Need help with PC Power & Cooling PS

[RacerX]

Just got in this 510wt turbo cool PS. I have opened and examined the inside. I have figured which pot is the 3.3 and 5 &12 rail. I have maxed the 3.3 pot and still only get 3.31 according to MBM. I find this alittle low for my liking. Anyways there are 4 more pots lined up in a row and I have no idea what they do.

Can anyone explain what the 4 other pots do? As of right now I not very happy with this PS and will return it if I can't get my 3.3 rail higher without modding.

I've noticed my 3.3 rail when running 3dmark at 235fsb takes a beating and goes as low as 3.25 making my system unstable.

thanks

[saaya]

are you sure that 3.25 on the 3.3v rail makes your system unstable???

even if the extra pots help you, id ask them to replace the psu.

[STEvil]

My brothers 230 watt ATX on his XP1800+ @ 1200mhz (darn psu!) runs the 3.3v at 3.75v, lol 8-)

[RacerX]

Originally posted by saaya
are you sure that 3.25 on the 3.3v rail makes your system unstable???

even if the extra pots help you, id ask them to replace the psu.

Its been well documented here that in able to run high FSB DC on a NF2 board especially on the NF7 its very important to have a good solid 3.3 rail. Anything above 230fsb running 3d apps kills my 3.3 rail and lowers it to 3.25 or so.

I'm trying to get my 3.3 rail to 3.4 so I can run 3d apps around 237-240 or so.

[WxChaser]

I'm sure the PSU is fine, you must be adjusting the wrong pots. Call Bob Roark at PC Power & Cooling tomorrow and ask him which ones you should adjust, he'll be very cool about it. I have the same exact power supply and it's rock solid, best I've ever owned. I have the 3 external pots however.

Edit - measure your outputs with a multimeter from a molex on the PSU, not with MBM5. You can jumper the PSU to start up without being connected to the motherboard.

Good luck.

[Dissolved]

yea software temps arent always the best reading..

but the problem with testing the psu not connected to the pc is there is no load..

but email them or call them.. they answered my email within a day.. very nice guys there

[Kunaak]

can anyone with a PCpower and cooling PS tell me the max voltages you can get with your PSU?
all rails ;)

[Major]

Here ya go:

[Major]

Overvolting your rails offers no advantage ! the adjustments on the PC Power & Cooling are designed to adjust the rails to the proper voltage, I doubt you will be able to adjust them more than +/- 5%

The voltages reported by winbond monitoring on mb's are BS, you have to measure with a VMM on the mb to see what your rails are actually like. If you measure on the molex connectors you will see that your rails are probably high, there is a typically a small voltage drop across the atx connector that the sense lines cannot correct for since they are on the psu side of the connector.

[Vatt]

Exactly what Major Slaughter said....
I also have the PC Power&Cooling Turbo Cool 510 Watt and MBM reports 3.3 as 3.3 with the pot completly maxed out.
However when measuring with a digital multimeter it showed my 3.3 as 3.45.

So just try and measure at a connector , that is way more accurate than what MBM is saying.

I have 3 motherboards tested with this PCPower&Cooling and all 3 report different voltages even when I have not changed the pots.
They range from 3.2 upto 3.45 when my digital multimeter says 3.33 exactly.

The same applies to the 5v and 12v rails.

[az'rail]

my PC P&C 510 was also under volting the 3.3 rail. at idle it was at like 3.28. adjusted the pot and it when up to a mobo killing 3.29......:( well sending it back was out of the question because i live so far away. so i did some psu surgery. i changed out the 3.3v rail pot (5kpot) for a 10k pot. got a safe distance away and fired her up, after i poked at it with a stick:stick: and found it was good to go i slowly adjusted the pot and go her up to 3.38v. now she is good to go. 233fsb rock stable

[xigfrid]

to Major Slaughter and Vatt

Don't be so rude with ppl, i know that with older boards MBM's or any software monitoring can reports wrong values, but with NF7-S rev2.0, the values are more than accurate

I measured them directly onto the board, if MBM said 3.23V, it is exactly 3.23V with voltmeter on the Motherboard volt check point

I must congrate az-rail for modding such an expensive PSU ;)

regards.
Xigfrid

[saaya]

the nf7-s voltages are not accurate, i just read an article about how the traces from the points where the voltage is meassured to the winbond chip is interupted with some small resistors wich makes the readings not very accurate.

i think the only way to find out what your rails are like is to meassure them on the running board. if your fire up the psu without any load the rails youll get diferent readings than under load.

[xigfrid]

any link bro ? I would be glad to read more about NF7 components, trying to find a new thing to Vmod on this §§§§ that won't do 250Mhz ;)

[saaya]

i think it was @ lostcircuits.com ... not sure

[xigfrid]

Originally posted by saaya
i think it was @ lostcircuits.com ... not sure

yeah found it here (http://www.lostcircuits.com/motherboard/abit_nf7-s12/3.shtml). It is a rev1.2 review from march.

Btw, I did measure the V on the mosfets and not on the sensor, i confirm what i said about MBM+NF7-rev2.0

I wouldn't trust lostcircuits review, it is too old and some improvements have been realised since march ;)

quote:"though, that it is hard to justify anything beyond a real voltage of 2.75 V for both VDD and VDDQ. Contrary to popular belief, there is neither any gain in operating frequency nor in performance at higher voltages beyond 2.8V unless some really old components are used" :stick:

[RacerX]

Ok guys got a new NF7-S 2.0 in today. I always like to keep a back up.:)

This one is better than the first for sure not that my other is bad. This is what the board does right out of the box no mods peroid. Vdd set to 1.7 in bios and 2.9 on the vdimm, cas2,2,3,5. http://service.futuremark.com/compare?2k1=6929972

Note I also measured my 3.3 rail while 3dmark was running. It fluctuated from 3.33 to 3.27, this was done with a meter. Measured the middle leg of the mofset right next to dimm 1. MBM is reporting the same also.

I'm not complaining but I have a feeling my 3.3 rail at 3.4 and this board will do 245 no problems.

I didn't stop there either. The real test came when doing alittle Nascar 2003 racing. Normally this game will crap out way before 3dmark does. Ran 100 laps at Bristol without any issues.

Gotta love Abit.

[RacerX]

http://home.comcast.net/~phippy1/NF7.bmp

The highest I can go without mods. Sorry for the low bandwidth but any multiplier under 10 Sandra reports it incorrectly. Done at cas2,2,3,5

Gonna put my vdd mod on and see what it will do. I have a feeling my TwinMos PC3200 bh-5 chips needs more than 3.0v to make anything over 240 stable.

[Nico]

Originally posted by az'rail
my PC P&C 510 was also under volting the 3.3 rail. at idle it was at like 3.28. adjusted the pot and it when up to a mobo killing 3.29......:( well sending it back was out of the question because i live so far away. so i did some psu surgery. i changed out the 3.3v rail pot (5kpot) for a 10k pot. got a safe distance away and fired her up, after i poked at it with a stick:stick: and found it was good to go i slowly adjusted the pot and go her up to 3.38v. now she is good to go. 233fsb rock stable

I'm keen to replace the VR on mine also, does anybody know what wattage VR they are using? I've found some Nidec VR's that will fit and are 500mw so I think that would do?

[Evil_Spork]

my PCP&C425 is great. i assum eits similar to teh 510 in alot of ways.

my 3.3 is solid as a freight train! and if i turn the pot up all teh way i can get something like 4.1v on my 3.3!!!!!!!!!! :eek: :explode:

wish u luck

[Nico]

Originally posted by Evil_Spork
my PCP&C425 is great. i assum eits similar to teh 510 in alot of ways.

my 3.3 is solid as a freight train! and if i turn the pot up all teh way i can get something like 4.1v on my 3.3!!!!!!!!!! :eek: :explode:

wish u luck

Wow!:eek: , that's high. i'd be happy with a top end around 3.5 - 3.6 just to give me some room to move.
By the way, does yours have the external VR's to achieve that?

[Ragnarok]

just void the warranty and cut the sense line.. and do the mod

my TC550 gives up to 3.8V without tripping, and I ran 3.5 everyday on the NF7-S 2.0...

[Penti]

Originally posted by xigfrid

but with NF7-S rev2.0, the values are more than accurate


For me it measures fairly true, better then most boards. But the winbond-chip doesn't report the same Voltages as on the board and there are no reason to trust a motherboard anyway.


Anyway as major says. No point measuring at the molex, the voltage drop over the atx-connector can be hugh enough to make a difference. I guy i know that i helped out had over 0.4V voltage drop on all rails at the atx-connector, becuse of bad contact ;) Probably becuse it's bad psu too though. But i have some own experiance, with my own stuff with voltagedrop over the atx-connector to..

[Nico]

I have taken out the 3.3v VR, it was 500ohm in my case. Should have some replacement VR's tommorrow and I start with 1k and see where I get too:) fingers crossed.

[Nico]

The 1k VR is in and I can go to 3.6v and it's still got more to spare :D . I set it at 3.45v and I put it back in when I get home from work.

[esoteradactyl]

sweet :) you wouldnt happen to have any pics would you?

[Nico]

Originally posted by esoteradactyl
sweet :) you wouldnt happen to have any pics would you?

There's not much to see :p: as it looks almost untouched (slightly different color VR), but I'll take a pic for you ;)

[Nico]

It's all working fine, it dropped a bit once the 3.3v rail was loaded, but I adjusted it upto 3.4 in bios which was 3.45 at the ATX connector. Have gone as high as 3.45 in bios/ 3.5 at the ATX.

[Nico]

Another angle!, the darker blue VR is the new 1k for the 3.3v rail.

[pelikan]

These boards must vary in their voltage readings. My NF7 v2 does not read accurately at all through MBM5. I have taken readings with a DMM off the board.

[esoteradactyl]

i got too anxious and just did the mod using the sense line. good job though :)

[bachus_anonym]

is that 1k ohm pot rated at 0.75W (just like the ones used for vmods for mobo) or 0.1W just like the one here (http://www.radioshack.com/product.asp?catalog%5Fname=CTLG&category%5Fname=CTLG%5F011%5F002%5F011%5F000&product%5Fid=271%2D280) from RadioShack? would that really matter ?

EDIT: oh, i found this pot on their website - TOCOS (http://www.tocos.com/thruholesealedsingleturn.htm). i'll order some of those and 3.3v rail - here i come! :-)

[Nico]

It's rated at 0.5w, and was the only type readily available. I use 0.25w for motherboard stuff, but I think 0.1w might be a little low although it would probably work ok.

Go for the Tocos GF063X 1K MODEL

[bachus_anonym]

great, i found it here (http://www.tocos.com/thruhole_pdf/GF063.pdf), page 66... they have whatever one wishes... thanx man :toast:

one more question, when u have your 3.3v rail set to 3.45v what does the middle leg of the mosfet read? does it still fluctuate ?

in my case i have 3.3v rail maxed at 3.32v (with other nf7-s v2.0 it was 3.39 - too bad it's dead now :-( ), middle leg fluctuates from 3.31 to 3.22 idle and to 3.15 under prime... that's why whenever i set vdimm to higher than 3v prime95 generates error, even at 217FSB which is stable at 3v...
seems that it's true that vdimm can't be higher than middle leg reading under load (high voltage) minus 0.15v.

thanx again for help...

[Nico]

Haven't check it to be honest, but it could explain why anything over 2.9vdimm use to be worse stability, not better :D

[`schr0et]

You guys should be very careful when messing with your rails.

NEVER, EVER rely on your windows based PC health probe to give you accurate readings of your lines!

Always check thru mosfets/molex connectors to make sure.

[bachus_anonym]

u r absolutely right, PRanganathan... DMM is your best friend and any serious overclcoker should have one in his "toolbox".
i'm using both MBM and DMM for monitoring but multimeter is what i trust the most...

[Nico]

My bios reads near spot on what I get of the Mosfet on my video card.