• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Max 'hot plugging' power draw capability?

NoxTek

NoxTek

The Geek Redneck
Joined
May 27, 2002
Messages
9,300
I've got a Corsair TX750 that's about 6 months old and has worked great for me sofar, but recently I've run into a problem and I'm looking for advice on whether it's something to be worried about or not. If Redbeard (Corsair rep) still visits it'd be great if he could weigh in...

My main question is basically:

What's the highest amount of added amperage draw that a PSU such as my Corsair TX750 can handle (and still keep a stable +12v rail) when a 12 volt device is connected while the unit is powered on. The reason I ask is that I've got a 12v device that draws 1 amp of power, and about half the time when I connect it to the PSU in my (powered on and running) machine it causes a spontaneous reboot.

Meanwhile I can plug in a Seagate SATA drive in the same manner and everything is just fine. Only difference is the Seagate says it only draws 0.5 amp on the label (unsure if that's at spinup, normal r/w operations, or idle).

It's just kind of hard for me to believe that the TX750 wavers in stability enough to cause a reboot / crash when an amp's worth of load is added to the +12v rail.

The system specs that this issue is happening on is in my signature. Anyone want to weigh in?
 
Are you sure that is the PSU causing the reboot? If it is, it must be the PSU is defective. A healthy 750W PSU provides way enough juice for what you have. Post back if you find out something.
 
Bruce2 said:
Are you sure that is the PSU causing the reboot? If it is, it must be the PSU is defective. A healthy 750W PSU provides way enough juice for what you have. Post back if you find out something.
Click to expand...

+1, Make sure it's the PSU and if it is then it's most certainly defective. It shouldn't do that with just 1 amp. Also you might want to make sure that your not overloading a rail either.
 
Blue Falcon said:
The reason I ask is that I've got a 12v device that draws 1 amp of power, and about half the time when I connect it to the PSU in my (powered on and running) machine it causes a spontaneous reboot.
Click to expand...

What 12V device that draws 1A of power are you connecting?

Uberbob102000 said:
Also you might want to make sure that your not overloading a rail either.
Click to expand...
The Corsair 750TX is a single rail design. So the chances of overloading a rail is pretty damn low.
 
Uberbob102000 said:
+1, Make sure it's the PSU and if it is then it's most certainly defective. It shouldn't do that with just 1 amp. Also you might want to make sure that your not overloading a rail either.
Click to expand...

This is a Corsair TX750 which is a single rail design, so no worries about accidentally overloading a rail. Also the load being applied is actually 1.3 amps and not 1.0 amps like I originally stated. The device in question is actually a NiMH battery charger, a 1-hour quick charge Energizer model that the original wall-wart AC/DC transformer broke on some time ago.

Me being the cheapass and inventive guy I am, I discovered that:

1. Finding a cheap 12v universal wall wart that provides 1.3amps isn't very easy.
2. A PC power supply can happily pour on 12v at 1.3 amps without even breaking a sweat.

So I chopped the end off the original AC/DC adapter and used it to make myself a little molex adapter. I've been using it with an ancient spare Antec PSU I keep on the test bench for like 6 months, but I recently decided to consolidate and clean up my work area, so I've begun using my main rig's PSU instead since it's on most of the time anyway.

Again, I'd think the TX750 would be able to take a transient load increase of 1.3amps on the +12v rail while running without making the system wig out, but I can reproduce it almost every time.
 
So after several hours of troubleshooting I discovered that my system instability and crashing problem wasn't caused by connecting the battery charger at all, it was only a slightly related symptom. I discovered that I could make the system crash not only by plugging in the battery charger, but I could also make it crash by running any 3D program or running Vista's performance check / benchmark. It would black screen and crash reliably at the point where it tests the graphics card for Aero capability.

So I went to troubleshooting.... first I thought it was memory related so I swapped memory around and even put a couple of brand new sticks in. No dice. Then an idea struck me and I decided to check the PCIE power connections on my GTX280 to make sure they were secure... in fact I disconnected them completely and reconnected them, and also swapped the two cables in their connectors on the card. Sure enough, next reboot the system is rock solid even in 3D apps / benches / games.

So then I shut down and swap the cables back around just to make sure, power on, and try to run Vista's performance check. BAM. Fails at the Aero check again. Reboot, try and run Furmark... BAM. Black screen, Dead PC. I then noticed that the power status LED on the GTX280 switched from green to red whenever these crashes occured, signifying that the card wasn't recieving enough power.

So I pulled out my digital multimeter and began testing voltages on the two PCIE connectors. All voltages are reading the same, around 12.27v. I tested all three yellow wires on each of the two PCIE connectors. Then I decided to check the ground connection on each pin from the PSU (10 in total counting the 'extra' 2 ground pins on the PCIE 6+2 connectors). One of the PCIE connectors reads like -320 ohms of resistance between ground (the case interior) and the negative wires. The other reads -280ohms using the same method of measurement except for one of the extra ground wires on the +2 connector, which reads around -800 ohms. Yes, negative ohm readings... I haven't figured out what that's about yet since the only time you usually see that with a DMM is when there is existing current in the lines. (The PSU was unplugged during the ground tests).

So I am officially stumped. I'd suspect the GTX280 except why the hell would it work perfect if the power cables are swapped around? I'd suspect the PSU but why would this problem exist if the +12v reads good and the grounds read (relatively) good. I even shook the cables around while I was taking measurements just to see if there was an intermittent connection problem or something.

This is incredibly weird.

And before someone says it: No I'm not "overloading a rail" - the Corsair TX750 is a single rail PSU meaning it has a single +12v rail with the capability to handle something like 70 amps. :eek: Although I can understand how someone would think that a rail was being overloaded by the symptoms and troubleshooting results.

I need a beer and some tylenol. Ugh.
 
Well I'm not quite the expert, but you measured resistance, not impedance.

Negative ohms are quite possible with op-amps.

Bet you if you measured impedance you would get a positive number. No clue how it would display... 14 + j30?
 
And things continue to get weirder....

I got out a magnifying glass and examined every single part of the PCIE power connectors... there are absolutely no mangled pins, no signs of wear, no indication that the video card should have trouble receiving power from either connector.

But the plot thickens!

I decided to try something different. As some of you may know the GTX280 has two PCIE aux power connectors, one of them being a standard 6-pin and one of them being an 8-pin (2 extra grounds). So I decided to only swap the +2 connector... and what do you know everything works fine.

I am flustered at this point. There is absolutely no good damned reason this should be happening. The +2 connectors on both PCIE power cables are exactly alike and simply piggyback two extra grounds off the already existing grounds in the 6-pin connector. I've measured ohms across these grounds every which way you could possibly do it and they are all good. The only POSSIBLE reason for these symptoms is that the +2 connector on one PCIE connector 'fits' the socket on the GTX280 better than the other, but I swear they are EXACTLY alike - the pins / connectors are not bent in any way. Moreover if it wasn't making full contact in the first place the GTX280 wouldn't even boot, it would display the red power status LED indicating not enough power. For fark's sake, I wonder if the little female 'pins' in the +2 connector are scorched or dirty somehow, not allowing for full contact?

Damn GTX280 has this plastic shell on it so I can't just use my DMM to measure the actual solder points on the board.... grrrrh.

I swear I'm going to have to make a video of this just so someone will understand. It's that wacky...

And the two of you who actually understand what's going on might just say 'So swap the cables and live with it, or better yet try the other two PCIE connectors" (the TX750 has four in total)... but I'd rather fix something now while I know about it instead of ignoring it and having it bite me in the ass again later.
 
You must log in or register to reply here.
Back
Top