My mcp655 died... and it took my 4850 with it :(

[drizzt5]

How exactly doesn't it spin smoothly? Does it get blocked? Or is it hard to make it rotate?

When I spin it, it seems to get a little stuck on a part of the rotation every time I rotate it. I have never done this when it was 100% working, but I assume it should be a completely smooth spin.

Kind of hard to tell by picture, but is there any moisture in there?

There appears to be a few dark brown areas on the PCB, those could be heat indicators, but I'm having a hard time finding any pictures to compare for reference.

No moisture that I can see. I'd like to add that if I do plug the pump in it probably will still DO something... but last time I did that with it in my loop it made vibration noises and didn't pump any water.

[Martinm210]

When I spin it, it seems to get a little stuck on a part of the rotation every time I rotate it. I have never done this when it was 100% working, but I assume it should be a completely smooth spin.

It should spin like a really strong fan would. The magnets will make it hard to turn and sort of pull to each winding position, but it shouldn't make any noises when you do that.

If you put a drop of oil on the white bearing, you could probably do a quick test in the open to see how it spins. A second under power shouldn't cause any harm, but if the bearing is bad it could work itself loose and take off on you..

I'd try it, but I like it when stuff takes off like that. Just be careful...

EDIT: Wear a face shield, I take no responsibility for my suggestions!!...

[drizzt5]

Guys I am posting this from the emergency room. I lost an eye....














jk ill give it a try soon.

[Martinm210]

Guys I am posting this from the emergency room. I lost an eye....













jk ill give it a try soon.

Darn...You got me with that for a second...no fair!

Make sure you wear a face shield..

[sunnyville]

It does look like there's moisture in there - there are drops of liquid on the inside of the plastic housing by the looks of it - that could have caused a short and then the thing seized up. Just my guess. Did you use the standard housing or a custom top (e.g. EK or something)

[Quagmire LXIX]

drizzt5, sorry to hear about your card/pump (and your eye ) and I have nothing for your past, but I have a recommendation for your future.

I know most watercoolers frown on flow meters due to restriction, but I use them for this exact reason. I hook up the Koolance ones to my fan controller which will put out a loud alarm when there's no flow (achilles heal is if nobody is around to hear it), but you can also hook these into some mobo fan headers that will shut the pc down, you know, the right way .

It's a shame about the 4850, but view it as time to upgrade.

Good Luck.

[NaeKuh]

looks like the pcb got fryed..

overvoltage somehow though a power spike or a brown spike.

You can tell by that toasted resistor.
Usually the result when it gets too much voltage, or too little voltage.
that is my initial guess.

Meaning u had a bad pcb.

It should spin like a really strong fan would. The magnets will make it hard to turn and sort of pull to each winding position, but it shouldn't make any noises when you do that.

i think his pump is dead bud... he zapped that resistor good.

[drizzt5]

It does look like there's moisture in there - there are drops of liquid on the inside of the plastic housing by the looks of it - that could have caused a short and then the thing seized up. Just my guess. Did you use the standard housing or a custom top (e.g. EK or something)

I didn't modify the pump at all. Just the standard pump out of the box the whole time I was using it with tubes and worm clamps.

It's a shame about the 4850, but view it as time to upgrade.

Tbh the 4850 @800/1000 pretty much covered all my gaming needs. -.-

Yea I guess in retrospect a flow meter could have helped but I never thought that a component would over heat so quickly. My temps were good that whole day and up till that point
(idle at around 32c's and load at maybe 50c's if its really intense... I always load up HWmonitor or core temp to check them out).
The pump must of called it quits in the middle of the game or something cause after I took my headphones off I heard the subtle vibration noise.


So I plugged the pump in. The white ball doesn't appear to be spinning... but makes the vibration noise when I have it all assembled. I did this for no more then 10 seconds and then immediately turned it off.
So it is still doing something... it is not completely dead, although obviously not functioning.

I don't know how water would of gotten in there either (if it is even really in there. I don't see any)... I run a tight wc setup in my case and I always run a leak test and I've got my worm clamps securely applied on some thick tubing.
Still waiting for a reply from swiftech.

p.s. it would be embarrassing if I lost an eye after joking about it...

[Waterlogged]

If the pump still runs, it's not the electronics, and the electronics certainly wouldn't cause a vibration.

Here's my theory...somehow, air got into the loop and vapor locked the pump causing either the bearing (white ball) or the mating part of the impeller to over heat and deform. Considering the bearing is ceramic, I'm guessing that isn't the weak link, more likely the mating piece in the impeller. The bearing likely transferred it's heat into the surrounding stainless steel as did the mating part until the stainless could no longer shed the heat it was getting and that's when failure occurred.

[drizzt5]

If the pump still runs, it's not the electronics, and the electronics certainly wouldn't cause a vibration.

Here's my theory...somehow, air got into the loop and vapor locked the pump causing either the bearing (white ball) or the mating part of the impeller to over heat and deform. Considering the bearing is ceramic, I'm guessing that isn't the weak link, more likely the mating piece in the impeller. The bearing likely transferred it's heat into the surrounding stainless steel as did the mating part until the stainless could no longer shed the heat it was getting and that's when failure occurred.

Sounds logical enough to me...
I guess I am at the mercy of swiftech.

[3dchipset]

Don't those pumps need to be completely flooded with water to work? I thought any air can kill the pumps?

[drizzt5]

Don't those pumps need to be completely flooded with water to work? I thought any air can kill the pumps?

If they are on for long periods of time ya, but sometimes there is gonna be little pockets of air when your bleeding the loop right?

[NaeKuh]

Don't those pumps need to be completely flooded with water to work? I thought any air can kill the pumps?

just the chamber.

Inline pumps can not self prime.
And priming without water can cause chamber damage.

But this is a electronic damage.

[Waterlogged]

just the chamber.

Inline pumps can not self prime.
And priming without water can cause chamber damage.

But this is a electronic damage.

Please tell me how electronics causes vibration? Pump still runs as well so it can't be electronics.

[drizzt5]

just the chamber.

Inline pumps can not self prime.
And priming without water can cause chamber damage.

But this is a electronic damage.

I think it might be physical damage and not electronic. Once again it still does something when plugged in.

Swiftech still hasn't responded. Should I contact them again?

[gabe]

I think it might be physical damage and not electronic. Once again it still does something when plugged in.

Swiftech still hasn't responded. Should I contact them again?

Email received SAt 18, @ 1:56pm
Copy of my answer today:

Hi Dillon,

Read your post on XS.

Customer Service was temporarily unmanned Saturday, and we are closed Sunday.

What is you exact date of purchase? You warranty is 2 years from purchase date. We’d need copy of your invoice.

Best regards,


Gabe

[drizzt5]

Email received SAt 18, @ 1:56pm
Copy of my answer today:

Hi Dillon,

Read your post on XS.

Customer Service was temporarily unmanned Saturday, and we are closed Sunday.

What is you exact date of purchase? You warranty is 2 years from purchase date. We’d need copy of your invoice.

Best regards,


Gabe

Thx for the quick reply!
I sent you a reply via email.

[NaeKuh]

Please tell me how electronics causes vibration? Pump still runs as well so it can't be electronics.

huh im completely lost in what your asking.

He looks like he fryed a mosfet, so it could be a power trigger for the magnetic motor.

So only half his cycle could be firing which is causing the impeller not to spin.

This is all assumption..

And the pump can cause vibration if there is something or was something stuck in the impeller.

If the pump is run dry, friction from the ceramic bering and sidewalls can also cause vibration.

If the impeller was defective and and off balance, it would cause vibration while spinning at 2k+ rpms.


But he toasted a Mosfet along with another chip, and part of the pcb.
What more is there to say?
That usually means electrical damage. :P

[Waterlogged]

huh im completely lost in what your asking.

He looks like he fryed a mosfet, so it could be a power trigger for the magnetic motor.

So only half his cycle could be firing which is causing the impeller not to spin.

This is all assumption..

And the pump can cause vibration if there is something or was something stuck in the impeller.

If the pump is run dry, friction from the ceramic bering and sidewalls can also cause vibration.

If the impeller was defective and and off balance, it would cause vibration while spinning at 2k+ rpms.


But he toasted a Mosfet along with another chip, and part of the pcb.
What more is there to say?
That usually means electrical damage. :P

I've seen similar scorching on DDC PCB's that are brand new and they run just fine.

Look at the first pic again and tell me that impeller stud and bearing looks normal, this is airlock/mechanical/heat related.

[drizzt5]

I'm still talking to the swiftech guy via email. He is very helpful.

But I think my foray into watercooling has reached an end for now :/ Gotta work on getting a replacement gpu first.