• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

The Storm G5, 50 microns from perfection

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Joe Camel

Senior Camel Kicker
Joined
Aug 6, 2003
Location
---> NEW HOUSE 7/17/09 !! <---
this is a PM i sent to Cathar:




so today i got the G5 :attn: and just for grins & giggles, i put a straight edge on it....ummm....


@ this angle every thing is GREAT:



BUT then we move to the other....

pict01452ci.jpg



i am NOT going to post these on the forum until we "talk"...

please give me your input!!


@ a loss for words,

(aka Joe Camel)


PS this is the same straight edge/test/light source etc etc as in this thread LINK .


Cathar said:
What can I say? Not very good is it?

Something I noticed with a few of the blocks is that came back slightly curved from lapping, and again, once assembled also gained a small amount of curvature. I had to re-lap a number of blocks by hand.

If you could just do a quick test for me and see if the base is still like that when the block is disassembled (base-plate by itself).

Due to the softness of the silver it seems that with manufacturing variances in the O-ring channel depth that some O-rings squished a little too much upon assembly and caused the warping which you can see. Looks rather dramatic with a straight edge, but is probably about 50 microns in "bow" from edge to edge.

Disassemble the block and redo the test. If the plate is fine after you do that, then this is an assembly issue and can be fixed by some lapping with 600-grit paper with the block assembled.

I tried to catch all these but I had someone helping me who was assembling and packing blocks and let some slip through. The other possibility is that this is heat related, and again, disassembly and reassembling in your local environment may improve things.


.
 
Last edited:
here are the pics of the base unassembled:



i am almost SURE this bow is caused by the block "permanently" pulling the silver out of true. (since the silver is SOOO soft)



these next 3 shots show: TOTAL PERFECTION!



("same" shot/axes as above)




the tubes came out PERFECT !!!! (too)






baseline the disk i stone wheel surface ground flat.






MUCH MUCH BETTER!!!!

i did a test today with a Micrometer , you can see a .0015" (INCH) gap!!

that is about 1/2 the thickness of a sheet of copy paper!!!!

(ill take/post a pic of that later this weekend too, i hope)

so yes im fully aware of the "possible" accuracy of this test and the #'s are VERY small.




Cathar said:
As you can see in your pictures though that the issue is caused by assembly as I suspected. Try re-assembling and not tightening down the bolts that hold the block together too much and see what happens then, if the gap is worse than unassembled.

Quantify that, then lap with the assembled block.

The other alternative that may work is lapping the top side of the silver plate, which may be slightly uneven which is what is causing the bow when assembled.

0.0015" is 38 microns. I said it looked like about 50 microns, so overall not too bad of a guess.

Sure, you can post it and point it out. Many blocks aren't affected at all, some may be. I was somewhat at a loss with it as an issue though, because if it happens on assembly and you lap it, then pull it apart, then it's "out" again.

This is something that I've (hopefully) addressed by way of a small extension to the design, whereby the base-plate is "supported" by the middle plate better.

Stew.
 
Last edited:
here is the pic of the block reassembled:
(i tightened them as tight as i would considering: too loose and there can be a leak IN my PC)


pict01559bg.jpg


although a lot better, there is still a "bow" that a small amount of careful lapping can help.

more to come ;) (nice lil weekend project)

id like to thank Cathar for allowing me to post these (pics AND PM's) and his input on how to fix the "problem"
 
Last edited:
What is the thickness on the base plate?

Swiftech has something like a 5mm plate on their blocks. HSFs are obviously much thicker.

Very interesting. I will be watching to see how this all works out. With a soft metal and thin plate this looks like a problem. Though in the big picture its just a nuisance.
 
all hosted by imageshack....odd that some work some dont...looking into it...

all but those 6 (out of the 63 pics they host for me) are working....give it some time...im SURE they will come back soon...if not back up by tonight, ill re host them!!

damn they are down again...came back from lunch and they loaded...but now dont, again....SUP!!? imageshack! dont do me like this!

as soon as i get home from work, ill rehost...this is *%$(%^@#
 
Last edited:
Joe, you must assemble the block then relap it... that will fix er up. It will take some time but it will be worth it.
 
hmm, i'm goin to have to check mine out, b/c i'm getting some abnormally high temps

nice post
 
nikhsub1 said:
Joe, you must assemble the block then relap it... that will fix er up. It will take some time but it will be worth it.


like this...?


1 page of 600grit


pict01609dz.jpg



2 pages of 600grit


pict01613th.jpg



3 pages of 600 and 1/2 1500....(all the 800 is used)


pict01623pa.jpg



still have a long way to go IF i decide to go for "mirror" finish


pict01674ju.jpg


gotta love the shine off my :cry: balding head :cry:

EDIT/notes:
silver lapps "somewhat" like copper...loads up the paper FAST!!...i made the jump to 1500 too soon or somthing, i thought i would have more "miror" than i do @ this point...1 "bad" pass on "hot" (fresh) paper can really do some dammage to silver (worse than copper)
 
Joe, my trick is to take a Sharpie and make lines all over the base of the block, start lapping but use soapy water or kerosene or some lubricant, if it gets hot it will not be good. Slowly and gently start lapping, keep an eye on the sharpie marks, see where they disappear first. Once all the marker is gone, remark the base. The marker 'should' sand away evenly. IF it does check the flatness and you are good to go. Another note, mark the orientation of the 3 parts to the block, mark them with tape or something. This way, if you ever disassemble the block, you will know the way it needs to go back together. If you don't and something is reversed, the base may not be flat... never lap the base unassembled after lapping it assembled. I hope this makes sense.
 
Yah, keep going until it looks like this. I wish that I had the time to do all of the bases like the following, but there's well over an hour of effort and $15 worth of sand-paper in each one. Instead I had to opt for machined finishes on most of the bases otherwise I'd still be sending the blocks out by Christmas.

mirrors.jpg
 
Yeah, lapping and grinding take the most time out of any one operation when putting together the STASIS heatsinks as well... :(

Time well spent however! :)
 
WOW Cathar, thoes look.....:drool:....what else can i say?
have YOU noticed a "loss of performance" going to a mirror finish?

nikhsub1, all my lapping has been done DRY. i played around with wet lapping once, but the parer loaded up in like 10 passes. did i do something wrong or is that normal? i lap for about 3-5 min (wipe paper 3-5 times) and then take a break and let the block (my arm) cool down.

to ALL, thanks for your input!!
 
you should do it wet so you don't grind away too much of the silver, also so you don't leave too many scratches
 
lilneel12 said:
you should do it wet so you don't grind away too much of the silver, also so you don't leave too many scratches

the "too many scraches" i think i understand, but how can it grind away "less" if its wet?

if im @ the mirror finish part, i can see the advantage, but if im still in the "going for flat", wouldnt it just take longer?

edit: i totally understand the "expand when hot" theory
 
well, I didn't lap mine, is that something I should consider doing? even though I have everything done w/ my setup?
 
Joe, try lapping with some water, just a few drop will do, but make a mixture in a cup of water and a few drops of dish soap. Mix it up, dunk yer hand in the water and sprinkle a few drops on the paper. As the water starts to disappear, add more. I always duct tape the paper down, i dont get any loading issues... if you do just work through it.
 
Back