Overclock.net › Forums › Case Mods & Cases › Builds & Case Mods › Build Logs › [Build Log] - The Big Budget Boomer Box (aka, the "BBBB")
New Posts  All Forums:Forum Nav:

[Build Log] - The Big Budget Boomer Box (aka, the "BBBB") - Page 328

post #3271 of 3982
Thread Starter 
Quote:
Originally Posted by fantasticgcg View Post

Hi Awesome build!!! biggrin.gif
Could i ask where you ordered those white bitspower fittings from?

Thanks! Got most of mine directly from Bitspower....but FrozenCPU and PPCS both have a pretty big collection now.
post #3272 of 3982
Could you not have just wrapped the outside of the rubber with a paper sleeve so it would slide easily against the metal? Well, that back side is where the holes are at, so it's liable to be the loudest part anyway.

Are you going to combine the wiring (minus the rpm wire) of the two pumps since they run together in series? You'll want the modulation wire from both to be attached to the same AQ5 pwm header. Then someday you could replace the other two pedestal series pumps with pwm and put them onto the other AQ5.
post #3273 of 3982
Quote:
Originally Posted by cpachris View Post

I hope they work for you, but if they have the same LED spacing that the double density Hitlights do.....I think you may have the same problem with those lights also. They look almost identical to Hitlights. I also have the same normal density Hitlight connectors as you do, as well as the Hitlight connectors made for the high density LED strips that I showed earlier. The connectors you have will not work with the high density Hitlight strips....it's because of how deep into the connector the high density strips can (or in this case can't) be pushed. These pictures are of my trying to put the high density strips, into the normal density connectors that you have.
Warning: Spoiler! (Click to show)

Not sure how clear these pictures are, but because there are twice as many LED's on the high density strip...there is MUCH less distance between the contact spots on the strip and the LED. So...when using the high density strips, you can't push the strip far enough down into the connector to make contact with the contact spots. The LED keeps it from going in deeper. There is more distance on the normal density strips....and they slide down further into the connector. The only way the normal density connectors could be used with the high density strips....is either to cut out some of the plastic lid that snaps down on the connector (so that LED clears), or to remove the first LED light on the strip closest to the connector.

I'd be interested to know if you get them to work. Good luck!
I understand now. thumb.gif
Thanks for taking the time to post the pics and fully explaining. + rep for that. Do you happen to have a link to the proper connector?
post #3274 of 3982
Thread Starter 
Quote:
Originally Posted by Electrocutor View Post

Could you not have just wrapped the outside of the rubber with a paper sleeve so it would slide easily against the metal? Well, that back side is where the holes are at, so it's liable to be the loudest part anyway.

Are you going to combine the wiring (minus the rpm wire) of the two pumps since they run together in series? You'll want the modulation wire from both to be attached to the same AQ5 pwm header. Then someday you could replace the other two pedestal series pumps with pwm and put them onto the other AQ5.

Hrm. Paper actually may have been better than vaseline. May give that a shot next times pumps are disconnected.

Right now, I have not combined the PWM wires. Each pump is being controlled by one of the AQ5's. If I ever were to have more PWM pumps...I would probably need to do this. But why replace the two in the pedestal? Why not just add two more to the mix? 6 pumps. smile.gif


Quote:
Originally Posted by mandrix View Post

I understand now. thumb.gif
Thanks for taking the time to post the pics and fully explaining. + rep for that. Do you happen to have a link to the proper connector?

This is the one on the Hitlights website. I did a quick look and couldn't find them on Amazon. But Hitlights ships pretty quickly also.
post #3275 of 3982
Thread Starter 
Backup and running! Running overnight....no leaks at all this time.









Part of the loop layout change was this section down here underneath the from AMS 360 radiator. I've put the two AMS 360 radiators in parallel flow, and cleaned up the loop layout a bit in the process.

Previous Layout (Click to show)

In the old layout picture above, you'll see that the connections coming up from the pedestal (not actually plugged in in the picture above) have a long reach around the reservoir line down to the pedestal. I wish I had the same angle picture...but the new layout does not take up near as much space. The connection (dplug) coming up from the pedestal goes to a T-block and then splits to each of the two AMS 360 radiators. On the PSU side...there is a similar setup for the out ports. They join in a T-block and then head up to the 360 SR-1's on top of the case.





And in this picture, you'll see that all of the pump connections are done on the ports behind the pumps. In the old layout, I used the top port for the final out, and I just never liked the way this looked. Much cleaner look now. Playing with some red/white combination of 3/8 inch adaptor and stop plugs for the top ports. Those with sharp eyes will also see a Bitspower QDC to the right of the pumps. I'm using a Y-Block that splits into a drain line (used with the QDC) and a line into the motherboard subloop. If I need to drain it, I just attach the otherside of the QDC with a tube...and it starts draining.





In this picture, you'll see the Bitspower quick disconnect I'm using for the tubing into the board. I was worried how the black QDC would look with the other white fittings, but I think it kind of disappears into the board. I like it, and it sure makes the motherboard tray easy to disconnect and take out.





...and one more shot of the full layout of the board. This picture gives you a good view of how the new waterblock on the Areca raid card fits into the grand scheme. It goes: VRM>RAM>CPU>RAM>GPU's(parallel)>RAID>Chipset





More results later...but initial results are that moving the AMS rads into a parallel setup increased flow just a little bit for the radiator portion of the loop....but not the motherboard loop. Because of the connection I have between the two reservoirs....the least restrictive subloop (motherboard vs radiators) can get more flow than the other one. The radiator subloop gets 1.4 gpm now (was between 1.2 and 1.3 previously), while the motherboard subloop can get up to 0.9 gpm.

Based on what I had seen previously with the CPU subloop and the GPU subloop....I think it's the Aquacomputer Aquagrafx 690 blocks that are the most restrictive part of the loop. I already have those two in parallel. Temps are great....so I'm not too worried about increasing flow even more....but there is a part of me that wants to. devil-smiley-019.gif Other than moving the RAM blocks to parallel....any other ideas on increasing flow in the motherboard subloop?
post #3276 of 3982
Awesome!
I need to read this whole thread soon biggrin.gif
post #3277 of 3982
Quote:
any other ideas on increasing flow in the motherboard subloop
RAM blocks in parallel

AQ5 blocks in parallel

RAID card and chipset in parallel (if their restriction is very similar)

If your VRM block is very restrictive (doubtful), you could add a partial bypass valve.

Add a valve between the two reservoirs so you can have all four pumps working in series if you desire it.

It looks like the bent clear tubing under the AMS rad is too tight of a corner and is pinching the tube?

The least restriction for pumps is to go in the front center and out the side edge (in an L shape); I'm not sure how much this would change, but that would remove 3 right angles from your current setup I think.

How restrictive is your flow meter? You could create a straight bypass using a full ID valve if it would help when not taking measurements.

Probably the most restriction is being caused by how many G1/4 connections, angles, and fittings you are going through (each is small, but there are a lot to add up); but I'm not sure you can do much about that without sacrificing aesthetics.

Use a pair of MCP35X or an Iwaki.
post #3278 of 3982
Thread Starter 
Quote:
Originally Posted by Electrocutor View Post


Add a valve between the two reservoirs so you can have all four pumps working in series if you desire it.

This idea intrigues me the most. I've seen how severely flow diverts to the path of least resistance when given a choice. That's why I'm leery of running different block in parallel when I don't know how similar the restriction is. But I'm wondering if having this connection between the two reservoirs is kind of creating that situation for me. The radiator subloop clearly has less resistance, and I'm getting up to 1.4 gpm through there. But...flow stays closer to 0.9 gpm through the motherboard subloop....because the coolant chooses the path of traveling this connection between the reservoirs and goes back through the radiator subloop again instead of being forced through the motherboard subloop. This isn't necessarily undesirable, since it means that the focus is on removing heat from the coolant...and less on removing heat from the components. Would work fantastic in low heat (idling) situations where I'm trying to spin the fans very slow (or not at all) and keep the coolant temp low.

But it probably is not as desirable when really stressing the components and trying to remove as much heat from the CPU/GPU's as possible. For these situations...more flow through the motherboard subloop would clearly be desirable. Closing the valve would utilize any remaining pressure after traveling through the radiator subloop and force more coolant through the motherboard subloop. I've played around with AQ5 temp data enough to know that more flow reduces the component temperatures....sometimes significantly.

I think I'll definitely add the valve. It's quick and easy....don't even have to drain everything all the way. Just a little bit.

Quote:
Originally Posted by Electrocutor View Post

RAM blocks in parallel

This may be the next largest increase in flow. I would guess these are pretty restrictive. Very thin. Plus...they are completely for aesthetics. Early on when I was benchmarking on air....the memory sticks never heated up much at all. Maybe 35c when I was running memory benchmarks. So these blocks are unnecessary....but I like the way they look. I may play with parallel....and I may just take them out and see if they improves flow much. I could always put back on the aluminum fin heatsinks...and paint them white. Or have them copper plated. biggrin.gif


Quote:
Originally Posted by Electrocutor View Post


How restrictive is your flow meter? You could create a straight bypass using a full ID valve if it would help when not taking measurements.

Martin tested this model awhile back and it is fairly low restriction. About 0.25 psi pressure drop at 1.0 gpm. I've got two of them...but that's not a lot. Extra connections for building bypass may negate any beneficial impact.


Quote:
Originally Posted by Electrocutor View Post


AQ5 blocks in parallel

Haven't ever seen any restriction testing for these....unsure of impact. They just fit together so perfectly right now...I hate to mess with that setup. Plus...they are actually in the radiator subloop already. Flow comes out of the top 480, then to the AQ5 blocks, then to reservoir. So they are not really on the part of the loop that needs to improve. Unless...the valve idea between the reservoirs (discussed above) really helps...and then I guess any improvement serves the entire loop once it is full serial.


Quote:
Originally Posted by Electrocutor View Post


The least restriction for pumps is to go in the front center and out the side edge (in an L shape); I'm not sure how much this would change, but that would remove 3 right angles from your current setup I think.

I've seen testing of impact of using front center for inlets....but haven't ever seen anything definitive on whether the outlet impacts anything. I'm using the front center inlets already. And I think that changing to using the top port for the outlet...would actually add another right angle for me. I don't see any benefit to changing this setup.

Quote:
Originally Posted by Electrocutor View Post


It looks like the bent clear tubing under the AMS rad is too tight of a corner and is pinching the tube?

It pinches just a tad...but nothing extreme. Probably still larger than any G1/4 connections....so not adding much to restriction. And...it's on the radiator subloop which already has pretty good flow. Might play with options next time it's drained. Could do away with any pinch if I added another angle connector....but not sure this would improve anything.

Quote:
Originally Posted by Electrocutor View Post


Probably the most restriction is being caused by how many G1/4 connections, angles, and fittings you are going through (each is small, but there are a lot to add up); but I'm not sure you can do much about that without sacrificing aesthetics.

May be true.


Quote:
Originally Posted by Electrocutor View Post


Use a pair of MCP35X or an Iwaki.

I'd probably go the route of more D5's before adding a different or more powerful pump. I like the way they look...and I like the way they sound. I actually do have the perfect spot for 1 or 2 more D5's on the PSU side....right on top of the AMS rad. That would be the ultimate in redundancy....six pumps. smile.gif



Appreciate all the ideas!
post #3279 of 3982
Remove as many 90deg as possible and use 2 45deg fittings. I know when my dad did this for our pool there was a huge increase in flow.
post #3280 of 3982
Thread Starter 
Quote:
Originally Posted by m1ndb3nd3r View Post

Remove as many 90deg as possible and use 2 45deg fittings. I know when my dad did this for our pool there was a huge increase in flow.

Two 45 degrees together to make a 90 degree? Or a 45 degree...separated by some tubing....and another 45 degree?
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Build Logs
Overclock.net › Forums › Case Mods & Cases › Builds & Case Mods › Build Logs › [Build Log] - The Big Budget Boomer Box (aka, the "BBBB")