d5 DOA Pump?

[evil_noodles]

Good afternoon everyone, 

 

I've been a super long time lurker but I've now the need to ask a question, so here I am all registered.

 

I've been slowly building my SFF hardline water cooled system for a good 5 months now - the final parts arrived this weekend so I've slowly and methodically been building it up. This afternoon, I finished the loop and am ready to get this shindig underway but after priming the reservoir with 1/2 bottle of cooler-water (dye hasn't been added yet) I flicked the PSU switch to engage the pump to draw the water around the loop - nothing happens. There's no buzzing in the pump, there's no bubbles and no vibrations.

 

I'm running a PrimoChill Rez / Pump combo (D5) and it's never been used before (it was pre-constructed bought from PrimoChill, so I've not incorrectly put it together). The pump has a 4-pin PWM and a molex, both of which are connected. I watched a bazillion videos around this, and I can't see what I'm doing wrong.

 

I have two 6-pin to molex adapters, and have confirmed that they can power a spare fan I have lying around. I have currently tried the following configuration mixtures;

 

Connected to Molex1 / Fan connected to CPU_Fan - silence.

Connected to Molex1 / Fan connected to Molex-3pin_Fan - silence.

Connected to Molex1 / Fan connected to CPU_Fan via splitter, the other fan is working, the pump is not.

Connected to Molex2 / Fan connected to CPU_Fan - silence.

Connected to Molex2 / Fan connected to Molex-3pin_Fan - silence.

Connected to Molex2 / Fan connected to CPU_Fan via splitter, the other fan is working, the pump is not.

 

Do you think this is a power issue with the pump? I have read around that the pump should engage even without a fan input. Or do you think that the pump is dead?

 

I have connected the whole system up (powered on the MoBo and got to BIOs very quickly and very short span, the fan outputs are reading N/A so I'm guessing it is power related...

 

Happy to provide more information, picture, the works. Any thoughts?

 

Cheers

Nik

 

[W-L]

11 minutes ago, evil_noodles said:

-SNIP-

Just to make sure you are using the paperclip method to jump the PSU so it provides power? The PSU will not provide any power if the 24 pins isn't plugged into a board or has the pins jumped to think it's connected.

http://www.silverstonetek.com/downloads/QA/PSU/PSU-Paper Clip-EN.pdf

 

[For Science!]

33 minutes ago, evil_noodles said:

-snip-

Lots of pictures can help, but I'm just going through your post now

[evil_noodles]

Thanks guys, yeah, I wasn't sure if the paperclip was working either (darn idle PSU fans!) so I connected some 120mm fans up to the second molex and they spin fine.

 

Since then I've plugged the MoBo in, and that lights up too so it's not a faulty PSU by the looks of it. I'll nab some pictures and add to the post when I can (away for work for a few days). If it is the pump, are all D5's the same? I had this shipped from the US and it took ages to arrive, and cost a bomb on import fees so I'd like to source a replacement thats more local if we can't figure something out.

[For Science!]

4 hours ago, evil_noodles said:

-snip-

Ok I've digested your post now, and would request for more information:

 

- Which power supply are you using?

- What exact 6-pin to molex adapter are you using, and perhaps more importantly, why? Are there no inbuilt molexes from the PSU?

 

This reminds me a previous post that I sorted out where the user had a completely crazy 6-pin adapter which was meant to be a molex --> PCie, but he was trying to use it the otherway around and the pin layout was all wrong.

 

 

[evil_noodles]

2 minutes ago, For Science! said:

Ok I've digested your post now, and would request for more information:

 

- Which power supply are you using?

- What exact 6-pin to molex adapter are you using, and perhaps more importantly, why? Are there no inbuilt molexes from the PSU?

 

This reminds me a previous post that I sorted out where the user had a completely crazy 6-pin adapter which was meant to be a molex --> PCie, but he was trying to use it the otherway around and the pin layout was all wrong.

 

 

I'm using a Corsair SF600, I have tried both the included PSU > molex cables and my cablemod PSU > molex cables and the outcomes the same (I also wondered if it was a incorrectly wired custom cable).

 

The PSU is completely modular - the adapter is a Noctua 'Molex to 3pin' fan extension and was to ensure the MoBo wasn't causing the problem (as at the time I didn't have the MoBo connected, which I now do). Experiencing the same thing no matter what combination I try. I've also read that the fan doesn't actually need to be connected, and the pump should still produce 50% efficiency, but this clearly isn't happening either.

 

I have read your post, that is some next level paint skills, and I can confirm that I am using the 'PERIPHERAL & SATA' connection point on the PSU. Really wish I had a multi-meter right about now!

[Brett_Bst]

could you post some pics of the connections ? is there any was to test the pump without the adapter ?

[For Science!]

26 minutes ago, evil_noodles said:

-snip-

I don't know how feasible it is given you pump is sleeved, but can you just peek in to make sure that the cable on the outermost part of the molex is a red cable, and that the middle one is black. I have seen on a different company that these cables were put in the wrong way.

 

I would recommend using the standard molex cable that came with the PSU or the cablemod one and not use the 6-pin adaptors.

 

I would recommend not powering the motherboard since your cooling is not working yet. So stick to the 24-pin jumper and an unpowered PWM (50% should be fine to fill the system).

 

So at this point apart from all the photos you have of you system, these in particular can be useful:

 

-Overview of your loop and system

-Connection of the pump to the molex adaptor / molex cable

-Zoom in on the molex end of the pump connector, in particular if you can, the color of the cable

-A view where the jumper in the 24-pin cable is clearly visible

 

[evil_noodles]

3, 2, 1 - I'm back in the room.

 

Okay - so I can confirm that I have this as an overview:

 

Reservoir / D5 Pump Outlet > 240mm Radiator > GTX 1080ti > 92mm Radiator > CPU Monoblock > Reservoir

 

From my PSU, I connect the pump molex to the PSU molex output (there are no adapters).

 

I have shorted the 24pin with the correct pins, the PSU does start but the pump does nothing. 

 

It's also worth noting that I have only ever started the system when the reservoir is in the upright position so that it does not run dry. The PMW from the pump is currently just hanging loose out the top of the system.

 

[Brett_Bst]

your 100% sure the molex is powering  the pump and that the jumper wire is in the correct postions, some are the 3+4 pin and some are the 4+5 pin on the 24 pin connecetion

[Brett_Bst]

nvm I have seen a connection in colour and looks correct

[Brett_Bst]

I think you should contact the people you got the pump from they might have some ideas or sort out aa new one as it seems like that's the issue, if you have no other way to test the pump, a wall socket for a phone or somethink that runs the same volts can be used to try power the pump, no one near you with a power supply ?

[Brett_Bst]

looks nice though

[evil_noodles]

Thanks for the ideas. I've emailed the company but they have yet to respond to me (apparently they are known to be terrible at post-sale support) but until then I'm just going to try a few things out. Like on the weekend purchase a multimeter and see whats actually coming out of the PSU, and see if I can get at the pump without taking the whole system apart (although, I will if I have to) there could be a simple loose connection internally.

 

Not entirely sure I want to plug anything other than a PSU cable into trying to power the pump in case it does more damage than good. And I'm confident that the PSU is working.

 

I actually started this project as something to do, as I moved to a new country - I know absolutely nobody else that games here! In hindsight probably not the best life choice, heh!

[Brett_Bst]

7 minutes ago, evil_noodles said:

Thanks for the ideas. I've emailed the company but they have yet to respond to me (apparently they are known to be terrible at post-sale support) but until then I'm just going to try a few things out. Like on the weekend purchase a multimeter and see whats actually coming out of the PSU, and see if I can get at the pump without taking the whole system apart (although, I will if I have to) there could be a simple loose connection internally.

 

Not entirely sure I want to plug anything other than a PSU cable into trying to power the pump in case it does more damage than good. And I'm confident that the PSU is working.

 

I actually started this project as something to do, as I moved to a new country - I know absolutely nobody else that games here! In hindsight probably not the best life choice, heh!

well even though no friends yet they will come juust takes a bit of time, were are you living now trhen ?

[For Science!]

6 minutes ago, evil_noodles said:

-snip-

I couldn't spot any glaring errors apart from the fact that the connector for the 12V on the pump cable looks slightly more "out" than what I have in my mind, but may be significant.

 

From your description, since the PSU is jumpered correctly and there is no reason to believe the molex output is dead, the natural next step is to dissassemble your pump/res combo if it doesn't scare you too much and it doesn't void warranty (it shouldnt, but you know your terms better than me)

[Brett_Bst]

1 minute ago, For Science! said:

I couldn't spot any glaring errors apart from the fact that the connector for the 12V on the pump cable looks slightly more "out" than what I have in my mind, but may be significant.

 

From your description, since the PSU is jumpered correctly and there is no reason to believe the molex output is dead, the natural next step is to dissassemble your pump/res combo if it doesn't scare you too much and it doesn't void warranty (it shouldnt, but you know your terms better than me)

yh get a multi meter to test its getting power etc

[evil_noodles]

I'm living in The Netherlands now, moved from the UK back in the summer for work - and to be fair it's been on the whole an incredibly positive experience (outside of this damn PC!).

 

Looks like this is a job for the weekend. I'll take the blasted thing out and see what I can do in getting pictures to you guys. Once I have it out do you have any particular tests you think I should run?

[Brett_Bst]

well I would frist test the molex connection is sending power to the pump first

[evil_noodles]

I've got the second output on the same molex cable powering a fan when it's on.

[For Science!]

5 minutes ago, evil_noodles said:

-snip-

Yeah, I wouldn't go crazy dissassembling the D5 pump itself, its a fairly stand-alone unit unless you want to break open the shell which I definitely do not recommend. I would only go as far as to taking it out of the reservoir shroud and making sure there aren't any pinched cables and that the power supply cables are actually making contact with the molex connectors.

 

Something you could do to entertain me would be with a tweezer or nose plier to push the red cable into the plastic molex shroud a bit more to see if there could be a loose connection. Obviously do it with the power supply off, and disconnected from the connector to not electrocute yourself.

[Brett_Bst]

1 minute ago, evil_noodles said:

I've got the second output on the same molex cable powering a fan when it's on.

sol its safe to assume that the pump is getting power atleast to the molex connection

[evil_noodles]

Yeah that's what I was thinking. The cables coming out of the pump are at a pretty hard right angle, although overall they're very straight everywhere else. I'm wondering if that has possibly pulled the wires of the PCB.

[evil_noodles]

So the reservoir pump combo from PrimoChill is a CTR. Found a thread from 2014 from someone experiencing the same issue as I am and it turns out that the ring keeping the pump attached to the reservoir was overtightened, stopping the impeller with the attachment pressure. Tomorrow, when I have sunlight on my side I'm going to give the ring a twist to see if I can unscrew it. If I can't move it, then I believe that's probably the cause. 

 

Crossed fingers everyone!  

[evil_noodles]

So.... positive news! After managing to drain the system, without taking the whole thing to pieces, I was able to undo (with great difficulty I might add!) the D5 pump compression ring. After completely removing the ring, displacing the pump from the reservoir very slightly and giving the system a split second of power (I know how dangerous it is to run them dry) to my surprise the whine of a dry pump filled the room. It was not dead, the ring was waaaaay overtightened and stopping the impeller spinning. 

 

I replaced the pump, and replaced the o-ring and compression ring to a 'hand-tight' level, filled the reservoir with some coolant water and flicked the switch, it burst into life sucking all the water through. So I've currently just finished the priming of the loop. I've doused the case in tissue and will leave it running to see if I have any leaks.

 

Thanks for all the input guys, and I'll be sure to post the finished item when I get everything sorted!