UPS - Poor Guy's almost-UPS

[jebatty]

Just a thought for experience or reaction. Could not a person make an almost-UPS with a DPDT relay and an inverter hooked up to a deep cycle battery(ies) by having the coil energized by line power, having the load connected to the poles, having the side of the relay energized by the coil hooked up to line power, and having the other side of the relay hooked up to the inverter?

Line power energizes the load circuit until power failure. Line power fails, relay switches to other side and inverter kicks in to energize the load circuit.

 

[raybonz]

Just a thought for experience or reaction. Could not a person make an almost-UPS with a DPDT relay and an inverter hooked up to a deep cycle battery(ies) by having the coil energized by line power, having the load connected to the poles, having the side of the relay energized by the coil hooked up to line power, and having the other side of the relay hooked up to the inverter?

Line power energizes the load circuit until power failure. Line power fails, relay switches to other side and inverter kicks in to energize the load circuit.

Hi Jim,
Yup what you described would work fine.. Essentially that is how an automatic transfer switch works.. You can buy a good size inverter for less than 100 bucks.. Amazon has some good deals on inverters.. If you decide to something like that I would suggest breaking the cable between the battery and inverter because most inverters will drain a battery with no load attached at the inverter.. Basically whatever your AC load will be multiply by 10 for battery current..
If your AC load is 2 amps then your battery load will be 20 amps plus whatever losses the inverter has..

Ray

 

[jebatty]

Why not then a 3PDT, and the extra pole switch the battery on at the same time the load shifts to the inverter? Just need to make sure the contacts are rated for the possibly high DC load.

 

[raybonz]

Why not then a 3PDT, and the extra pole switch the battery on at the same time the load shifts to the inverter? Just need to make sure the contacts are rated for the possibly high DC load.

DC is rough on contacts and personally I would feel better using 2 relays to be sure the 2 voltages are isolated.. My background is licensed electrician that has done industrial control for about 25 years.. Better safe than sorry.. You'd be better off with a contactor instead of a relay in my opinion particularly controlling the DC side..

Ray

 

[jebatty]

Makes lots of sense. Not good to mix circuits/voltages in the same device/box.

My thought in heading in this direction is that my intuition tells me that a computer UPS system, even if sufficient VA and big batteries, is not designed to operate for several hours, but only minutes, and I would not want to rely on a UPS that trips a thermal breaker during an extended run time. Maybe you know more about whether this is a valid concern.

Thanks for the advice.

 

[raybonz]

Makes lots of sense. Not good to mix circuits/voltages in the same device/box.

My thought in heading in this direction is that my intuition tells me that a computer UPS system, even if sufficient VA and big batteries, is not designed to operate for several hours, but only minutes, and I would not want to rely on a UPS that trips a thermal breaker during an extended run time. Maybe you know more about whether this is a valid concern.

Thanks for the advice.

The larger the capacity of a UPS the longer the run time.. To size the UPS you need to know the load that it will be supplying and how long you want it to supply power to the load..The advantage of a UPS is it always plugged in and does the switching for you and many will act a voltage regulator protecting against sags, spikes etc. I'd say it would be safer to use a UPS as everything is self contained as opposed to a large marine battery with an inverter. On another note the UPS batteries will eventually need replacement You will probably get 3 or 4 years of service from the batteries and UPS's monitor battery status via onboard electronics. I liked the idea of a pellet stove until I realized they always need power to run.. Wood stoves while primitive in comparison to a pellet stove give many years of service and rarely require parts.. At least that's been my experience..

Ray

 

[jebatty]

I guess my question is this. I have a 950VA-1400watt UPS, with 3-12v internal batteries, 7ah. I have pulled the batteries and re-wired to 3-12v 105ah gel cells. As designed, the UPS on internal batteries normally would run at full load only for 3-5 minutes, I estimate, before battery cut-off voltage would be reached. If I use the big gel cells, I may get an hour or more at full load. Do you think the UPS is designed for an extended run like this? Or might a thermal breaker trip out on an extended run?

Actually I don't need to run at full load to keep my gasification boiler and two circ pumps running, only about 300VA.

 

[raybonz]

I guess my question is this. I have a 950VA-1400watt UPS, with 3-12v internal batteries, 7ah. I have pulled the batteries and re-wired to 3-12v 105ah gel cells. As designed, the UPS on internal batteries normally would run at full load only for 3-5 minutes, I estimate, before battery cut-off voltage would be reached. If I use the big gel cells, I may get an hour or more at full load. Do you think the UPS is designed for an extended run like this? Or might a thermal breaker trip out on an extended run?


Actually I don't need to run at full load to keep my gasification boiler and two circ pumps running, only about 300VA.

Most breakers are designed to handle up to 80% continuous load for an indefinite period. I suggest you try running your system and give your UPS a real world test and run it until the UPS cuts out this way you'll know how it will perform and also give you a baseline to compare in the event of a problem with the UPS..

Ray

 

[Redox]

I guess my question is this. I have a 950VA-1400watt UPS, with 3-12v internal batteries, 7ah. I have pulled the batteries and re-wired to 3-12v 105ah gel cells. As designed, the UPS on internal batteries normally would run at full load only for 3-5 minutes, I estimate, before battery cut-off voltage would be reached. If I use the big gel cells, I may get an hour or more at full load. Do you think the UPS is designed for an extended run like this? Or might a thermal breaker trip out on an extended run?

Actually I don't need to run at full load to keep my gasification boiler and two circ pumps running, only about 300VA.

I did this to an old APC 1000va that I picked up at a hamfest for next to nothing. I found that the heat sink on the output transistors got very hot on extended runs. I ended up putting a small muffin fan on the side of it and now it runs for 3-4 hours without a problem. The fan runs continuously plugged into the UPS. One day, I might put the fan on a relay that only turns on the fan while the UPS is operating on batteries, but this is down low on the list of priorities.

 

[Gooserider]

As stated, real world testing is the only way to be sure, but typically a computer UPS is designed to give 10-15 minutes minimum on full load when new. This number will drop somewhat as the units age, mostly because of battery degradation.

As to power handling ability, and the likelyhood of a thermal shutdown, I would say that a lot would depend on the design of the UPS - I would be less inclined to trust a basic grade unit designed for a home system than I would one that was rated for high power output in a server-room type environment. Also you are less likely to have problems with a unit that you are not pushing to the max.

The batteries can be replaced in most UPS's, as long as you match the voltage, but caution is advised if changing from the standard sealed gel-cell type batteries the units are built with to any home brew alternatives like marine batteries... A gel-cell battery is designed not to emit gasses, or corrosive / toxic fumes, but any sort of automotive or marine battery can do so, and provisions should be made to ensure adequate ventilation, corrosion protection, and so forth.

In terms of the OP's question about doing a "poor-man's UPS" the proposed idea would work, but may well not be any less money than doing something with an equivalent computer UPS. There is also a question of how well it will work with the equipment it's connected to, which is a function of what the equipment is... A significant part of the complication in a computer UPS design is to minimize the "transfer time" - or how long it takes the UPS unit to take over the job when it detects a power loss. A good unit will respond in less than 1/2 an AC wave cycle, or about 1/120 of a second. The relay design would probably take closer to a full second - fine for light bulbs or motors where a power drop won't really hurt anything, but it won't work for anything computerised where a power interruption will case a reset and reboot if you need to worry about preserving the existing machine state...

Gooserider

 

[sleepie]

find a 2000xl smartup on craigslist and hook deep cycle batts to it and all will work fine--it even has the internal fan that comes on when its runnning also it does a bi-weekly check of it self --automaticly---pat