Low 3.3v rail

hi guys,
my pc has been having trouble running cs source so i did a bit of troubleshooting as i know my pc should run that game with ease but it doesn't,
after a long time searching i found my 3.3v rail drops to as low as 2.9v during heavy cpu and ram usage which i thought is very strange as i have a decent psu.
what uses the 3.3v rail? is it only the ram these days? my psu is only a year old but i cant rma it cos i opened it so i could make the fans run full speed :( lost 2 years warranty cos of my stupidity
not running o/c or anything.
my case allows for dual psu's and has an adapter to connect them both, would it be feasable to take the 3.3v line from another psu and use that?

The 3.3 volt line getting to low will cause plenty of headache. While it's possible in some cases to tie in a different PSU to work in tandem with the current one? I would NOT recommend it.
Most likely cause? Flacky ATX power connector on the motherboard/power ATX connector, or weak PSU. You can clean up the connectors by just unplugging and plugging 'em back in about 10 times or so (being carefull of course not to allow the mobo to flex to much)
If you want to isolate this down more, a general test can be made with a digital multimeter if you have the skills. I'd settle on cleaning up the connector first and seeing if that improves matters any.
S-

Similar Messages

Curiously low 3.3v rail?

According to Everest (and speedfan) my 3.3v rail never budges past 3.07. Is that anything to be concerened about? Its been that way since I put the system together over a month ago and I havent had any stability problems associated with that AFAIK. But it does strike me as odd so I thought Id finally get around to inquiring about it...
FWIW the other rails seem steady enough. 12v is almost contantly @ 11.98 idle or under load while 5v hovers around 4.92.

Quote from: Liver Kick on 20-June-05, 03:35:58
Thats two questions. And the answer would be no. I take it I ought to do some research first, will do.
Well my mistake instead of a comma I should have write the word 'or'... well...
Quote from: Liver Kick on 20-June-05, 03:35:58
If my rails really are 3.07 (Everest) or 11.67 (Bios) do you think I would have noticed some stability issues by now? Or could it be a ticking time bomb type of deal.
As I stated earlier, some posts a go...
Quote from: TheBigMan on 19-June-05, 07:30:19
... but if the system is stable don't worry about it...
Be well....

Low voltage radings? read this 1st

I was concened about a problem with the rig in my sig, I had a low 12 v rail of 11.43
so the brand new shiny 600w PS I just bought , or so I thought, was at fault, so I pop in my old trusty Antec 550 True control , same problem.
humm what is up? I bring home from work my good old trusty DVOM and check voltages on the molex connectors with the system powered up, now I get 12.35 and the 5V rail was 4.99
so before you run out and get another PS best check it 1st , might save you some bucks!
I emailed MSI about the issue , so we will see how that goes

this issue goes for most motherboard out there.
its partly because they use cheap parts.
a cpy temp can easily be 20celcius off.
also some programs shows wrong from time to time so always check in bios.
and 11.4 is borderline.
5% -+ on 3.3v 5v and 12v rails is within acceptable.

Why do I see -28C??

For awhile now I have been questioning this motherboard (865PE Neo2 Platinum) because of the low+3.3v rail readings, crappy 3D graphics, and now I see a reading of -28C on speedfan for "Temp 1". I get the negative reading whether I am in Vista or XP. Can anyone tell me what this is and why it's reading in the negaitves?
Thanks,

Quote from: Exo on 19-April-07, 04:20:23
It's an unconnected sensor input on the monitor IC - Disregard it.
Temp 2 = CPU and temp 3 = chipset probably, thats the only readings there are.
About the low 3.3V line - Don't blindly trust speedfan, it does not support every type of sensor chip, and onboard sensor chips themselves are highly inaccurate. It's far more likely a wrong reading then the line beeing really low.
The only real way to know is to measure with a DMM.
Temp 3 appears to be my system temp...
I'm not blindly trusting SpeedFan. Hmonitor says the same thing and although my BIOS doesn't show nearly as low, it is still lower than it should be. I'm not sure where to test with a multi-meter.

To all user

Hi. Merry Christmas & Happy New Year.
Excuse me, can I know which one is true & 100 % correct for comparison PSU.
1) Example for my Enlight 420 watt PSU:
+5V= 40A; +3.3V= 30A; +12V= 18A
-5V= 0.3A; -12V= 0.8A; +5VSB=2A
Use the Formula :[(5V*40A)+(3.3V*30A)/3]*2
=[(299)/3]*2
= 99.667*2
=199.334 ##
2) Manufacturer label +3.3V;+5V combined as 220 watt
3) I) PSU more HEAVY
II) GOOD quality blower Or cooling fan
III) Connection cable (Label)
Recommended Min
ATX 16AWG 18AWG
AUX Power 18AWG 18AWG
AUX 12V 18AWG 18AWG
HDD/FDD 18AWG 20AWG
One question ,anyone using ICUTE (8310 SL-AL)casing come with 420 watt ,it is ok????
For Example (1) & (2), NOT 100% for me and some of cheap psu label with not corrcect. how about u??????? Please give me a reason .
IF above is wrong , I say sorry for yours, and help me make a correct please.
Thanks!!!!

Those are estimation, to guide you to estimate your PSU.
As in the place you live, most cheaper casings are bundled with el-cheapo PSUs. Those PSUs eventhough "stickered" as 420W, they have very low and fluctuating rails, ie 5V only around 4.7V, 3.3V @ 3.1V , and over-volted 12V @ 12.40V (if you use P4, it would drive 12V down to less than 10V!!), and without any surge protection feature. So it's expected the lifespan of these PSUs are relatively very short compare to the branded.
It is not always the PSU label put it wrongly or over-rated, but the component quality can hardly make the PSU to deliver as it is rated. There are some users happily live with 420W the brand you mentioned with a power-hunger rig, so we can't argue if it doesn't deliver 420W. But one cannot deny is the days are counting.

K8N Neo2 Plat power and bios question

I have a K8N Neo2 Plat. I use a UltraProducts 500W xconnect PSU and a BFG 6800 GT OC.
A couple weeks ago the 6800 died. I contacted BFG and they sent me a new one.
The new one died the same way after a week. They said to check the power levels on the +12 rail.
In windows using softare tests it reports 13.58
in the bios it reports 12.22
both are over the allowed variance.
Ultra products told me to check with my motherboard manufacturer to see if there was a BIOS setting that would let me set the power on the +12V rail. Is that possible?
Is that something my K8N Neo2 Plat can do?

No, you cannot lower the 12V-rail by a bios setting. I really don't see what they mean by that.
Your motherboard does, however, regulate the power that comes from the power supply. Within certain margins it can lower the voltage and keep it steady. Your videocard can do the same. Now, depending on the tolerance (quality) of your hardware, you can vary on those margins. But on any piece of hardware, there is a point on which it cannot correct/lower a voltage that is too high. It then simply shuts down or dies.
To my opinion, the PSU-maker gave you a very funny piece of advice. What he actually says is that their psu may give too high a voltage (I talk about the 13+ now), but your hardware should correct this. That's not the way it works: they should make psu's that give 12V where it should. So, I'd demand a replacement. But, before you do: use a multimeter to measure the 12V-rail. You cannot trust the board's measurement too well.
But then, like Stick says, I think that 12.22 is within mentioned margins. 13.58V is too high. So, that's why you should first measure using a multimeter. Measure for a longer time. System idle, system full load. Write everything down and see how steady this psu is. If 12V dances arround, replace it.

12v rails low?

CoreCenter is showing my 12v rail at 11.75 to 11.8. Is that low? I have an Antec True Power 480.

Thats fine

Video Card unstable, PSU the cause? Low V rails in BIOS

Ive recently ran into some Video card trouble, I started getting artifacts all of a sudden and random 5 second pauses. I rebooted, the artifacts went away but the pauses persisted before I finally got a BSoD with a nv4_dsp Ininite Loop error. Not good. At first I thought it was my temps because they have been higher than normal for some reason. Normally Im barely above 70c under load but lately its been around 77c all the way up to 80c for no reason I can discern. I figured maybe I knocked the heatsink around (VF700) by accident while cleaning and should reapply it.
But then I noticed my voltage ratings. In Everest and Speedfan the 12v rail is reported @ 11.31 and the 3.3v rail is 3.06. I checked the BIOS reading and that says the 12v rail is fluctuating between 11.55 and 11.61, and the 3.3v rail was down to 2.96. I dont have a multimeter or really know how to use one, for the experts out there how much of an indication are the BIOS and Windows readings to go by that something is really wrong? This PSU is supposed to be among the best out there (PC Power and Cooling have a great rep) but I cant think of what else would cause the problem besides by card going belly up (and the card is still pretty new). If the PSU is truly a dud are the rest of my components in any danger operating at these low voltages?

Another possibility that could lower voltages is the amp demand being too high in the system. Now why this would happen now and not before could be due to the PSU slowly dying, or something new added to the system, or in the worst case scenario, a hardware component breaking down and causing some sort of short... this would be noticed by an increased in heat, but don't panic, just because your temp readings on the card have been higher doesn't mean that's the case, your card could be hotter for other reasons (like maybe the thermal grease shifted or something and isn't conducting heat as much as it used to).
With voltages going low, you might face system instability anyway, so this should be a concern. I think it's most likely the PSU, best way to check is borrow one from another computer that you know is in great shape (and sufficient) and try it out, if the readings are much better, you got yourself an almost definite culprit.

PSU putting low voltages out on 3.3v rail

I had this problem for a while and always thought it was PSU's problem, now got a new and more powerful PSU and still same probblem.
The PSU only putting 3.10v~3.15v out on 3.3v rail, the pins 1,2 and 11 on the ATX got burned, everything else seem working fine. this is my third PSU for last 8 weeks. please help me.

Quote
do you mean RMA both PSU and MOBO?
YES = i would,if nothing else just to have the connectors replaced,if not replaced the prob WILL return.
If there were no foreign material present on the connectors ! ,then a NOT fully locked connector caused the prob to began with.
=owner / operator / assembler prob.

MSI GTX560ti Hawk low 3dmark11 score

I recently purchased MSI GTX560ti Hawk. I ran 3dmark11 on it and on performance mode, I only got P1650 score. The graphics score is very low and the FPS averaged around 6-7.
Here is my specs:
Asus P5Q motherboard
Gigabyte Odin GT 550w PSU
Intel Q9550 oc@ 3.67GHz
Kingston DDR2 2x2GB @ 800MHz
Intel X25-M 160GB SSD
MSI GTX560ti Hawk
I know the package box suggested using a minimum 600W PSU but I managed to boot into windows and play games with my 550W. I can run Assassin Creed Brotherhood and Shogun 2 on very high graphic setting.
Could the very low 3dmark11 graphics score due to the low PSU power supply? Please help
Thanks!

Quote from: Kuane on 21-May-11, 01:48:18
I just bumped the core voltage up block by block from +0 to +150 on Afterburner 2.2.0 beta 3.
I tested Heaven DX11 benchmark on all of those voltages and there were artifacts in all of them (running VGA default clocks). But there were less artifacts as the core voltage went up.
Is my PSU not supplying enough power?
Quote
Is my PSU not supplying enough power?
could be , but basically should be more than enough if the stats can be trusted.
Test the VGA in another machine[who has a better PSU, with strong single rail] then see how its goes there
Quote
I just bumped the core voltage up block by block from +0 to +150 on Afterburner 2.2.0 beta 3.
I tested Heaven DX11 benchmark on all of those voltages and there were artifacts in all of them (running VGA default clocks). But there were less artifacts as the core voltage went up.
try next voltage step and retest

EVO:RAIL - Use Cases?

I think as soon as the world becomes aware of EVO:RAIL then folks will start asking what the use cases are. It's clear that there are couple of immediate use cases, but I'm betting that customers will bring to the table applications that perhaps we hadn't considered. The other night I was our Tech Summit event (its private event for VMware SE/TAMs to attend prior to VMworld to get them up to speed on what the announcements are all about), and I was approached by a number of SE/TAMs who are interested in running EVO:RAIL in the basements and garages as the foundation to their homelabs. Even with the onsite of Cloud - and we have an internal cloud that SE/TAMs can use demo products to customers - just like the VMware Community they still have a passion for building out their own demos, just how they like them - that they can tailor make to handle customer questions that come up. I'm the same, despite being at VMware for 2+ years, I'm still running my own gear at home. It just worked out more cost effective form my needs than keeping my 42U rack at a colo.
I doubt that my might the "WAF" would apply to having a 4U box with 4 nodes inside it - in the spare bedroom. So I'm already looking at running EVO:RAIL in a nested configuration on my home lab gear for testing, videos and demos to customers. That work has already been done for the VMworld HoL, by William Lam (the undisputed king of nested ESX, well, perhaps Simon Gallagher comes close!) - who's done some excellent work on this already. So hopefully it won't take much work to get hold of the bits from the HoL, and pump the lastest build of EVO:RAIL into it...
Anyway, I digress use cases. It's clear that ROBO, SMB and VDI are the top use cases - situations where an appliance like model, with zero configuration (almost!) is needed. I think we perhaps be a bit careful with the SMB use case. EVO:RAIL is 4xNodes in a 4U configuration with each node presenting 192GB of RAM. One thing I've always been concerned with is how variable "SMB" is as term. Over here in a the US an SMB would be classed as company with <1000 employees. Over here in Europe, that would be regarded as still quite large for an SMB.. The same goes ROBO. So I guess what I'm saying is size is relative, and so we have to be careful when use terms like "small".
Folks who know me will know that I'm a big fan of VMware Site Recovery Manager and vSphere Replication - having written books about the version 1.0, 40 and 5.0 versions. I think there's use case where an EVO:RAIL is put into some colocation facilitiy and used as DR target. The important thing to remember about EVO:RAIL as its based on vSphere 5.5 U2, other products in the vCloud Suite family will work with it - like SRM, View, vCloud Director or vCAC... So its not some special "moded" version of the platform. Perhaps a better model for DR and EVO:RAIL is to run VR locally on it, and use vCloud Air as the target for recovery...? I guess we have to ask the question what's going to be more popular/easier. Two site with EVO:RAIL at either end, or EVO:RAIL enabled for DR to the cloud. Just to be clear. That isn't baked into EVO:RAIL today, I'm just blue sky thinking about the possibilities in the future.
So what do folks think here? Can you dream up any other scenarios where an EVO:RAIL would be good fit...?

You make some valid points there, but there's a couple of things that are worth clarifying....
Firstly, EVO:RAIL is pre-installed at factory. So really the only "installation" is racking up, and connecting to the TOR. In fact I think an EVO:RAIL is easier to setup that a VSA is, and gives more usable disk space, and lower cost per gigabyte..
IF the customer has given the OEM the IPs, IP Ranges, Hostnames, and Passwords - then it literally is a 15min setup after clicking "Build Appliance". On the other hand if the customer wants to customize the EVO:RAIL prior to build - such as changing the VLANs Tags (That they supplied) that would add a couple of minutes...
vDS is NOT a prequistie. EVO:RAIL uses standard switches... Although an EVO:RAIL is licensed for Enterprize Plus. There's no requirement to configure a Distributed Switches.
The real "asks" here is the customer needs 10Gb switch - I think its fair to say that isn't common in SMB/ROBO environment. I suspect many of the OEMs will be asking the customer to validate their switch configuation. Multicast for IPv4/IPv6 is used by the "Zero Network Configuation" on the management/VSAN networks. As the name suggest no configuration is required on the EVO:RAIL, its used as part of the discovery process for finding the nodes, and finding new appliances. One most switches this tick off box option on the switch for the VLANs that require it...
Anyone of my plans is to order an EVO:RAIL from one of our suppliers just like a customer would, and document/record that process. So i can get a feel for a real world customer experience would be. In fact I might see if my wife, Carmel can set it up... :-)

EVO:RAIL and VSAN

Are there VSAN benchmarks on EVO:RAIL compared with traditional server based VSAN?

Sorry, I'm a bit confused about the question - you mean a comparison between EVO:RAIL and VMware Virtual SAN - or do you mean benchmarks between EVO:RAIL/VSAN - and conventional Storage Array connected over ethernet (iSCIS/NFS) or FC block storage?
For VSAN there's been a number of articles about performance which you might find useful - one caveat. In this release EVO:RAIL supports 16 nodes per cluster whereas Virtual SAN supports 32-nodes per cluster - so some of the sliding scale doesn't apply...
http://blogs.vmware.com/performance/2014/09/microsoft-exchange-server-shows-great-performance-vmware-virtual-san.html
http://blogs.vmware.com/performance/2014/08/web-2-0-applications-vmware-virtual-san.html
http://blogs.vmware.com/performance/2014/03/vdi-performance-benchmarking-vmware-virtual-san-5-5.html
My colleague Duncan Epping has written extensively on some of the considerations around VSAN:
http://www.yellow-bricks.com/2013/11/14/vsan-performance-many-sas-low-capacity-vs-sata-high-capacity/
and there's been some independent work too:
http://www.storagereview.com/vmware_vsan_posts_2_million_iops

EVO:Rail Training

For EVO:Rail to succeed vmware must make sure the availability of training for the same and not depend upon its hardware partners for Training.
Most of the hardware mentioned don't have a great track record in Training.
Vmware can also add a EVO:Rail Training Module in VSAN Deploy and Manage course.
Non Availability of the training was one of the reason that vmware VSA failed to get traction in the market.

You make some valid points there, but there's a couple of things that are worth clarifying....
Firstly, EVO:RAIL is pre-installed at factory. So really the only "installation" is racking up, and connecting to the TOR. In fact I think an EVO:RAIL is easier to setup that a VSA is, and gives more usable disk space, and lower cost per gigabyte..
IF the customer has given the OEM the IPs, IP Ranges, Hostnames, and Passwords - then it literally is a 15min setup after clicking "Build Appliance". On the other hand if the customer wants to customize the EVO:RAIL prior to build - such as changing the VLANs Tags (That they supplied) that would add a couple of minutes...
vDS is NOT a prequistie. EVO:RAIL uses standard switches... Although an EVO:RAIL is licensed for Enterprize Plus. There's no requirement to configure a Distributed Switches.
The real "asks" here is the customer needs 10Gb switch - I think its fair to say that isn't common in SMB/ROBO environment. I suspect many of the OEMs will be asking the customer to validate their switch configuation. Multicast for IPv4/IPv6 is used by the "Zero Network Configuation" on the management/VSAN networks. As the name suggest no configuration is required on the EVO:RAIL, its used as part of the discovery process for finding the nodes, and finding new appliances. One most switches this tick off box option on the switch for the VLANs that require it...
Anyone of my plans is to order an EVO:RAIL from one of our suppliers just like a customer would, and document/record that process. So i can get a feel for a real world customer experience would be. In fact I might see if my wife, Carmel can set it up... :-)

Does drive caddy include rubber rails?

I'm thinking of getting a drive caddy (0A65623) to install a second hard disk in my W520.
I presume the disk needs to be fitted with the small metal tray and rubber bumper rails to fit the caddy.
Are the tray and rubber rails included with the caddy? If not, does anyone know the part number for them.
Thanks
KD

@combthins:
- If you're going to be using your laptop indoors and dust or aesthetics are not a worry, there are bay adaptors on amazon and ebay as low as around 10 bugs, and usually those adaptors have screws in it so you can just place and attach your hdd, no caddy, no rubber rails; But it has the disadvantage that it'll leave some gaps above and below the adaptor which, as mentioned, will affect aesthetics and will allow dust entering into that space.
- those w series laptops are machines that are evaluated to very stringent specifications so they can sustain some abuse working in outdoors thus the original adaptor you're looking for (0A65623) fits in perfectly and has a layer (plastic or something) so as to prevent dust entering and as you can see in the original lenovo description (in the third picture) it shows the hdd being placed with an additional and non included caddy, also you can see the reviewers complaining about the fact that the necessary rubber rails are not included, thus the need for the part number. see in Amazon there is an user video that shows the adaptor and the need of rails.
the required parts are:
for 9.5mm hdd: 41V9756 (description, amazon (caddy indluded))
for 7mm hdd: 04W1640 (description, ebay (caddy included))
::: In conclusion if you want to configure your computer the way it is meant to, go for the original part and grab the caddy and rails. If you want to save money and stay indoors grab a cheap adaptor and don't complain about gaps.
(However you do have the option of buying an newmodeUS caddy for 45+ (around the same price of the original one) and that one won't leave gaps nor will it require anything else (but it is somehow inferior to the original one for about the same price)

My 12v+ rail

My 400w Antec seem to be within, though not much over the guidelines recommended by the experienced users here, and squeaks under the recommended PSU + .3 according to the PSU calculator webbpage linked above, but my 12v+ rail is consistently at 11.7v, and never above 11.8. Is this a problem, seeing as I have a GeForce Ti4600 in here?
K7N2-Delta-L
XP2100+
512MB Crucial PC3200 (2x256)
GeForce4 Ti4600
Antec case/400wPSU
Coupla HDD's
Coupla CDRW
Onboard sound

Quote
Why, is there a problem with its operation?
Well, the other day I had a hardlock, opened the case and the damn HSF had fallen off the GeForce. I had to use thermal epoxy to reattach it, and have had one other hardlock since then, similar to the previous. Was wondering if my card was damaged (possible, no?), or if perhaps my 12v+ was after all, kind of low. It's just always bothered me, so I took the opportunity to ask. Thanks.

Low 3.3v rail

Similar Messages

Maybe you are looking for