Planning Custom DDC pump top

Hi hi

If you've seen my Asteria II project log, you'll know that I'm having a few minor niggles with my DDC pump placement. To cut a long story short my original requirements had a pump top with ports on the side since I can't realistically use any ports on the top. Not being able to do that is causing some headaches and untidiness.

So I thought "let's just make my own".

Having dissected a few DDC tops, it really does seem to be nothing more complex than a 40mm round space in the middle of the top 8mm deep for the pump head, the outlet port positioned to the outside edge of that round space (presumably for maximum speed from centrifugal force) and an inlet port put largely anywhere to get the water in and spinning.

Is it really any more complex than that?


The only thing I'm looking at is the vertical placement of the outlet. Some tops, such as the XSPC one, have the outlet port exactly half way over the base of the round area, so half of the "tunnel" is aligned with the space. Other tops, such as the Alphacool top, have the outlet port aligned central to the top's thickness, which would align the "tunnel" mostly away from the round area.

Tips, tricks and advice please? Smashing a forstner and couple of twist bits through acetal isn't likely to be an issue if I take things very slowly with plenty of lube, it's more the placement of everything I want to confirm.

Cheers.

 

Nooooooooooooooooooooooo!

With any and every DDC top the water feed is central to the impeller, it may be ported elsewhere for packaging reasons, but with shrouded impeller pumps such as the DDC your inlet needs to be at the centre of the spinning puck of doom.

Whats more the inlet feed is preferably the same diameter as the impellers inlet hole which is smaller than most pump tops - 'cause it's easier to reduce tool changes you'll find that usually all the long holes are 11.8mm.

I've not got a DDC and even if I did impellers do vary between versions (whether the inlet diameter does or not I don't know) so if you open up your pump and measure the hole in the top of the impeller that's the dill size you want to use for porting which can "flare" to accommodate whatever thread size you opt to use for your couplers. It looks like about 8mm guessing by pictures.

So you've got your 8mm deep recess for the impeller, then lets say 2mm of material before an 8mm (guestimate) port and 2mm of material on top - so 20mm minimum thickness. Unless you do something tricksy*... 8mm diameter is circa 25mm²and we'll be meeting an 8mm wide opening so we need a depth of 3.125mm - so use a 3.5mm drill bit to drill two holes and "breakout" the middle section. That gives 8+2+3.5+2=15.5mm

Making sense?

*hobbit

 

Right, so regardless of what happens I need to feed water directly over the centre of the impeller. That'll be annoying given I don't want anything on the top face of the top.

So essentially then I have to drill a tunnel from the side of the top into the center directly over the impeller, then bore from the recess a little bit more to meet the tunnel, thereby dropping water onto the centre of the impeller?

 

I've done a sketch...



First bore your impeller cavity. Then drill your inlet to the impeller. Then drill your connecting holes - I've drawn roughly what the 3.5x8mm version might look like. You also (as you previously noted) want to get the output as close to tangential to the impeller cavity as possible.

I'm also guessing that 2mm thickness of delrin between the impeller cavity and the input porting and between the input porting and the exterior of the top will be mechanically sufficient to withstand the pressure produced by the pump.

 

Thanks for the sketch, it's pretty much what was in my head after your clarification on feeding the impeller. Does the output need to be fully in line with the impeller recess as you've draw in, or can I vertically-center it so half the tunnel is above the recess?

I'll go back and update my scale drawings.

 

Nope doesn't have to be like I've drawn at all - I hadn't been going to put an output on the sketch, but then threw one in for the sake of input-output* in fact the ideal would be a steady taper from a square opening the height of the impeller blades, to an 11.8mm diameter hole for tapping. In reality a slightly oversized round hole (diameter = diagonal of the square) will be as close to perfection as you're likely to get. If you were feeling very through (and have viable length/thickness) then once you've drilled your (let's call it) 6.5mm hole to the impeller void, and your 11.8mm section for tapping, you might consider measuring up the inner diameter of your fittings and using a cone burr to taper the 6.5mm up to 10mm** with a steadily increasing bore rather than a sharp step which will induce eddy currents and vortices...

but that's a bit ocd, I haven't noticed any of the commercial manufactures doing that...

slackers...




*to the badger-badger-badger-badger-badger-badger-badger-badger song... mushroom mushroom = input-output... you had to be there

**[/i] as a guess at what the inner opening of the fitting might be.[/i]

 

Good info, very much appreciated.

I'll play it safe for now given I'll only have a drill press and a couple of drill bits. The goal of this discussion is to get a DDC top that only has ports on the sides, and given the pump is hidden behind my graphics card it won't matter if the top isn't super-pretty with vertically-aligned holes and whatnot. Function over form in this specific instance opens up a whole world of improved form for the project in its entirety.

I'll probably post up some basic layout drawings of the thing later on for you to cast your eye over, if that's OK?

 

I understand that you need a top with nothing on top but would a top with the top port with a plug work for you?

 

That would be most factory tops then

Alas no, a top plug will get in the way of the graphics card unless the pump is placed at a very specific point. Also, most DDC tops with side ports are greater than 20mm thick (so they won't fit) or are larger in area than the DDC itself, so again won't fit. That pretty much leaves the Aquacomputer top which I already have, which in itself was a compromise.

I do have a placement that will work, and I've made a new steel plate for it to remove the top inlet port, but it leaves me no margin for error and tube routing that's ugly and very awkward. Hence looking at making something specific.

Ultimately the issue is I cannot go larger than 45mm tall for the entire pump assembly, which leaves 20mm for the top itself. That in turn means no top plugs. That gives me the flexibility to squeeze the DDC behind the graphics card as originally planned, but raise and lower the vertical position of the pump unencumbered for best fit of tube routing, anti-vibration pads, etc.

It's a tricky one.

 

So asura, quick layout based on what we've discussed.

Side view to illustrate feeds

• 8mm deep impeller recess with an 8mm diameter feed (yellow)
• 6mm diameter inlet into the impeller feed; threaded hole vertically offset a bit (red)
• 8mm diameter outlet (green)

No crazy angles since I'll just be stabbing acetal with some drill bits in a drill press, but is this the sort of thing I should be looking at?

Cheers.

 

Looks good - just be sure that your 8mm outlet fully covers the depth of the impellers - it can be offset to the 11.8mm hole like your inlet is...



... so measure them up, otherwise you'll be down on flow and pressure.

 

So the green tube needs to be as wide, if not wider, then the spinny blue bit?

 

Missed the 20 mm thick criteria. You might want to take a look at the conclusion's from a Martin's Liquid Lab Review of custom DDC tops.

Arg, your diagram just gave me a flash back to when I did ASIC design in university. Lots of classes lead to working late on projects, but that was the only lab I had were the entire class stay past midnight working on the finale project. Around 3 am was when the group sing along started, kicked off with a rousing rendition of Teddy Bear's Picnic.

 

Funny you should mention Martin's roundup because I was discussing spiral volutes with JR23 about an hour ago. No way in hell could I do a spiral volute with a drill press and forster bit though

Now of course there's no point in doing this if I'm going to totally ruin my pump performance, but I'm OK with "good enough" to get this done.

Worst case scenario I can make the Aquacomputer top work and just put up with ugly plumbing. Most of it will be underneath the GPU so not immediately obvious, but it'll irk me if it's too scruffy.

 

The more critical detail that I thought you should consider was matching input and output hole sizes to the impeller. Until I read Martin's comments I would have guessed the opposite to what he says optimizes the flow rate.

 

Missed by that much… err I guess you can't see my fingers… We'll I'm holding them 1mm apart. I got a spare Danger Den ddc top and they are 21 mm thick. Yes there is a hole top centre, but I figure that it can be pluged could be glued into place and be level with the top.

 

I have some grub screws floating around after sealing up some ports on my Alpahcool radiator, but that was a faff and lots of mess with Fernox LSX. They don't work on the Aquacomputer DDC top I have at the moment because they're quite long and distort some of the threads on top (it's a weird internal construction).

But yes, sealing up top ports with other material and glue is an option too. I have routes to explore.

 

D'oh! My brain is jelly. Area of a circle is pi radius squared, not pi diameter so if the impeller inlet is 8mm than its a sectional area of 50mm2 not 25 as I mentioned above... oops.

 

Hi, me again.

I've done some musings in Fusion 360 based on tools I have access to to stab some acetal. If I put an STL file up somewhere, would somebody be able to view it and tell me if the design would work?