Uneven temp shenanigans

Uncle Petey · 26 Aug 2014 at 22:02

So I just moved my main build (in sig) over to a test bench with a MO-RA 1080.

First of all, I ran the cards in parallel (using tubing/fittings) which provided even temps within a degree of each other.
After a marathon session of bleeding the blocks by turning the bench every way possible, I found that while temps were even, they were a lot higher than they were when running in serial in my case with 360/240. They were also a fair bit higher than when I ran 3x290 in parallel (using a triple FC bridge) with the same 1080 rad.

So I then drained it, ran them in serial and spent a while making sure they were bled. There's still a small bubble, but whatever

Now I'm seeing the second card running at up to 10°C higher than the first, when on my old setup there was around 3-6°C difference.

Apart from a possibly unseen air pocket, what in blazes do you think is going on here?

Pneumonic · 26 Aug 2014 at 22:19

what's the flow rate like? could be the MO-RA 1080 is much more restrictive than the two rads you were using?

Uncle Petey · 26 Aug 2014 at 22:25

Flow rate's fine as far as I know.
The MO-RA didin't seem to have any flow issues when I was running 3 cards in parallel or when I was filling/bleeding the current setup.

I do have a couple of spare D5's though. I suppose I could rig up another in serial if need be.

pigeonguyuk · 26 Aug 2014 at 23:53

might sound crazy but anything is possible.

could the flow rate be too high. fluid moving to fast through the radiator meaning less heat dissipation?

Uncle Petey · 27 Aug 2014 at 00:17

Nah, the pump is only set to about 3

I've had it up to 5 just to help clear a few bubbles.

It seems that at stock clocks/volts, the difference is around 4-6°C.
The more I OC, the bigger the difference: Up to ~10°C
The blocks are now totally bled and it's still the same (voltage/usage is the same on both cards btw)

I've built /rebuilt a lot of loops with various cards and not seen this kind of nonsense until now...

I think I'll grab a dual parallel bridge seeing as the best results that I've seen came from the triple version and perhaps flow was dodgy with using tubing for some reason.
Hopefully it'll do the same for just two cards and I'm not just clutching at straws.

Uncle Petey · 2 Sep 2014 at 02:09

LOL!

So I redid the trifire thing with the parallel bridge, and now the first two cards are bang on even, but the third card is now the one that likes to be 10°C higher than the others

It's set up exactly the same as it was when I last had the three cards on there and they were all even back then... I haz bin trolled

ED: I actually saw a +19°C gap in the third card at one point :eek:

Turning the bench on it's side and orienting the cards horizontally seems to have sorted it, with all three cards now within a degree or two of each other.
Not ideal, but hey...