How do I test a Water Pump from a Corsair H110 Closed loop water cooling kit.

sir_bacharach

Reputable
Feb 9, 2015
34
0
4,530
Hi,

I bought a cable from Ebay to power my H110 water pump as well as the rear fan for a constant maximum speed rather than mother board variable fran control, I did it so I could have variable speed fans on the side of the case separate from the rear fan and pump which I always want to be on max.

When I plugged it in I tested it to make sure the fan span that was plugged directly into the power supply. It did not, so I checked the connections, unplugged them then plugged them back in again and tried again. Still nothing happened. I turned off the PC again and went to check the wires, the one that was connected to the water pump was hot, I guessed there was a short somewhere.

So I unplug the new cable and plug the fans back into the motherboard where they were in the first place. I switch on and the PC turns on for a about 4 or 5 seconds and then turns off before even reaching the bios screen.

I turn off and try again, it does exactly the same thing. wondering if the fan or pump were damaged I remove them from the board in case it's causing a failsafe. I try booting the computer again but the same happens, 5 seconds and off. I hope that perhaps the processor is just shutting down the computer due to a temperature failsafe due to a non running pump.

So I try one last time to plug the pump in and restart the computer. 4 seconds in and I hear a large pop and a flash of light comes from the board just below the pump roughly where the wire goes into the pump case. I wasn't sure if it came from the board or the actual pump, but I'm guessing it was the pump. The pump now smells of burnt electrical components.

Is there any way I can test with a multimeter to make sure the pump is definitely faulty, should there always be a certain resistance between some of the connectors on the pump/fan?

I'm scared now that my PC is also damaged, but I guess I won't know until I get a new CPU fan to test.

I also don't know if the pump was already faulty or if the cable caused the fault, it did seem that the new cable was hot next to where the pump plugged into it.

Can anyone help suggest what to do next?

Thanks.

Rob.
 
Solution

On the diode range, the meter attempts to pass a small current in the order of 1mA with a voltage limit typically around 3V and the reported value would be this voltage at the given test current.

The usefulness of this test is that semiconductors do not start conducting until some threshold voltage. For the bipolar transistors used in many brushless fans, that threshold is around 0.6Vbefore anything starts to happen. On the ohm...

InvalidError

Titan
Moderator
If you saw a flash coming from somewhere near one of the pump wires, I would guess the wire fused. The pump motor may have seized or stopped working for some reason, drew excessive current, heated up the wires, then the wires shorted and blew up.

If you use a meter on the ohms range, the wires will likely read either, shorted, open or some unusual high value.

Since the problems started with your aftermarket cable, you might want to check that the cable is wired correctly. Maybe its polarity was inverted and you ended up blowing the fan's brushless motor driver, same for the pump if it has an electronics-driven motor.
 

sir_bacharach

Reputable
Feb 9, 2015
34
0
4,530


Yeah, I thought the wire must have something to do with it since it was all working fine before.

I can see the new wire seems to have melted where it plugged into the pump, it was hot to touch when I tested, it seemed like it was shorting there. but the spark definitely came from near the pump where the wire goes into it.

I used a multi meter set for resistance and tested two prongs on 2 of my working fans, then the same two prongs on the pump and I get absolutely no resistance value. Would this prove that the pump is now damaged?
 

InvalidError

Titan
Moderator

That depends on the sort if motor is being used. Brushless DC motors use transistors and those would read open if you apply less than 0.6V or so to them. You might want to try the diode testing range too, see if the circuit will pass some current at somewhere under 3V or whatever the max diode testing voltage your meter can do. This might still not be enough to get a reading though.

But after a presumably quite noticeable flash if it sparked your attention, which implies a rather high current arc, my bet would be on burnt open due to some length of metal getting vaporized.
 

sir_bacharach

Reputable
Feb 9, 2015
34
0
4,530


Since the flash was outside the actual pump, I'm guessing something inside has arced out onto the motherboard. Do you think there is a chance the board is still OK? It's an ASUS M5A99X EVO R2.0 I hear some boards have voltage protection in etc. The PSU is a Corsair HX 750i too, apparently that has over voltage protection also. One of the settings on my multi meter has a picture of a Diode and what looks like a wifi signal next to the setting, it doesn't mention a voltage though, when I use that setting I can get readings on both working fans but not on the pump still, not quite proficient with the multi meter to be honest.
 

InvalidError

Titan
Moderator

On the diode range, the meter attempts to pass a small current in the order of 1mA with a voltage limit typically around 3V and the reported value would be this voltage at the given test current.

The usefulness of this test is that semiconductors do not start conducting until some threshold voltage. For the bipolar transistors used in many brushless fans, that threshold is around 0.6Vbefore anything starts to happen. On the ohm range, the test voltage is often in the neighborhood of 0.3V to avoid biasing transistors and diodes, so using the ohm range to test semiconductor will often read open, depending on the meter.

If you get open circuit / off-scale on the diode and resistor range, the wire or something in whatever it is attached to is very likely blown open.
 
Solution