Flow Restriction
In this test we measure the pressure drop across the waterblock while varying flow rate through it. This is done by hooking up the block to a house faucet. Flow is measured by a King Rotameter, while the pressure drop is measured by a Dwyer Manometer. If you want to learn more about what this plot means, then check out our guide.

We’ve tested so many CPU blocks now that our comparison plot is getting out of control:

flow

It’s far easier therefore to pick a flow rate and compare the pressure at that flow rate.  Let’s choose 1GPM – a flow rate that most users will be close to:

flowbar

The HeatKiller IV has been highlighted in orange.  The Basic acetal version is very low restriction indeed.  In fact it’s the lowest restriction block we’ve seen that uses a jetplate rather than parallel flow paths.

The Pro version on the other hand is higher in restriction, but they are not as bad as some.  My rule of thumb is that below 1PSI at 1GPM is good, and above it is bad.  With this metric in mind we can see that both are acceptable.  It’s interesting that the nickel plated version is measuring higher in restriction.  Normally we discount the effect of the nickel plating on the copper having any measurable effect because it’s so thin.  However these channels are very very thin already.  We’ll have to see if there is any thermal effect.

6 COMMENTS

  1. Greetings from Germany!

    Awesome review, thorough as usual and very much appreciated.

    While it may be impractical to add another block to the restriction graph, a graph with select blocks would be nice.

    i remember a tweaked/moded HK3 on skinnees (http://skinneelabs.com/heatkiller-3-0-lt-lc/2/) for those with a HK3 it might be interesting to see how it compares. Vapor had an improvement of 1.3 deg C which would in theory even beat the HK4. 😮

    all the best
    DX2
    p.s. i had a question about your testing, i sent an email.

  2. greetings from Germany!

    Awesome review, thorough as usual.
    A restriction graph of the HK4 as a link or compared to a select few would be nice.

    i remember on skinnees, Vapor reviewed the HK3 LT. He tweaked/modded it getting 1.3 DegC improvement. It would be really interesting to see if the tweak delivers, assuming it does, it would even beat the HK4 :-O

    all the best keep the reviews coming!
    i had a question about your testing. i sent a email. ^_^

  3. greetings from Germany!

    Awesome review, thorough as usual. Keep up the good work!

    A restriction graph ith a select set of blocks would still be nice tbh.
    The HK3 LT review from skinnee included a modded HK3. It was 1.8 Deg better than unmodded if those result hold true the modded hzk3 might best the HK4. It would be awesome if you tried the mod!!

    many kind regards
    DX2
    p.s. i sent a question via email.

  4. I’m curios about the price differences in these blocks. Was the universal back plate necessary at all? It adds ~12USD more to the overall price. Currently, this will make it ~100USD simply for the cpu waterblock. Just curios to know if they’re any help, or if they’re required.

    • Honestly I haven’t tried it, and of course its only needed for socket 115x, but I would be reluctant to not use it. CPU blocks can mount with a fair bit of pressure, and I’d want to avoid bending the board.

  5. One thing I can add after learning from experience. The HK3 water block works real well. Then the plasticizer from the flexible tubing clogs it. This happens anywhere from 6 to 9 month intervals with a OC’d AMD FX 9590 and 2 OC’d R9 285s. After the 2nd cleaning you notice the fine cuts in base are filled with crud. I found the best cleaner is tooth paste mixed with vinegar. The real problem is finding a brush with bristles fine enough go in between the ridges and stiff enough to remove it without scratching or bending them. So if you are not using rigid tubing, the HK basic maybe be better when it comes to cleaning. It is best not reuse the old tubing (leaks). Have not done a ridged reinstall, yet,,, but I believe careful disassembly and attention to orientation can be done without heating and bending new tubes. It is possible that the heat from the 9590 is more concentrated than the heat from the 285s because I did not notice as much crud in them?

Comments are closed.