SMPS output power upgrades and limits

[Dannyx]

Good day folks. Having messed with power supplies a lot, getting shocked and whatnot and constantly trying to improve my knowledge of them, I was wondering: what exactly gives the output power figure of a SMPS, let's say a wall-wart or power pack for a laptop for instance. Every power supply out there has a wattage stated on it in W, so what exactly gives this upper threshold ? There's also a current rating stated, so basically the power is the rated voltage, times the maximum current, so since we can't alter the voltage, we want to increase the current capacity.
We know SMPSs contain several blocks, so is it the switching block (the FETs), or the rectifier block that's the bottleneck? What should be upgraded with components of higher specs to increase the power of a SMPS or at least make it more reliable after it blew and you fixed it ? The FETs, the diodes, the coil, or a combination of these perhaps ? When replacing trannies, I always choose ones higher up the range in that family for instance, so if the original was a MJE13003, I'd got with a 13009, even though these are not FETs...just a random example. Cheers guys.

[stj]

Re: SMPS output power upgrades and limits

i would say the bottleneck is when the transformer saturates.
i would also say that most psu's are already running on the limit for cost reasons.

[Dannyx]

Re: SMPS output power upgrades and limits

So you say the transformer would be the decisive factor. I was curious because I got hold of a generic adjustable power brick which claims 120w (I shall post some pics soon). I hung onto it because of its reasonably strong construction, the case being all aluminum and stuff and also because it's also a step-up converter, taking a 12v input and stepping it up to anywhere from 15 to 24v, which is handy for laptops since I can run them off a battery or in my car. Thing is, the electronics inside look pretty unconvincing, I mean you take a look inside and it's clear it's not gonna take those 120w, is it ? :| There are no other figures on it, like the amps (maybe they were written on the box it came in), so we can only take guesses. One other option would be to test it to destruction by loading the hell out of it and observing at which point it (and what) fails

[Dannyx]

Re: SMPS output power upgrades and limits

Here's the brick in question...you're probably seen these before....the soldering is pretty piss-poor on some of the components on the back. It gets the job done at least.

There's two transistors: one on the primary and one on the secondary, plus a diode next to the one on the secondary. From what I can make out, when plugged into the mains, it operates like a regular SMPS: the primary transistor runs the transformer which is then rectified and fed to the output. When powered from 12v, it operates in a boost converter configuration using the transistor and the coil on the secondary side. I don't know what would happen if you hooked up BOTH the mains and the 12v input A resistor network switches between the different outputs. Pretty cool idea...I like it actually. Very hackable too, like if I wanted different voltages, I could hack that resistor divider to get whatever I want, maybe with a pot or a trimmer or something. I already did this with a cigarette socket mobile phone charger and turned it into a temperature controlled driver for a fan: by using an NTC in place of one of the feedback resistors, the output voltage would go down/up depending on the temperature, but I digress

[sam_sam_sam]

Re: SMPS output power upgrades and limits

Where did you get this power supply from
Could you please provide me with the link to it
Thanks

[Dannyx]

Re: SMPS output power upgrades and limits

Where did you get this power supply from
Could you please provide me with the link to it
Thanks

Found it in a junk box at work

[stj]

Re: SMPS output power upgrades and limits

i think i'v seen them on ebay, or atleast something similar with a line of leds on it.

[Dannyx]

Re: SMPS output power upgrades and limits

It's a generic thing as far as I can tell, probably going by many different makes and models, but the concept's the same: universal power brick.

[momaka]

Re: SMPS output power upgrades and limits

i would say the bottleneck is when the transformer saturates.

^ That.
Also, for PC ATX PSUs, especially the cheaper ones, the size of the output toroid and size/capacity/count of output filter caps are another two factors. After all, who cares if your PSU can output XX Watts of "raw" power if the output ripple will be so terrible that it kills all of the attached components to it or makes your system crash.

i would also say that most psu's are already running on the limit for cost reasons.

Either that, or crappy APFC circuit designs are making everything run marginal.

Regarding that universal power adapter:
You have one heatsink for the primary and secondary-side silicone. This is a *terrible* idea, especially with the way it is running close to other primary components. Basically, the isolation of this adapter (if you can even call it that) is absolutely marginal, if not outright dangerous.

In terms of limits, that would mostly come from the transformer size and also the power dissipation from the rectifiers and primary-side MOSFET. With bigger heatsinks and/or active cooling, you may be able to push more power, but who knows how much before that transformer saturates.

[Dannyx]

Re: SMPS output power upgrades and limits

Yeh, no need to bother - I just closed it back up and keep it around for emergency purposes only, when I quickly need to test something and don't feel like busting out my bigger sh!t just for that...