How to test the power requirements of a given system ?

**justpassingby** · 05-10-2007, 01:32 AM

Hi guys !

I have this computer (check my system) powered by an Antec NeoHE430. The case is plugged to an APC RS800 UPS and that's the only device that's powered by the UPS.

Before you yell at me saying that I have to buy a bigger PSU for my rig, I have to specify that I'm in Europe on 220V, which is more efficient than 110V, and that the maximum power I've seen drawn from the computer using the APC bundled software was 151W, and that was when doing a GPU and a CPU stress test at the same time. When idle, my computer draws 97W.

I have several questions :

1) Are there softwares ways of knowing how many watts are drawn from a given rail ? If not, how do I use a multimeter while the computer is on to say test the power delivered on the 12V rail while stressing the GPU using Rthdribl. I'll need a detailed tutorial since I've never used a multimeter.

2) Are there softwares out there that could draw maximum power from a given system by stressing the CPU, GPU, RAM, hard drives, CD/DVD drives, USB devices and sound card simultaneously ? If not, which software combination would you recommend : Prime95, 3Dmark05, Rthdribl, HDtach, Everest stability test, superPI, ... ?

The idea would be to benchmark the power needs of a given system. The PSU calculators I've seen are based on 110V current and I believe they're likely to give overrated PSU requirements for computers that will be used on 220V (my system drawing a max of 150W would be a good example).

I'd like to be able to measure the max total wattage drawn from a given system, as well as the max amperage drawn per rail.

Edit : just noticed dai had added this to the stickies :
HOW TO: Measuring PSU Rails
I guess this should answer my first question, at least partly. I'd still like it if you could provide me with specific instructions as to how to measure the max amperage drawn by my video card.

**linderman** · 05-10-2007, 06:27 AM

Hello Just

the timing of your question is rather ironic.

there is a discussion in the tweaker section of HW where were are in the midst of doing a guide just as you are requesting

i would expect it to be done with in about a week

to answer your question #2 I would say running a stress test like prime95 or orthos ( for dual core) would load the system plenty to get a value for the cpu amp load

then you would need to go to 3dmark and test the max load on the video end of things

then you would know what each major player is drawing amp wise ?

as for your max load at the wall of 150 watts I agree with that assessment and basically where you are headed with that evaluation.

However there are some factors which you must identify as far as the "real" capabilities of your PSU

#1 ) the 430 watt rating: is this continious delivery or peak wattage at 100% output ????

#2 ) as you know; there is NO electrical device know to man kind that can be run at more than 75% to 80% of its capability without suffering a severely shortened service life.

#3) what was the operating ambient temperature of the environment during which the test unit was given its wattage and amperage "rating" this is the real stumper, many 70% efficient units are tested and rated for output at 25C ambient operating environment temps, this "trick" boost their capabilities substantially and gives us a false picture for reality operating, the manufacturer is not lying to us, but its borderline deceptive in my opinion. As no user will buy such a rated unit and operate the unit in a 25C temp case ?

so all this over kill in the units we suggest is really brought on by lack of "real" available information of most capabilities of units sold.

Thats why its common to see PSU testers that have discovered 580 watt units that tripped the overload protections at 366 watts

or 480 watt units that trip the overload protections at 277 watts

then you must add 10% loss per year due to capicator aging, in essence everyday of operation wears a PSU electronics out

the hotter it is in your case and the more hours you run your machine (like 24 hours a day, 7 days a week, 365 days per year) capacitor aging % will rise

10% being the most modest 30% being the most dramatic

so in essence, the average PSU is expected to live three years, so you must add 30% for cap aging, then add 30% for headroom, because no elec unit can run anywhere near its full rate. 80% can be expected from top quality units in the best built units! not the economy or affordable models most people prefer.

so in summary we are @ 430 watt rated PSU can really deliver 331 watts if the rating was at the same temp as the internal temp of YOUR computer

its not uncommon for the diff in 25C ratings to change(lessen) a PSU by 50 watts when the higher normal temps of 45C are applied to the unit

also can; we are assuming with the 331 watts above that the PSU has a continous ratign of 430 watts, again this is usually not accurate with an attractively priced PSU; usually these are "peak" ratings

in summary; sure your system has a very modest load at 150 watts of power draw, and the 430 watt PSU you are currently using, IS suited for the job, however I am VERY confident you have no were near the head room you may "think" you have.

And the vast reason why you are doing well; is the NEO is a high quality unit.
its 85% efficient and I will be its "rated" at 50C instead of 25C
but I am pretty sure the 430 watt moniker is peak wattage not sustained wattage

just like your car could be sold as rated for 120 miles per hour, yes it may actually go 120miles per hour, but will it do 500 laps at daytona race before the motor blows >>>>> NOPE

**justpassingby** · 05-10-2007, 08:09 AM

Thx a lot for that feedback Joe ! I guess I can wait a week to read that guide.

Thanks to you and other hardware techs (remember this topic ?) I'm now aware of most of the factors that come into play as to determine the real capabilities of a given PSU.

The NeoHE is a very good quality PSU that is supposed to deliver 430W continuous output at 50°C ambient temperature and boasts a very high efficiency (85% on the box, more like 75% in real conditions under full load). Those results will off course decrease as time passes by. See this link for some real testing on that PSU :
http://www.silentpcreview.com/article273-page4.html

I wouldn't rely on those numbers too much and would say I'm fine with my NeoHE430 as long as my system doesn't draw more than 250-300W. Should it draw closer to 300W I would consider upgrading my PSU. My current system showed to only draw a maximum of 150W from the UPS which means my system really uses only around 75% of that power (75% being the PSU's average efficiency). I guess those good results are partly due to the more efficient 220V input.

But what could cause problem with my rig is the amperage on the 12V rail : most manufacturers recommend using a PSU that can deliver up to 18 or 20A on the 12V rail for use with a 7600GT. The NeoHE430 only has 16.

I used everest to monitor the voltages while I was running the 3Dmark05 benchmarks and all voltages on all rails were stable and within specs (and I got a normal score compared to similar systems). I did the same test with the latests games I've been playing and got stable voltages as well. I guess I may have just enough power for my video card but I really want to determine precisely the max power drawn on the 12V rail at full load. If I find it to be close to 13 or 14 amps I would change the PSU.

As a side note, the only PSU available at my local retailer that has more than 18 amps on the 12V rail is the Enermax Infiniti 720w. This is a great power supply with 3 12V rails @28A, but I fear it might be a little overkill for my system (not to mention it costs more than double the price of my NeoHE430). How bad would it be if you use only 15% of a PSU nominal power ? I suppose the efficiency would decrease. Would the PSU wear faster ?

**linderman** · 05-10-2007, 08:46 AM

your current PSU can deliver more than 16 amps

your video card can draw as much as 75 watts ( 7 amps) off the pci slot itself which is feed by 12v 1 rail and the 20 plus four connector

then if your video card wants more than 7 amps ( it will want more than that) it gets the extra power off rail 12v2 which has 16 more amps

you must look at the combined 12v rail capabilities of your unit

yes' the efficiency of the 720 watt unit would be horrible for you, your unit must be in the curve of the power to get to the 85% efficiency, your load would not be in the "sweet" spot for efficiency

floydfan · 05-10-2007, 09:45 AM

a 7600gt requires 36w at full load, it does not need any more power than the pci-e slot gives it.

**linderman** · 05-10-2007, 09:57 AM

Floyd is correct, your video card is not a power hungry version

**Kalim** · 05-10-2007, 10:20 AM

Do a disk defrag or bench your HDD using HD Tach (etc) while looping 3DM05/3DM06 and running RTHDRIBL/Prime 95 and you'll see what your max power pull is.

But to accurately simulate it, you will need to get into safe mode, disable all extra processes/services (no net/firewall/av etc), system restore off, set those stress applications as "real-time" in Task Manager, and then load them up for about 30 minutes while using something like a Weibo, Extech or Kill-A-Watt to measure VAC.

Make sure your core is fully loaded. That'll give you values of what your system will pull at max. loads at present.

Yea, EU efficiency is quite higher than America's. The NeoHE will be over 80% consistent.

**justpassingby** · 05-10-2007, 05:24 PM

Thx linderman. Yup, my video card takes all its power from the pci-express slot, no additional power cable needed.

Do you belive a PSU that's used outside of its sweet spot (on the lower end of the wattage) would wear down faster ?

@ floydfan : thx for the information about my 7600GT wattage. Here are two links that give the same figures for those interested :
The Truth About Graphics Power Requirements V2
Video card power requirement review : Radeon X1300PRO to X1900XTX, Nvidia 6600GT to 7900GTX
Found it funny that the 7600GT is less power hungry than the 6600GT.

I guess the high power requirements stated by manufacturers like eVGA (400W 12V@18A) and BFG (350W 12V@20A) for their 7600GT are meant for single-rail PSU's where all devices draw on the same 12V rail.

My XFX 7600GT XXX is factory overclocked (a bit more than its eVGA and BFG counterparts) but that shouldn't make a huge difference... or would it ? I'm happy to know I'm on the safe side with it drawing less than 3 amps from the +12V

@ Kalim : thx for providing me with that test bench. Wouldn't have thought to do that in safe mode and to tweak those priorities. Everest stability test + RTHDRIBL were drawing around 120W from the PSU in normal mode (151W from the UPS at 80% PSU efficienty). I guess my system could draw 10-15 more watts from using the Raptor and Hitachi HDD at the same time, + maybe another 10 from adding prime95 to push it to its limits.

I don't have access to those measurement equipment but will try to borrow a multimeter from a friend some time. I currently only have softwares ways of measuring the power, could I rely on the PowerChute software (bundled with the APC Back-UPS RS800) to give me precise figures ? How accurate can you expect the power output reported by the UPS to be ?

**Kalim** · 05-10-2007, 09:31 PM

1: That data is not all conclusive. Like it's been covered many times before, on different test benches you have different figures. The maximum you have will be different in most cases to others' maximum. People tend to take them all conclusive and that's very wrong. For instance the G80GTX has been shown as pulling 145W on other bench testing.
2: XFX usually make the most powerful GPUs. The stock nV cards do not pull less or equal to the AIB OC'd card. The OC'd version will pull more, quite easily upto 5-20 watts more on that.
3: IDK about your APC software TBH, maybe it is accurate depending on if it has a voltage/current measure built inside and the transmission is clean, which it usually will. But I have come across something similar before in a study where they found some discrepancies. I seriously do not recall what those faults were, whether minor or large, I don't think they were large at all, something to do with the accuracy affected by communication between the two.

4: A PSU has been given a value of "typical load" for a reason. That typical load is at 50% load, as MFGs expect the users of any PSU to be running most of the time. Much more is harsher on the PSU and less than 25% is usually also bad and not as efficient nor clean. The best spot would be roughly 35-65% loads.
It's no different to most electronics. The harder, faster, hotter and longer you make it work, the quicker it degrades to less quality and dies out.

**linderman** · 05-11-2007, 04:46 AM

if you were to buy a 700 watt unit and your power draw (amp load) is not in the "sweet spot" for efficiency / it will not hurt the PS its just a plain ole poor choice for the system >>>> why waste elec power, it can add up $$$$ over a three year period of time

also dont be fooled into thinking a generic 430 watt unit like say (Apevia, Aspire) would power your system without any problems for a three year period of time

you now have the "right" size unit and a quality built unit

for the most part I have found the APC software to be rather accurate, maybe not down to the last 2-3 watts, but do we really need to be that accurate ??????

I havent found a recently built APC unit that was wrong by much when compared to more elaborate equipment

in reality I think we easily have a 25 watt + or - minus forgiveness factor for what we are doing and evaluating

05-10-2007, 01:32 AM	#1 (permalink)
justpassingby Moderator, Microsoft Support Join Date: Mar 2007 Location: Belgium Posts: 4,965 OS: XP Home SP3 / Ubuntu My System	How to test the power requirements of a given system ? Hi guys ! I have this computer (check my system) powered by an Antec NeoHE430. The case is plugged to an APC RS800 UPS and that's the only device that's powered by the UPS. Before you yell at me saying that I have to buy a bigger PSU for my rig, I have to specify that I'm in Europe on 220V, which is more efficient than 110V, and that the maximum power I've seen drawn from the computer using the APC bundled software was 151W, and that was when doing a GPU and a CPU stress test at the same time. When idle, my computer draws 97W. I have several questions : 1) Are there softwares ways of knowing how many watts are drawn from a given rail ? If not, how do I use a multimeter while the computer is on to say test the power delivered on the 12V rail while stressing the GPU using Rthdribl. I'll need a detailed tutorial since I've never used a multimeter. 2) Are there softwares out there that could draw maximum power from a given system by stressing the CPU, GPU, RAM, hard drives, CD/DVD drives, USB devices and sound card simultaneously ? If not, which software combination would you recommend : Prime95, 3Dmark05, Rthdribl, HDtach, Everest stability test, superPI, ... ? The idea would be to benchmark the power needs of a given system. The PSU calculators I've seen are based on 110V current and I believe they're likely to give overrated PSU requirements for computers that will be used on 220V (my system drawing a max of 150W would be a good example). I'd like to be able to measure the max total wattage drawn from a given system, as well as the max amperage drawn per rail. Edit : just noticed dai had added this to the stickies : HOW TO: Measuring PSU Rails I guess this should answer my first question, at least partly. I'd still like it if you could provide me with specific instructions as to how to measure the max amperage drawn by my video card. __________________ HJT - 5 steps against malware. Post your HijackThis log there and not here ! Posting system specs Last edited by justpassingby : 05-10-2007 at 01:57 AM.

05-10-2007, 06:27 AM	#2 (permalink)
linderman Moderator, Hardware Team Join Date: May 2005 Posts: 14,055 OS: win 2000 pro / Win XP pro My System	Re: How to test the power requirements of a given system ? Hello Just the timing of your question is rather ironic. there is a discussion in the tweaker section of HW where were are in the midst of doing a guide just as you are requesting i would expect it to be done with in about a week to answer your question #2 I would say running a stress test like prime95 or orthos ( for dual core) would load the system plenty to get a value for the cpu amp load then you would need to go to 3dmark and test the max load on the video end of things then you would know what each major player is drawing amp wise ? as for your max load at the wall of 150 watts I agree with that assessment and basically where you are headed with that evaluation. However there are some factors which you must identify as far as the "real" capabilities of your PSU #1 ) the 430 watt rating: is this continious delivery or peak wattage at 100% output ???? #2 ) as you know; there is NO electrical device know to man kind that can be run at more than 75% to 80% of its capability without suffering a severely shortened service life. #3) what was the operating ambient temperature of the environment during which the test unit was given its wattage and amperage "rating" this is the real stumper, many 70% efficient units are tested and rated for output at 25C ambient operating environment temps, this "trick" boost their capabilities substantially and gives us a false picture for reality operating, the manufacturer is not lying to us, but its borderline deceptive in my opinion. As no user will buy such a rated unit and operate the unit in a 25C temp case ? so all this over kill in the units we suggest is really brought on by lack of "real" available information of most capabilities of units sold. Thats why its common to see PSU testers that have discovered 580 watt units that tripped the overload protections at 366 watts or 480 watt units that trip the overload protections at 277 watts then you must add 10% loss per year due to capicator aging, in essence everyday of operation wears a PSU electronics out the hotter it is in your case and the more hours you run your machine (like 24 hours a day, 7 days a week, 365 days per year) capacitor aging % will rise 10% being the most modest 30% being the most dramatic so in essence, the average PSU is expected to live three years, so you must add 30% for cap aging, then add 30% for headroom, because no elec unit can run anywhere near its full rate. 80% can be expected from top quality units in the best built units! not the economy or affordable models most people prefer. so in summary we are @ 430 watt rated PSU can really deliver 331 watts if the rating was at the same temp as the internal temp of YOUR computer its not uncommon for the diff in 25C ratings to change(lessen) a PSU by 50 watts when the higher normal temps of 45C are applied to the unit also can; we are assuming with the 331 watts above that the PSU has a continous ratign of 430 watts, again this is usually not accurate with an attractively priced PSU; usually these are "peak" ratings in summary; sure your system has a very modest load at 150 watts of power draw, and the 430 watt PSU you are currently using, IS suited for the job, however I am VERY confident you have no were near the head room you may "think" you have. And the vast reason why you are doing well; is the NEO is a high quality unit. its 85% efficient and I will be its "rated" at 50C instead of 25C but I am pretty sure the 430 watt moniker is peak wattage not sustained wattage just like your car could be sold as rated for 120 miles per hour, yes it may actually go 120miles per hour, but will it do 500 laps at daytona race before the motor blows >>>>> NOPE __________________ If you cant be polite and friendly; then please be silent. Last edited by linderman : 05-10-2007 at 06:31 AM.

05-10-2007, 08:09 AM	#3 (permalink)
justpassingby Moderator, Microsoft Support Join Date: Mar 2007 Location: Belgium Posts: 4,965 OS: XP Home SP3 / Ubuntu My System	Re: How to test the power requirements of a given system ? Thx a lot for that feedback Joe ! I guess I can wait a week to read that guide. Thanks to you and other hardware techs (remember this topic ?) I'm now aware of most of the factors that come into play as to determine the real capabilities of a given PSU. The NeoHE is a very good quality PSU that is *supposed* to deliver 430W continuous output at 50°C ambient temperature and boasts a very high efficiency (85% on the box, more like 75% in real conditions under full load). Those results will off course decrease as time passes by. See this link for some real testing on that PSU : http://www.silentpcreview.com/article273-page4.html I wouldn't rely on those numbers too much and would say I'm fine with my NeoHE430 as long as my system doesn't draw more than 250-300W. Should it draw closer to 300W I would consider upgrading my PSU. My current system showed to only draw a maximum of 150W from the UPS which means my system really uses only around 75% of that power (75% being the PSU's average efficiency). I guess those good results are partly due to the more efficient 220V input. But what could cause problem with my rig is the amperage on the 12V rail : most manufacturers recommend using a PSU that can deliver up to 18 or 20A on the 12V rail for use with a 7600GT. The NeoHE430 only has 16. I used everest to monitor the voltages while I was running the 3Dmark05 benchmarks and all voltages on all rails were stable and within specs (and I got a normal score compared to similar systems). I did the same test with the latests games I've been playing and got stable voltages as well. I guess I *may* have just enough power for my video card but I really want to determine precisely the max power drawn on the 12V rail at full load. If I find it to be close to 13 or 14 amps I would change the PSU. As a side note, the only PSU available at my local retailer that has more than 18 amps on the 12V rail is the Enermax Infiniti 720w. This is a great power supply with 3 12V rails @28A, but I fear it might be a little overkill for my system (not to mention it costs more than double the price of my NeoHE430). How bad would it be if you use only 15% of a PSU nominal power ? I suppose the efficiency would decrease. Would the PSU wear faster ? __________________ HJT - 5 steps against malware. Post your HijackThis log there and not here ! Posting system specs

05-10-2007, 08:46 AM	#4 (permalink)
linderman Moderator, Hardware Team Join Date: May 2005 Posts: 14,055 OS: win 2000 pro / Win XP pro My System	Re: How to test the power requirements of a given system ? your current PSU can deliver more than 16 amps your video card can draw as much as 75 watts ( 7 amps) off the pci slot itself which is feed by 12v 1 rail and the 20 plus four connector then if your video card wants more than 7 amps ( it will want more than that) it gets the extra power off rail 12v2 which has 16 more amps you must look at the combined 12v rail capabilities of your unit yes' the efficiency of the 720 watt unit would be horrible for you, your unit must be in the curve of the power to get to the 85% efficiency, your load would not be in the "sweet" spot for efficiency __________________ If you cant be polite and friendly; then please be silent.

05-10-2007, 09:45 AM	#5 (permalink)
floydfan Registered User Join Date: Oct 2006 Posts: 367 OS: xp pro My System	Re: How to test the power requirements of a given system ? a 7600gt requires 36w at full load, it does not need any more power than the pci-e slot gives it.

05-10-2007, 09:57 AM	#6 (permalink)
linderman Moderator, Hardware Team Join Date: May 2005 Posts: 14,055 OS: win 2000 pro / Win XP pro My System	Re: How to test the power requirements of a given system ? Floyd is correct, your video card is not a power hungry version __________________ If you cant be polite and friendly; then please be silent.

05-10-2007, 10:20 AM	#7 (permalink)
Kalim Roaming To Help Join Date: Nov 2006 Posts: 5,609 OS: Many	Do a disk defrag or bench your HDD using HD Tach (etc) while looping 3DM05/3DM06 and running RTHDRIBL/Prime 95 and you'll see what your max power pull is. But to accurately simulate it, you will need to get into safe mode, disable all extra processes/services (no net/firewall/av etc), system restore off, set those stress applications as "real-time" in Task Manager, and then load them up for about 30 minutes while using something like a Weibo, Extech or Kill-A-Watt to measure VAC. Make sure your core is fully loaded. That'll give you values of what your system will pull at max. loads at present. Yea, EU efficiency is quite higher than America's. The NeoHE will be over 80% consistent.

05-10-2007, 05:24 PM	#8 (permalink)
justpassingby Moderator, Microsoft Support Join Date: Mar 2007 Location: Belgium Posts: 4,965 OS: XP Home SP3 / Ubuntu My System	Re: How to test the power requirements of a given system ? Thx linderman. Yup, my video card takes all its power from the pci-express slot, no additional power cable needed. Do you belive a PSU that's used outside of its sweet spot (on the lower end of the wattage) would wear down faster ? @ floydfan : thx for the information about my 7600GT wattage. Here are two links that give the same figures for those interested : The Truth About Graphics Power Requirements V2 Video card power requirement review : Radeon X1300PRO to X1900XTX, Nvidia 6600GT to 7900GTX Found it funny that the 7600GT is less power hungry than the 6600GT. I guess the high power requirements stated by manufacturers like eVGA (400W 12V@18A) and BFG (350W 12V@20A) for their 7600GT are meant for single-rail PSU's where all devices draw on the same 12V rail. My XFX 7600GT XXX is factory overclocked (a bit more than its eVGA and BFG counterparts) but that shouldn't make a huge difference... or would it ? I'm happy to know I'm on the safe side with it drawing less than 3 amps from the +12V @ Kalim : thx for providing me with that test bench. Wouldn't have thought to do that in safe mode and to tweak those priorities. Everest stability test + RTHDRIBL were drawing around 120W from the PSU in normal mode (151W from the UPS at 80% PSU efficienty). I guess my system could draw 10-15 more watts from using the Raptor and Hitachi HDD at the same time, + maybe another 10 from adding prime95 to push it to its limits. I don't have access to those measurement equipment but will try to borrow a multimeter from a friend some time. I currently only have softwares ways of measuring the power, could I rely on the PowerChute software (bundled with the APC Back-UPS RS800) to give me precise figures ? How accurate can you expect the power output reported by the UPS to be ? __________________ HJT - 5 steps against malware. Post your HijackThis log there and not here ! Posting system specs Last edited by justpassingby : 05-10-2007 at 05:26 PM.

05-10-2007, 09:31 PM	#9 (permalink)
Kalim Roaming To Help Join Date: Nov 2006 Posts: 5,609 OS: Many	1: That data is not all conclusive. Like it's been covered many times before, on different test benches you have different figures. The maximum you have will be different in most cases to others' maximum. People tend to take them all conclusive and that's very wrong. For instance the G80GTX has been shown as pulling 145W on other bench testing. 2: XFX usually make the most powerful GPUs. The stock nV cards do not pull less or equal to the AIB OC'd card. The OC'd version will pull more, quite easily upto 5-20 watts more on that. 3: IDK about your APC software TBH, maybe it is accurate depending on if it has a voltage/current measure built inside and the transmission is clean, which it usually will. But I have come across something similar before in a study where they found some discrepancies. I seriously do not recall what those faults were, whether minor or large, I don't think they were large at all, something to do with the accuracy affected by communication between the two. 4: A PSU has been given a value of "typical load" for a reason. That typical load is at 50% load, as MFGs expect the users of any PSU to be running most of the time. Much more is harsher on the PSU and less than 25% is usually also bad and not as efficient nor clean. The best spot would be roughly 35-65% loads. It's no different to most electronics. The harder, faster, hotter and longer you make it work, the quicker it degrades to less quality and dies out.

05-11-2007, 04:46 AM	#10 (permalink)
linderman Moderator, Hardware Team Join Date: May 2005 Posts: 14,055 OS: win 2000 pro / Win XP pro My System	Re: How to test the power requirements of a given system ? if you were to buy a 700 watt unit and your power draw (amp load) is not in the "sweet spot" for efficiency / it will not hurt the PS its just a plain ole poor choice for the system >>>> why waste elec power, it can add up $$$$ over a three year period of time also dont be fooled into thinking a generic 430 watt unit like say (Apevia, Aspire) would power your system without any problems for a three year period of time you now have the "right" size unit and a quality built unit for the most part I have found the APC software to be rather accurate, maybe not down to the last 2-3 watts, but do we really need to be that accurate ?????? I havent found a recently built APC unit that was wrong by much when compared to more elaborate equipment in reality I think we easily have a 25 watt + or - minus forgiveness factor for what we are doing and evaluating __________________ If you cant be polite and friendly; then please be silent. Last edited by linderman : 05-11-2007 at 04:49 AM.