• 20.08.2025, 05:41
  • Registrieren
  • Anmelden
  • Sie sind nicht angemeldet.

 

GTXJackBauer

Full Member

Dienstag, 29. November 2016, 10:53

My apologies. Woke up not too long ago. You're correct, parallel splits the flow like my GPUs are set to even out the temps versus serial.

t1nm4n

Junior Member

Dienstag, 29. November 2016, 11:07

glad you said that, I was looking everywhere to make sure I hadn't shoved my smelly shoes in mouth, or is that foot in mouth ;(

InfoSeeker

Senior Member

Dienstag, 29. November 2016, 14:43

I did what you are talking about with a different product, it works great, and I'd prefer parallel flow over serial as serial can build pressure and parallel you get the same temp water cooling your gpus. Here's my shoddy workmanship, but it works and you can't hardly notice the bad job from a distance :)

Pipes in Series
   The pressure loss is the sum of the individual losses:
      dp = dp1 + dp2 + .. + dpn (1)
      where
      p = total pressure loss (Pa, psi)
      p1..n = individual pressure loss (Pa, psi)
   The mass flow rate is the same in all pipes:
      m = m1 = m2 = .... = mn (1b)
      where
      m = mass flow (kg/s, lb/s)

Pipes in Parallel
   The pressure loss is the same in all pipes:
      dp = dp1 = dp2 = .... = dpn
   The total mass flow is the sum of the flow in each pipe:
      m = m1 + m2 + .. + mn (2b)

Edit: reference

Personally, I prefer series, and having full pressure on all water blocks.
Mostly in case a bit of debris is present in the coolant, it has a better chance to be pushed through the micro channels of the block.
I also prefer having full flow over the channels, not a big preference, but no reason not to.

And lastly, in parallel, if you do get some blockage on one of the blocks micro channels, the water distribution will be uneven between the blocks.

Dieser Beitrag wurde bereits 4 mal editiert, zuletzt von »InfoSeeker« (29. November 2016, 14:51)

GTXJackBauer

Full Member

Dienstag, 29. November 2016, 15:12

That's why I made sure to always have a flow meter for the loop as it's saved me in the past. :thumbsup:

InfoSeeker

Senior Member

Dienstag, 29. November 2016, 15:25

That's why I made sure to always have a flow meter for the loop as it's saved me in the past. :thumbsup:

Yes, a flow meter is very useful, but it will not tell you if 1 of 2 cards in parallel has blockage... unless you have a flow meter on each card.

GTXJackBauer

Full Member

Dienstag, 29. November 2016, 15:48

That's why I made sure to always have a flow meter for the loop as it's saved me in the past. :thumbsup:

Yes, a flow meter is very useful, but it will not tell you if 1 of 2 cards in parallel has blockage... unless you have a flow meter on each card.

Any blockage will show through the flow meter as well as pin point to a specific block based on it's performance in temps. ;)

InfoSeeker

Senior Member

Dienstag, 29. November 2016, 16:08

That's why I made sure to always have a flow meter for the loop as it's saved me in the past. :thumbsup:

Yes, a flow meter is very useful, but it will not tell you if 1 of 2 cards in parallel has blockage... unless you have a flow meter on each card.

Any blockage will show through the flow meter as well as pin point to a specific block based on it's performance in temps. ;)

Temperature readings certainly can indicated underperformance, but a flow sensor will not provide any indication, as the flow decrease in the blocked block, will be increased proportionally in the clear block. Total flow will remain the same.

t1nm4n

Junior Member

Dienstag, 29. November 2016, 20:15

what he was saying is that the blocked block will have higher temps and therefore the temps will point you to which block has the blockage, but if your gonna take your loop apart you might as well go to town and take them all apart, it makes sense to me at least, clean one, clean them all.
But really it's too each their own, I've proven to myself that parallel is the way to go with my setup and current configuration, plus I think it looks better, on top of having multiple mounting points it looks more secure, might just be a mind trick but I think it's more secure so I feel better (and that's all that really matters).

GTXJackBauer

Full Member

Dienstag, 29. November 2016, 22:08

what he was saying is that the blocked block will have higher temps and therefore the temps will point you to which block has the blockage


This. :thumbup:

It happened to me and I knew exactly what the issue was. Of course I did a complete tear down and cleaned what was needed to be cleaned but fortunately only the CPU block needed cleaning. Basically overtime in a 6 month span, I was losing about .1-.2 GPM of flow every 1-2 months till it got down to .2 GPM at full load and the CPU temps were too high for my liking, mind you the CPU temps were rising overtime as well while the GPU wasn't at the rate as the CPU was since the blockage will increase the temps to the rest of the loop a bit but not as obvious to the block it really was affecting. It ended up being a piece of tubing I must've missed when cutting the tubes, melted on the block and blocked the inlet side of the block pretty well.

While I did know something was wrong from the start thanks to the flow meter, I was too lazy to get started and figured, I'd do it at the right time. Well, there it was. lol

Dieser Beitrag wurde bereits 1 mal editiert, zuletzt von »GTXJackBauer« (29. November 2016, 22:09)