Replies

1. 
   DaveRicheson | Sep 12, 2004 08:40pm | #1

   I hope this kicks you  question back to the top.

   We are looking for something to test small UPS units with. Right now all we are doing is charging them up and then plugging a work light into them to approximate the amp draw of a pc, and time how long they last.

   Gotta be a better way.

   BTW I found out that the only way to know if a ups unit will work when needed is unplug it from the ac source. The test button on the front of the unit doesn't mean a thing.

   Dave

1. 
   RenaissanceRestorations | Sep 20, 2004 04:36am | #5

   Most of the UPS power systems have chart showing the load and amount of battery life remains with a certain load on the UPS. I've used APC UPS units in some communication wiring installations on older homes, which have a self test and calibration feature built into the unit, as well as the software. http://www.apcc.com is APC's web site...Renaissance RestorationsAntique & Victorian Home Restoration Serviceshttp://www.renaissancerestorations.com

1. 
   DaveRicheson | Sep 22, 2004 01:19pm | #7

   Thanks for the link.

   Our It department already has a bunch of the UPS units. They just ask us to test them, to find out which ones are still good and how long they might last. I ask for performance and maintenace literature on the several different models and manufactures they have.....got a blank look.

   Guess I'll gather mnfg. and model numbers and do my own search for information on them. Seems like Imformation Technologies doesn't have any information on what they purchased.

   So far our bench tests have found 10 units that won't charge or hold it under load for more than 3 min., and 16 unit that work for 12 to 18 min.. Still have about 25 more to check out, but may not need them.

   We are upgrading our building UPS unit and batteries. $900K + if everything goes as planned. More if an unexpected outage occurrs.

   Dave

1. 
   BillHartmann | Sep 22, 2004 06:11pm | #9

   You might check with Gates, Panasonic, and the other manufacutres of the sealed L.A. batteries.

   I think that you will find them fairly generic in characteristics.

   I am guessing at this that they will have a high internal resistance when bad so a test under load would be the best way to determine this.

   But it might be easier to test the unit as you are doing it with a some lights or heater elements. With those you just plug in. But you have to disassemble the unit to test the battery.

   BTW, when I replace the battery in my APC a got a replacement for a local industrial battery distributor or about 2/3's the cost of one from APC.

   Edit, on 2nd through the APC did not have a battery for my unit, it was obsolete and they would give me 20% off a new one or something like that.

   Edited 9/22/2004 11:12 am ET by Bill Hartmann

2. 
   OneofmanyBobs | Sep 23, 2004 12:48am | #10

   About 3 years is all the useful life you can expect from a sealed lead-acid battery.  Less if you allow them to discharge all the way down to zero.  I've used a lot of UPS units for a lot of years.  Rarely you find a battery thats maybe 6 years old and still fairly good.  I've not found much difference between brands.  We write the install date on the battery and on the UPS and replace every 3 years.  Not much sense in having a UPS that dies when you need it most.  Not cost effective to spend time testing.  We know the answer already.  If its old, its bad.  Wet cells can last much longer but are more complicated to deal with.  Not good for individual user UPS units, but practical for facility systems.

   The quickest and most reliable test is to read the date code on the battery.  An hour (or much worse) to recover lost data is more expensive than a new battery.  We don't buy units where the battery can't be removed without major surgery.  The little home-office units are basically disposable.

2. 
   4Lorn1 | Sep 19, 2004 08:50am | #2

   On most AA batteries there is a set voltage curve. This primarily depends on the chemistry of the battery. The standard alkaline curve is different from the lead-carbon curve and different from the carbon-chloride curve and different from the nickel-cadmium curve. It will also vary between manufacturers. Most curves will shift with temperature. keeping the flashlight inside your coat in frigid weather can extend the usable life and power output of most cells.

   Most curves will have a 'knee', a bend in the curve transitioning between a very gradual decrease in voltage before and a much faster loss of voltage after, in voltage. This is typically about 1/3rd remaining capacity. A simple voltmeter can identify which side of the knee your on and a rough estimate of where you are on the curve.

   A very sensitive, and calibrated, voltmeter can give a reading that when combined with the manufacturers documentation of the typical curve for their batteries will give you a reliable and much closer estimate of the power remaining. Identifying the state before and after this knee to a close tolerance.

   This issue has troubled designs for a long time. Typically the answer is to base longevity under any set use by repeated experimentation, allow a little extra to account for variations between batteries and a multiplier to give a margin of safety. Once this limit is hit the batteries are replaced without any further testing. This is the general course for critical situations where the device just has to work to a set standard.

   The other approach is to have batteries made to a very small tolerance to increase consistancy, generally all batteries produced in the last ten years or so meet this standard, what used to be a special grade. Then a voltage was set to conform to the known consistent voltage curve. At this set point battery sets were cycled or replaced. The consistency of manufacture and the known curve allowed the batteries to serve reliably until near the very end of their ability to drive the device.

   Time in use and a voltage set point have been the main methods I know of for dry cell batteries. Liquid electrolyte batteries, like most automotive batteries, can be tested for charge by checking the chemistry of the electrolyte. An inexpensive tester tests the fluid density.

   Some dry cells, typically used as hearing aid or technical batteries, have little or no 'knee'. Silver cells are used for this reason as voltage standards in some voltmeters as they vary very little over their service life. Silver cells are not very highly efficient as energy stores, they are also expensive on a per watt-hour basis, but they are very reliable and consistent.

1. 
   HeavyDuty | Sep 20, 2004 04:06am | #4

   Thanks Lorn, there's a lot of knowledge to be absorbed here.

   When you mentioned to measure the battery with a voltmeter, do you put the battery under load? If so what kind of a load are we talking about?

1. 
   4Lorn1 | Sep 20, 2004 04:52am | #6

   Mostly, in my experience, they don't put it under load to read the voltage.

2. 
   moltenmetal | Sep 22, 2004 03:35pm | #8

   4LORN1:  sounds like you're the battery man.  From what I've read, there's not much difference between alkaline batteries between brands for most normal loading applications.  I saw a guy who measured his own discharge curves for various brands of batteries under both high drain and low drain conditions and found little difference between alkaline battery brands.  He recommended buying the cheapest alkalines you could find.  Do you have an opinion on the subject?

1. 
   4Lorn1 | Sep 23, 2004 04:20am | #11

   The few organized studies I have seen deal with AA size batteries as they are the most common in the consumer electronics the magazine, Popular Electronics, was centered on. Their conclusion closely matched what you said.

   Alkaline batteries are generally the best value in terms of power per dollar but, for the most part alkaline batteries are all the same. There was, this study is years old so the information could have changed, some small advantage to Duracells. Something like 2%. Of course this brand also sold for slightly more so, on a dollar per mAh, milliampere-hour, basis they offered no advantage.

   Another study I saw, if I remember correctly it was Consumer Reports, and dealing with a wider selection of battery sizes showed that most alkaline batteries were very close. On this basis price was the single best way of determining value. Sears batteries, produced by a major battery company but sold under the Sears name at a discount, rated high because they gave average performance at a reduced cost.

   I know a lot less about the newer alkalines that tout higher performance. I'm not sure how they define 'performance'. These units cost significantly more I notice. Will the additional performance pay for the added cost? It remains to be seen.

   Of course there are situations where a simple dollar per mAh is not appropriate. Unusual and critical conditions can shift things around. Lithium batteries can sometimes be justified for camping, even when carrying extra batteries is fairly easy and backup light sources are readily available, simply because lithium batteries work really well at reduced temperatures.

   A camping light is often turned on for a moment and then off again at low temperatures. Like to look at the map, find the gorp or untangle the tent line. You don't want to have to deal at 5 degrees F with a weak light until you warm it up. Lithium cells are also lighter than alkalines. Any hiker can tell you this is worth paying a little more for.

   Lithium cells also store well. Five years of storage leaves you a mostly dead set of alkaline cells but lithium are still strong. While the lithium cells cost a lot more you get to use more of their capacity and have a decent chance of having something to work with when you need it.

   The other consideration is criticality. If your likely to be in a situation with a flashlight, or other device, that will be unused for long periods but it absolutely has to work and there is going to be no chance to replace batteries. Military, rescue, escape from confined spaces, aeronautics are the major categories but something as prosaic as a glove compartment flashlight can, in some cases, qualify.

   In these cases having the most reliable and most capable batteries available can save a life. The extra cost of the lithium cells can easily be justified.

   As an electrician I use flashlights. Dozens of them. Mostly I use alkaline batteries but also keep spare batteries on nearby. Don't forget spare bulbs for your flashlights. Keep a spare on board the flashlight. Most Mag lights have this feature and I look for it on lights I buy. Be sure to replace any you use. Because I go through so many I keep half a dozen of each type on my truck. Cheap insurance.

1. 
   moltenmetal | Sep 23, 2004 04:55am | #12

   A most helpful reply.  Thank you very much!  I personally like NiMH rechargeables because they're so cheap these days.  I'm also fond of the very expensive Li-ion battery in my digital camera- worth every penny IMHO.

3. 
   Mitremike | Sep 19, 2004 09:15am | #3

   Hello Shoeman . saw ya earlier about that 9.6 bosch impact looking forward to them coming out with theirs--thinking 12 volt myself. did you score and goodies at the show? Mike