Storm Waterblock (FAQ in first post)

Discussion in 'Extreme and Water Cooling' started by Cathar, May 19, 2004.

  1. Cathar

    Cathar Storm Father

    Joined:
    Aug 16, 2001
    Messages:
    6,507
    Location:
    Melbourne, Australia
    Storm Waterblock Thread Introduction

    Most of the information in this post can be found at the website here: http://www.employees.org/~slf/lrwb/

    IMPORTANT ANNOUNCEMENT - Effective 11th June 2005

    All final payment request emails have been sent for the Storm/G5 ordering process.

    If you haven't received one and were expecting an email, then email me through my website:

    http://www.employees.org/~slf/lrwb/

    All of the emails have just been sent out to everyone from whom I've received a deposit.

    If you've paid me a deposit and haven't received an email from me, then email me.

    Also, since more than a few people sent emails to me from up to three different email addresses, I've just responded to the latest email address that you used in your correspondence with me. If you have multiple email accounts, check with them first. If you've since switched email accounts and haven't gotten an email from me, email me from your new account, citing your old email address for security, and I'll re-forward the final payment request to you.

    NOTICE

    At various people's request, I have organised the old mega-thread into two separate threads. The old Cascade XXX/XS thread can be found here: http://forums.overclockers.com.au/showthread.php?s=&threadid=263403.


    SCOPE

    In order to facilitate easy an easy introduction to the content of this now also very large thread, and also to make it easier to track new updates, I'll update this opening post with a summary of the pertinent information so people don't have to always read the whole thread content


    INTRODUCTION

    This thread details the ongoing work of the "Storm" waterblock. It evolved out of a concept embodied with the "Prototype 2" or "P2" block, for which early details and discussion can be found in the latter half of this thread.

    PICTURES

    While no internal shots of the block are available, I do have a few pictures to view:

    Socket A format picture #1

    Socket A format picture #2

    Jet acceleration in action spraying 6m distance

    Jet acceleration as a 3m high fountain

    FAQ
    Last updated 11 Aug 2004

    This section is purely an attempt to prevent the same questions being asked again, and again, and again.

    Q: What's with the G1, G3, G4, G5 things? Are they different blocks?

    A: See "STORM DESIGN CLASSIFICATIONS - G-level" below. The G-level is a reference to a particular sub-design complexity of the Storm design. The higher the number, the higher performing the design.

    Q: What sub-design have you settled on to make available for people?

    A: I've settled on making a G4 available in copper, and I will be making a very limited number of G5 sub-designs available in silver.

    Q: How much will these things cost?

    A: The final price has been set at $116 Australian dollars. Subject to exchange rate variations, this works out to around ~US$96, ~£49, ~€72 (as of 5th Dec 2004).


    Price for the silver Storm/G5 has not been considered yet, but at least double the price of the copper version would be a good bet.

    Q: When will they be available?

    A: Presently on target for last week September or first week October.

    Q: Will you accept pre-orders?

    A: No. At least not until I have a stronger indication of when they'll start being made.

    Q: Will you have international distributors/resellers?

    A: Maybe, maybe not. Depends on a number of things that are still to be resolved.

    Q: I live internationally. I want one. How much is postage if I get one direct from you?

    Postage is typically about US$10, and international delivery typically takes 5-10 working days.


    Other Questions Answered

    Q: How many barbs will it have?

    A: Two barbs only. One inlet, one outlet. There will be no 3-barb versions of the Storm, even on request.

    Q: Will it work well on CPU's with an IHS?

    A: It is designed with both bare die and IHS covered CPU's in mind.


    STORM DESIGN CLASSIFICATIONS - G-level

    The Storm design can be broken down into 6 levels of geometrical complexity, which I've coined G1 through to G6, with G1 being the easiest/fastest to make and lowest performing variant of the design, and G6 is the most complex to make and highest performing. The G6 right now is just a theoretical machining limit, and it is quite unlikely that I'll ever make it, but it's just there for a reference point.

    For tests I use the equivalent of an MCP600 sort of pumping power.

    The G1, of which I had one made, simluated out to around 1C worse than a White Water, and in reality performed about 1.5C worse than the White Water.

    The G2 I have not had made up. In reality it should perform about on par with a White Water.

    The "Prototype 2" referred to in the old Cascade XXX thread is essentially a G3 version of the "Storm" design. The G3 is basically slightly better than Cascade level performance, or roughly on par with a Cascade SS.

    The G4 is the "do I make these or not?" prototype block. Simulation-wise it works out to around 0.5C better than the G3. In reality, the G4 worked out about as expected, being a little more than 0.5C better than the Cascade SS given the same pump.

    The G5, which is the one that I'd make in silver, simulates out to around 0.8C better than the G4 (with the G5 in silver and the G4 in copper) when given the same pump. The G5 prototype has yet to be made.

    The G6, in silver, simulated out to a mere 0.3C better than the G5 (in silver), so its cost and complexity is unjustifiable except as a curiosity.
     
    Last edited: Jun 11, 2005
  2. phextwin

    phextwin Member

    Joined:
    Aug 30, 2003
    Messages:
    4,221
    Location:
    Geelong Velocity:Unknown
    So the complexity and effectiveness is scalable?
    Will (if you go ahead) the production run be of multiple designs or will you settle on the most effective design per cost kind of thing?
     
    Last edited: May 19, 2004
  3. hendot

    hendot Member

    Joined:
    Jun 27, 2001
    Messages:
    1,276
    Location:
    Sydney
    I reckon you should go with 2 designs ala silverprop evo and evo S.
    I'm sure you could make a great 'budget' model with only 5mins on the mill and a world-leading model (G5?) also available in silver.
     
  4. Benjo

    Benjo Member

    Joined:
    Mar 8, 2004
    Messages:
    240
    Location:
    VIC
    avaliable in silver? my god, that block will set ya back a fair amount..... three figures???

    Anyway, nice job cathar, keep up the good work!!!!

    cheers,
     
  5. kooliez

    kooliez Member

    Joined:
    Jul 30, 2003
    Messages:
    1,277
    Location:
    Brisbane - Qld
    benjo, yes Cathar has made blocks out of silver before and they tend to preform around .8 to 1 degree better than their copper equivilant.

    Cathar has stated a few times in this thread about the chances of a new copper block and has refered to the Cascade-SS.

    The Cascade-SS is silver also. i dont know how much they went for when he made them but i know that i personally would consider paying $200 - $230 for a Cascade-SS.

    to have a look at one click the link in cathars sig.
     
  6. OP
    OP
    Cathar

    Cathar Storm Father

    Joined:
    Aug 16, 2001
    Messages:
    6,507
    Location:
    Melbourne, Australia
    Sorry, am not really interested in offering a confusing range of models. I know that this also frustrates the crap out of resellers as well.

    If the G1 performs as expected, I may offer it. It'd be super-budget though, and would only be of a first-stage base-preparation, which is a basic grinding of the base. The base would not be guaranteed to be super-flat, but could be easily lapped properly by the end-user with some 600-grit paper in about 15 minutes. Basically built to a price, not to a particular level of finish, but if it could be done for a $50-60 final price, then I'd consider doing it. If it can be done for less, I'll try to do that too, but above $60 is where I'd not bother.

    The G4 is the block that I want to offer, purely from a price-performance standpoint. The CNC time is only a fraction of the total time that goes into producing a block. The machine-time difference between the G3 and the G4 is small enough that we'd be talking about less than a $10 difference, and I'd rather aim at a higher performance point in that circumstance.

    I will always end up making a G5, in silver, even if only for myself. Once I do that, I'll assess the situation for making more silver G5's.
     
  7. OP
    OP
    Cathar

    Cathar Storm Father

    Joined:
    Aug 16, 2001
    Messages:
    6,507
    Location:
    Melbourne, Australia
    They sold for $255 in the basic Socket A 2-barb format. A few dollars extra for the P4 or A64 models. The silver in them costs $125. With the G5, the weight in silver would be less, and we'd be looking at around $95 of silver in one. In fact, all things considered, a silver G5 could be done for under $200, but don't hold me to that.
     
  8. kooliez

    kooliez Member

    Joined:
    Jul 30, 2003
    Messages:
    1,277
    Location:
    Brisbane - Qld
    me no care bout cost, me just want a silver waterblock... because it has.... "Shining... performance...." you get it... "shining" becau...

    Nevermind...

    either ways, let me know when they are released and tell me how much cash to transfer!
     
  9. Benjo

    Benjo Member

    Joined:
    Mar 8, 2004
    Messages:
    240
    Location:
    VIC
    I know what cathar has done in the past, i was just commenting...
    imagine worldwide distribution of silver CPU water blocks.

    I and I imagine most of you would have never thought of this sorta thing happeneing about 6 months ago, i know that silver blocks would have to be ordered and made, but still....

    Cathar would be damn well known in the world of watercooling, that is increasing every month, and just overclocking in general!

    Cheers,
     
  10. Dilz

    Dilz Member

    Joined:
    Oct 20, 2003
    Messages:
    564
    Location:
    Australia
    Is the G4 going to be between $80-100?
     
  11. OP
    OP
    Cathar

    Cathar Storm Father

    Joined:
    Aug 16, 2001
    Messages:
    6,507
    Location:
    Melbourne, Australia
    Not really. The problem with the lower complexities is that they need to be quite tall. As it stands, I was struggling to bring the G1 in at 32mm tall, and even then made a couple of design choices to get it that low. To make it properly it really should be at least 40mm tall, so it'd be sucky as a GPU block, given that the barbs would then have to sit on top of that, be it sideways or whatever.

    The G3 I have here is 25mm tall, but really should be 30mm tall to do it properly. The G4 I have coming is being done properly at 25mm tall.
     
    Last edited: May 19, 2004
  12. Joe Camel

    Joe Camel Member

    Joined:
    Mar 27, 2004
    Messages:
    49
    Location:
    Ohio USA

    YES please!!

    (id say more but whats the point;) )

    GOOD LUCK

    PS will some of us who LOVE to go overkill (2x 16v D4 pumps in series with 5/8" and 3/4" tube) see any better #'s?
     
  13. OP
    OP
    Cathar

    Cathar Storm Father

    Joined:
    Aug 16, 2001
    Messages:
    6,507
    Location:
    Melbourne, Australia
    The larger tubing size is unnecessary. ;)
     
  14. MadCap666

    MadCap666 Member

    Joined:
    Jul 28, 2003
    Messages:
    19
    Location:
    Slovakia
    btw tube size ... for cascade is even 1/2 not needed, because the overall surface area of cups is smaller .. in fact 10mm or 3/8 ID is just enough ..
    MC
     
    Last edited: May 20, 2004
  15. Joe Camel

    Joe Camel Member

    Joined:
    Mar 27, 2004
    Messages:
    49
    Location:
    Ohio USA
    if i (already) went thru the effort of getting a 3/4" D4, over volt it and mod it so it runs cooler, im not going to go and put 1/2" tube on it.

    BTW im running an external setup so ive got a total of 6'-8' of tube.

    if Cather goes thru the effort to tweak the WB to get .5C why is it not worth it for me to TRY to get "just a little more" by using bigger fittings/tube?

    i know the barbs on the "Storm" WILL be 1/2" but things like the pump(s) and rad(s) dont.

    so unless im going to see higher temps or do damage to the WB, i dont plan on going to smaller tube.

    i will experiment with pump(s), tubes and rad(s) either way, i was just wondering if Cathar had an educated guess on the #'s...
     
  16. OP
    OP
    Cathar

    Cathar Storm Father

    Joined:
    Aug 16, 2001
    Messages:
    6,507
    Location:
    Melbourne, Australia
    MadCap666 the block is restrictive because that's where he restriction needs to be.

    The tubing is 1/2" because that's where the restriction doesn't need to be.

    Even assuming a generous 3m of 1/2" ID tubing, and flow-rates up around 10LPM, the tubing present about 0.5m of pressure drop, while a Cascade block is operating at around 5m of pressure drop.

    If we put 5/8" ID tubing on there, we'd gain about 30cm of pressure drop, which would result in about a 0.1-0.2LPM difference in flow rates, or near zero difference at the block.

    The reason why I don't like large ID tubing is main because it's really heavy, especially when filled with water, is that it hangs with a lot of weight off the waterblock.

    1/2"ID is IMO about the largest practical tubing you can use.

    I bought the Laing D4's with the 3/4" OD barbs, mainly because then the barbs are 1/2"ID, which is perfect when stretching 1/2" ID tubing over the barbs.
     
  17. MadCap666

    MadCap666 Member

    Joined:
    Jul 28, 2003
    Messages:
    19
    Location:
    Slovakia
    yeah i know there is no need to have restrictions anywhere else, but i can`t see any difference in 10/14 or 12/16 tubing (mm, ID/OD) in "normal" flowrates :)
    MC
     
  18. OP
    OP
    Cathar

    Cathar Storm Father

    Joined:
    Aug 16, 2001
    Messages:
    6,507
    Location:
    Melbourne, Australia
    Joe Camel wasn't asking about "normal" though. He defined the pumps and their speed, which works out to >10mH2O of pressure drop at 10LPM, as opposed to "normal" which is <2mH2O of pressure drop at 6LPM.

    Two totally different scenarios. 3/8" ID tubing at the 10+LPM that the twin-Laing's could push would be around 1/3rd of the total pressure drop, as opposed to 1/10th with the 1/2" ID tubing.
     
    Last edited: May 20, 2004
  19. MadCap666

    MadCap666 Member

    Joined:
    Jul 28, 2003
    Messages:
    19
    Location:
    Slovakia
    ough, sorry i didnt know that the pumps are so powerfull ..
     
  20. jo813

    jo813 Member

    Joined:
    Apr 12, 2004
    Messages:
    2
    Cathar

    What's up with the manufacturing?:)
    You apprear to be very busy.

    Regards
     

Share This Page