Overclock.net banner

Pump fried and began smoking after new PSU

3K views 27 replies 8 participants last post by  Skyl3r 
#1 ·
Hello All,

Just a few minutes ago I got a new PSU (1000w EVGA Gold) and hooked it up into my PC exactly as my old PSU was hooked up.
I pressed the power button and nothing happened. Pressed it again, nothing. Switched it off and on. Nothing.

I pulled the motherboard cable and jumped it and the power supply fan started spinning like it was on. A few minutes later I noticed a terrible smell from inside my case. I thought it was the power supply, but I was worried it might have been some of my other hardware, so I pulled the new PSU and put in my old one. I turned it on and immediately began watching smoke come out of my pump.

What could I have done to get my pump to smoke?
I'm thinking since I laid the case down (and the reservoir on its side) that it may have got air in it? But could that cause this kind of catastrophic failure?
Since I don't have a working cooling solution, I haven't yet tested if the rest of my hardware works.

I bought the pump (MCP35x) about a week ago.
I'm really trying to identify if I did something stupid or if the PSU or pump had a preexisting issue.
 
See less See more
#3 ·
Quote:
Originally Posted by cmpxchg8b View Post

Did the pump make any suspicious noise when it was powered on? Or any noise at all, the normal running sound?
I did notice a loud click. I thought it was the PSU coming on, but it may have been the pump sparking. Besides that, I didn't hear anything too abnormal. Actually, the pump never sounded like it was running at all.

EDIT:

Also, I'll add that I had been using my PC with this configuration for about a week without noticing any weird noises. It was working fine up to turning on the new configuration.
 
#4 ·
how much air is in your reservoir? There's really only 3 things that I can think of that happened. Air in the pump, a short in the pump, which is unlikely since you said you've been using it for a week, or the PSU is putting out too much power to the molex cable.
 
#5 ·
I would suspect either you plugged the molex cable into the wrong socket on the PSU causing it to deliver too much power. Or there is a fault with the PSU. I'm assuming you swapped all the cables to the new PSU's before turning it on?
 
#6 ·
Quote:
Originally Posted by l187l View Post

how much air is in your reservoir? There's really only 3 things that I can think of that happened. Air in the pump, a short in the pump, which is unlikely since you said you've been using it for a week, or the PSU is putting out too much power to the molex cable.
Well, it's a 250mm reservoir and there's about an inch of air at the top. So roughly 10% of the reservoir is air.
Quote:
Originally Posted by TGBM View Post

I would suspect either you plugged the molex cable into the wrong socket on the PSU causing it to deliver too much power. Or there is a fault with the PSU. I'm assuming you swapped all the cables to the new PSU's before turning it on?
I tested the perif cable with a multimeter and with the molex facing me, rounded corners up, the left two wires were putting out 5v and ground. I believe this is correct.
I can't be 100% sure that I plugged it into the right place the first time, as I've moved everything around a fair amount but I'm near positive that I plugged it into the PERIF port on the PSU every time.
 
#8 ·
Quote:
Originally Posted by TGBM View Post

And did you use the cable from the new or old PSU?
I used the cable that was from the old PSU (EVGA Supernova 850 GQ)
 
#10 ·
Quote:
Originally Posted by TGBM View Post

I would suspect that's the cause. You should always make sure it's the right cables for the right PSU I've fried HDDs before from using the wrong SATA power cables.
Ah, well, that sucks. I certainly hope I didn't fry anything else in the process.
I didn't even think twice about using the same cables. Well, I'll replace the pump and use the right cables and see what I get.

Thanks for the support
smile.gif
 
#11 ·
Quote:
Originally Posted by Skyl3r View Post

Ah, well, that sucks. I certainly hope I didn't fry anything else in the process.
I didn't even think twice about using the same cables. Well, I'll replace the pump and use the right cables and see what I get.

Thanks for the support
smile.gif
Well if you have an air cooler for the CPU you can do a quick post test to make sure that it is all okay I have found that normally its a peripheral that goes like my HDD's, much more of a ball ache than a pump, or in your case a pump. Otherwise I could just pickup a new pump as you know that is dead for sure and try and run it with either of the PSU's but make sure that all the cables are for the correct PSU.
 
#12 ·
Quote:
Originally Posted by Skyl3r View Post

Ah, well, that sucks. I certainly hope I didn't fry anything else in the process.
I didn't even think twice about using the same cables. Well, I'll replace the pump and use the right cables and see what I get.

Thanks for the support
smile.gif
Even changing PSUs in the same series I always run all new cables. Pinouts on PSUs aren't standardized so, not to scare you, but you could have fried everything in the system simultaneously. If you insist on using the same cables, triple check the voltages from the cable. My advice, just switch cables, it isn't that big a deal and it's at least guaranteed pinouts.
 
#13 ·
If my memory serves me correctly, EVGA power supply pinouts did change between power supplies and my room mate ran in to this when he upgrade EVGA power supplies and realized he couldn't use hiw first sleeved set of cables.
 
#14 ·
Quote:
Originally Posted by MightEMatt View Post

Even changing PSUs in the same series I always run all new cables. Pinouts on PSUs aren't standardized so, not to scare you, but you could have fried everything in the system simultaneously. If you insist on using the same cables, triple check the voltages from the cable. My advice, just switch cables, it isn't that big a deal and it's at least guaranteed pinouts.
I am a bit concerned. I would test on air, but that'd require me pulling a waterblock off one of my GPUs and putting the AIO unit back on (Fury X). I'll just wait for a new pump and make sure I do it right this time.
 
#15 ·
Quote:
Originally Posted by Skyl3r View Post

I am a bit concerned. I would test on air, but that'd require me pulling a waterblock off one of my GPUs and putting the AIO unit back on (Fury X). I'll just wait for a new pump and make sure I do it right this time.
You could if you are really nervous just see if it posts to a screen without having the pump running. Its not recommended but can be done if you don't have it running for more than 30s and could give you an idea if it is just the pump that died. If you have a motherboard with a post code readout like the EVGA boards you can easily see if it doesn't post what the cause is.
 
#16 ·
Quote:
Originally Posted by TGBM View Post

You could if you are really nervous just see if it posts to a screen without having the pump running. Its not recommended but can be done if you don't have it running for more than 30s and could give you an idea if it is just the pump that died. If you have a motherboard with a post code readout like the EVGA boards you can easily see if it doesn't post what the cause is.
The system in question is the one in my signature. It'll give me flashing LED's to indicate what the issue is but no display.
I did test by connecting my old PSU back up and try to turn on my PC. All my fans came on and the lights on my motherboard and such, but the computer shut itself off after a second. I think this is either a CPU overheating thing or a no CPU fan thing. I'll have to check when I get home.
 
#17 ·
It wouldn't instantly shut down from CPU overheat if you have a block with coolant in. I have run a system for about 5 mins with just a block on it with two tubed with some coolant in it so I could update the BIOS as I didn't want to rebuild the loop just for the BIOS update so I would say it defiantly isn't overheating if its just instantly shutting down. It should output to display unless there is something wrong with the hardware even with no fan connected to the CPU fan header it would output that to the display to tell you that.
 
#18 ·
Quote:
Originally Posted by TGBM View Post

It wouldn't instantly shut down from CPU overheat if you have a block with coolant in. I have run a system for about 5 mins with just a block on it with two tubed with some coolant in it so I could update the BIOS as I didn't want to rebuild the loop just for the BIOS update so I would say it defiantly isn't overheating if its just instantly shutting down. It should output to display unless there is something wrong with the hardware even with no fan connected to the CPU fan header it would output that to the display to tell you that.
I experienced the same symptoms when filling up my loop initially. The pump didn't seem to be pumping. I had it connected to a second PSU to fill the system with. I read that I might need the PWM connector on the pump plugged into the motherboard for it to work, so I connected it to my motherboard and booted up. The system would turn on for a second, all the fans would come on then go back off. I didn't see anything on the display. The computer is coming on long enough for it to be getting to the BIOS before shutting off. It's not immediate.
Once I finally got the pump to kick on I stopped seeing this problem.

So, I'm thinking now the pump is fried, this is likely the same thing.

Just saying, I agree with what you're saying. I'd expect it to post and I wouldn't expect it to overheat that fast. I don't know why this happens, but it happened before then worked 100% fine for a week. Now the pump doesn't work it's happening again.
 
#19 ·
To be very honest with you, the first thing i'd do first is to tear apart the PSU and inspect if any visible damage were done.
Then figure out what the difference of the pinouts of two PSU to learn how bad it was damaged.

Then try to test your mainboard. Most of the time components in your system would have safety protection when there's wrong power delivered.
Your pump doesnt have anything like that....
 
#20 ·
Quote:
Originally Posted by Skyl3r View Post

I tested the perif cable with a multimeter and with the molex facing me, rounded corners up, the left two wires were putting out 5v and ground. I believe this is correct.
The two black wires in the middle are ground. Red wire on the right is +5 V. Yellow wire on the left is +12 V.



I hope that's what you're reading with multimeter; if not, you're not using the right cables.
 
#21 ·
Quote:
Originally Posted by cmpxchg8b View Post

The two black wires in the middle are ground. Red wire on the right is +5 V. Yellow wire on the left is +12 V.



I hope that's what you're reading with multimeter; if not, you're not using the right cables.
This is sort of confusing me.

The two wires that were coming from the pump to it's molex connector were black and yellow and the pins they connected to from the PSU were the left two with the rounded edges up and they were reading 5v and GND. My understanding at the moment is this is wrong. Could too little voltage cause smoking though?

EDIT:
Okay, yeah 5v was being supplied to what is supposed to be the 12v line.
 
#22 ·
I can't see how undervolting a pump would cause it to burn... I never tried that though. Anyhow if you've seen smoke coming out of the pump, it's likely to be damaged beyond repair. I would just double check the PSU and all the wiring before connecting the replacement pump.
 
#23 ·
Quote:
Originally Posted by cmpxchg8b View Post

I can't see how undervolting a pump would cause it to burn... I never tried that though. Anyhow if you've seen smoke coming out of the pump, it's likely to be damaged beyond repair. I would just double check the PSU and all the wiring before connecting the replacement pump.
Haha, absolutely.

Since I'll be disassembling my loop anyways, I'm thinking of moving on to the 900D to support both of my PSUs. I've heard this case is particularly well geared towards watercooling (and dual PSUs
biggrin.gif
)
 
#24 ·
Now that I've verified results, I just wanted to post back.
I got a new pump and hooked it up with the correct wires and it worked as expected. At that point I was able to determine that my SSD was also fried.

So, little warning. Make sure to use the wires for your PSU. That was a $200 mistake
frown.gif
 
#25 ·
Quote:
Originally Posted by Skyl3r View Post

Now that I've verified results, I just wanted to post back.
I got a new pump and hooked it up with the correct wires and it worked as expected. At that point I was able to determine that my SSD was also fried.

So, little warning. Make sure to use the wires for your PSU. That was a $200 mistake
frown.gif
Glad to hear that it was not everything that had died but sucks that your SSD went did you lose any data?
 
#26 ·
its an honest mistake.

I tried to help a buddy out in a similar fashion by letting him use one of my extra psu cables to hook up his hard drives. Fried 2 hard drives. Luckily one was brand new and able to exchange but the other had a bunch of old data on it that he lost. I felt terrible. I was even about to buy an exact model hard drive off ebay to see if i could swap the controllers and get it to work but he said not to worry about it.

But it really should be more specified.

Do not swap around modular power supply cables between different power supplies.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top