UPS built into computers

The goal of this question is to only supply power to a computer long enough for it to stay up from a flash blackout. (no monitor) I am not talking about even enough power to allow for shutdown. At least 90% of the time the power goes down here is just from the flash blackout. This could happen 3 times with only 2 or 3 seconds down time. So what I am talking about is 10 seconds backup for a computer with say a 400w PS.

Anyone know what something like this would add to the cost of a computer? How small do you think the battery could be? Think it could be small enough to mount in a 5.25inch bay? Could it be built to just accept the connections from the PS and then furnish the connections to the mobo?

Reply to
Metspitzer
Loading thread data ...

formatting link

Reply to
George Jetson

Yeah. Something like that. It weighs 8lbs so I am guessing it still takes a pretty large battery.

It also looks like this one is switching the 120V. The one I had in mind would do the internal voltage after the power supply.

100 bucks is pretty pricey. Pretty cool though.
Reply to
Metspitzer

Just add a bunch of capacitors.

Reply to
tuinkabouter

Yawn. That might handle a few milliseconds. If they could handle a real outage the turnon currnt would trip the main breaker in the building, since they would charge on the first cycle. (1/60th of a second)

Reply to
Michael A. Terrell

Seriously? You can't figure a way to limit the charging current?

Reply to
dave y.

You're talking through the back of your neck. You're asking for a high-integrity backup system yet don't want to pay for it. Get real, for pity's sake.

John

Reply to
John Nice

Seriously? Show me how to do that, and still be ready for the next dropout a few seconds later. It's not like you have hours or days between dropouts. Calculate how large of a capacitor you need to hold the power supply at full output for just 10 seconds. then see how long it will take to charge at near the normal supply current. Also calculate the size of all the required capacitors and consider the safety implications if some or all explode from a line surge. I've seen electrolytics go through the ceiling, and others rip the screws out of a metal case when they exploded.

Reply to
Michael A. Terrell

The BEST large electrolytics are lead-acid :-)

John

Reply to
John Nice

We did that once. A 12 volt SLA battery was inside the power supply. Same circuits that created DC voltages also charged the battery and could obtain power from that battery. Battery only provided power on AC power loss - to preserve battery life expectancy. A UPS in a smaller package requiring less parts and money. The product had no market. People wanted a separate UPS anyway as if it does more because it was separate. Undoing well entrenched bias is difficult.

Reply to
westom

The UPS in the link posted earlier is supposed to furnish 300W for 8 min. This solves 100% of the power loss problem, but it is more expensive and heavy.

What I had in mind only solved 90% of the problem, but could be made lighter and cheaper. The one I had in mind would not need a serial port to shut down the computer either, although I am guessing that what ever holds the logic to do the switching from normal to UPS would do shutdown for almost nothing.

I think automatic back up should be built into software and not hardware. My computing is for fun and not work, though.

Reply to
Metspitzer

That isn't a capacitor, and do you want a lead acid battery to vent inside a PC case? :)

There was a "UPS on a card" for the original PC/XT computers that took three slots that I saw one ad for.

Reply to
Michael A. Terrell

XP has a built in monitor.

Reply to
Michael A. Terrell

The most parts efficient solution would be two 4700uf x 200v electros at the switchmode's line to DC conversion stage with diodes and resistors. The caps are about 40mm x 90mm in size, so plan on a suitable enclosure somewhere inside the computer case. It would likely run a few seconds, someone else can do the math. The Digikey prices are around 18 USD each so I would go buy the external UPS.

Reply to
George Jetson

Michael A. Terrell Inscribed thus:

I've still got one of those ! Its sitting back in its box somewhere in the garage. If I recollect correctly it was only capable of sustaining the original IBM XT for about 5 minutes. Yes it gave you time to park the HDD heads and switch off. But considering that during a power failure you couldn't see anything on the screen and it didn't activate a shut down automatically, it was worthless !

Michael, mine took three slots and two of those were covered by the sealed lead acid battery. It also had a tendency to get quite hot whilst charging the battery. All in all an expensive gimmick !

Reply to
Baron

That was my thought when I saw the press release. Of course, 95% of the early UPS for desktop computers were crap. Companies would spring up, then disappear a few months later.

Reply to
Michael A. Terrell

You still aren't dealing with the turn on surge. What good would it do, if you trip a circuit breaker when you turn the computer on? You can't just stick a big pair of caps in the computer and call it a day.

Reply to
Michael A. Terrell

Remember, there is often a fair bit of wire between your computer and your circuit breaker panel, it has resistance. All switch mode PSUs have a filter at there input with an inductor in line with the supply, inductance will limit current surges (If you did go down this road you might have to replace it with an inductor of higher current rating - but don't change the value).

How many other appliances or pieces of equipment do you have which also have high surge currents - motors, heaters, arc welder - do they trip your incoming breaker every time they are used? I doubt the size of capacitor mentioned here would have much effect in the grand order of things, especially as George mentioned resistors in his post - presumably current limiting.

Reply to
Stuart

Michael is discussiong a power on surge. That power on surge can have other consequences. One common solution is to also install an inrush current limiter. Low voltage is desireable when first powering on electronics.

Reply to
westom

And you don't get it. Adding a resistor makes a simple R/C timer. If you limit the current, it takes longer to charge. With that time limit, it can't recover fast enough for a second event. Most failures around here are in groups of three, as the reclosers attempt to restore power twice, before leaving the line open.

Reply to
Michael A. Terrell

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.