, bitrex
<bitrex@de.lete.earthlink.net> wrote:

>On 03/29/2018 10:32 AM, John Larkin wrote: >> On Thu, 29 Mar 2018 09:20:28 -0400, bitrex >> <bitrex@de.lete.earthlink.net> wrote: >> >>> I have a bunch of small 12 volt input ATX SMPS power modules each rated >>> for 120 watts. about 14 bux each, surplus. Fanless 200 watt modules are >>> harder to come by and significantly more expensive. >>> >>> I'd like to slave two together to get ~200 watts out, with the inputs in >>> series. 24 volt 8 amp power bricks are pretty common. >>> >>> I found this whitepaper on voltage balancing SMPS inputs in series but >>> it seems more applicable to from-scratch designs. >>> >>> <http://www.deltartp.com/dpel/dpelconferencepapers/apec11_yj.pdf> >>> >>> Got any suggestions? >> >> Switchers generally have a negative input impedance, so a series >> connection will be unstable. >> >> > >Yeah they would need the input voltages actively balanced, but even so >stacking these in series and taking the output in parallel is gonna be a >problem due to the input and output sides sharing a ground.
If the outputs are in parallel and the switchers are identical isolated converters, the inputs could be force-balanced without much power dissipation, but with some complexity. There would be some midpoint input voltage that would be the perfect balance point, so would require zero balance current. That point would be unstable but could be forced. Might not be too hard, actually. Sort of interesting, dynamically. If these are not isolated converters, the ground sharing problem wrecks things. -- John Larkin Highland Technology, Inc lunatic fringe electronics