Jump to content
Corsair Community

New Aussie Hydrocool owner - filters/A64 and tips


eva2000

Recommended Posts

I'm in Australia and yesterday I ordered the original Hydrocool unit, and today it was delivered :) Question where can i buy the 200EX unit filters and how much do I expect to pay for them ? Preferably a place which also sells the A64 waterblock as well :) Any beginner do/don't tips for a new Hydrocool owner ? I plan to run the unit on systems that setup on mobile utility benchies such as the one pictured here [url]http://www.fileshosts.com/pentium4/workbenches/utility1/web/[/url] ? Would positioning the Hydrocool unit at floor height, mid or above the motherboard in such a setup be more ideal ? thanks George
Link to comment
Share on other sites

Also how would you go about leak testing the unit /powering it on without actually powering on the pc ? I.e. connect the power on connector to PCI controller card and hook up all tubing/waterblock but not install it on cpu instead just existing heatsink and ran so when i power on pc, the existing heatsink and fan run but also the Hydrocool unit runs for leak testing ?
Link to comment
Share on other sites

George, By looking at the bench pictures on that web site, the HydroCool (HC) is about the same size as the shuttle computer shown in a few of those pictures. The HC should fit atop the bench or in one of the trays on the cart. Remember, the tubing and serial cable path to the computer is only four feet long. And make sure there is no clutter blocking the inlet/outlet air flow to the HC. As for the second question. Each HC is leak tested prior to leaving the manufacturing center. We test each HC with 20psi of air for about 10 mins. Any drop in PSI is a leaker that is then broken down for useable parts and the rest scrapped. The cold-blocks are pressure leak tested as well. This may be over-kill, but being safe is better than being sorry. The HC system's maximum psi is usually 3 psi. We chose special industry metal clamps and tubing to prevent leakers. Stev
Link to comment
Share on other sites

[quote name='stev']George, By looking at the bench pictures on that web site, the HydroCool (HC) is about the same size as the shuttle computer shown in a few of those pictures. The HC should fit atop the bench or in one of the trays on the cart. Remember, the tubing and serial cable path to the computer is only four feet long. And make sure there is no clutter blocking the inlet/outlet air flow to the HC. As for the second question. Each HC is leak tested prior to leaving the manufacturing center. We test each HC with 20psi of air for about 10 mins. Any drop in PSI is a leaker that is then broken down for useable parts and the rest scrapped. The cold-blocks are pressure leak tested as well. This may be over-kill, but being safe is better than being sorry. The HC system's maximum psi is usually 3 psi. We chose special industry metal clamps and tubing to prevent leakers. Stev[/QUOTE] thanks Stev i got it setup on a bench top system out in my garage right now powered up but not connected the cpu waterblock yet just running it to see how it goes - this is my first ever water cooled setup so can't wait to try it out on my different p4/a64 systems :) my first system i plan to test it on is Abit AI7 with P4 3.4C ... the AI7's 45 degree angled cpu socket won't cause any problems with HC's p4 waterblock right since it's the center core which still has contact ? You can see the 45 degree angled bracket on my web site at [url]http://www.fileshosts.com/pentium4/D1/abit/AI7/web/AI7_025.html[/url] [url]http://www.fileshosts.com/pentium4/D1/abit/AI7/web/AI7_026.html[/url] [url]http://www.fileshosts.com/pentium4/D1/abit/AI7/web/AI7_032.html[/url] i read numerous reviews of the HC original unit having a concave waterblock not making properly contact with the cpu heatspreader hence the need for Shin Etsu thermal paste but all these reviews were dated last year 2003, has there been any revisions on the original HC unit for the waterblock ? My HC model number HC200-1001 and unit's serial number is 3184 and seems to have a hand written E on the end of the serial number ? I plan to use AS5 first and then try a range of thermal pastes [url]http://www.fileshosts.com/thermalpastes/2/[/url] :) Lastly, where can i get the 200EX fan filters and A64 waterblock in Australia ? and how much would they retail for ?
Link to comment
Share on other sites

George, In all of our reviews of MOBO's and processors, we never ran into what the ABIT AI7 has. This is very strange with the 45 degree turn in the hold-down bay. There is no documentation at all in the Intel specs that show this 45 degrees. I'm forwarding your info to our development engineering. This indeed is shocking to us. The aluminum carriage works on a parallel socket. As for the concave cold-blocks, these were mainly pre-production prototypes that were sent out to the reviewer web sites (the first 24 HydroCools). The high temp brazing of the cold-block caused the micro-channel core to buckle upward, thus a concave bottom. We had then moved to low temp using silver solder and the concave problem went away. 3184 is the part number. "E" is the cold-block revision and part run allocation. B - E should all perform and function properly. As for the thermal grease, AS5 is conductive. We do not warranty the HC or the MOBO if users use this and the processor dies. We use the ShinEtsu-MicroSi non-conductive thermal grease that both Intel and AMD spec on OEM applications. Being only a degree off in performance isn't going to matter much vs. the AS5. Better to be safe than sorry. From the looks of the pictures on the filesshots web site, I take it that your either a reviewer or just really spending money and having fun with many toys. :biggrin: Stev
Link to comment
Share on other sites

[quote name='stev']George, In all of our reviews of MOBO's and processors, we never ran into what the ABIT AI7 has. This is very strange with the 45 degree turn in the hold-down bay. There is no documentation at all in the Intel specs that show this 45 degrees. I'm forwarding your info to our development engineering. This indeed is shocking to us. The aluminum carriage works on a parallel socket. As for the concave cold-blocks, these were mainly pre-production prototypes that were sent out to the reviewer web sites (the first 24 HydroCools). The high temp brazing of the cold-block caused the micro-channel core to buckle upward, thus a concave bottom. We had then moved to low temp using silver solder and the concave problem went away. 3184 is the part number. "E" is the cold-block revision and part run allocation. B - E should all perform and function properly. As for the thermal grease, AS5 is conductive. We do not warranty the HC or the MOBO if users use this and the processor dies. We use the ShinEtsu-MicroSi non-conductive thermal grease that both Intel and AMD spec on OEM applications. Being only a degree off in performance isn't going to matter much vs. the AS5. Better to be safe than sorry. Stev[/QUOTE] I see, so AS Ceramique would be okay as it's non-conductive as well ? As to the AI7, Abit says they angled the socket for optimal circuitry traces or something for prescott support etc ... seems to work well as AI7 is better than the IC7 i used and even better than IC7 Max3 heh... results of 3.4C on AI7 and 4PCA3+ so far [url]http://www.bleedinedge.com/forum/showthread.php?t=1748[/url] on air I think i'm the first person to use Hydrocool with AI7 so i'll be sure to provide some feedback as well :)
Link to comment
Share on other sites

a bit of both... i review stuff i buy on numerous forums informally as a way to find out what my purchase can do and at the same time inform others :D bit i've been asked to do a few ram reviews ... looks at my pile of 31GB of ram - 10GB of XMS2700/3200/3500 :D
Link to comment
Share on other sites

okay may have a problem... been leak testing the unit for 2hrs nearly where i have the waterblock dangling off a chair about 12-14cm off the ground with a paper towel directly beneath it on the floor. I just went to check it out and it seems the paper towel is slighly soaked in the center directly under where the waterblock has been hanging. I touch the block and it's slightly wet as well. The unit display temp says 12C while room temp is around 26-28C. I checked the waterblock connectors and they seem to be fine but slightly damp - alot of tiny bubbles gathered at the tube ending where it connects with the waterblock's barbs. I just changed a fresh new paper towel and this time wrapped each barb/hose connector in a paper towel as well and see if the paper towels get soaked. Is this normal? as i'd hate to have a little puddle on my motherboard which is horizontally positioned not vertical
Link to comment
Share on other sites

George, That is a surprise! :bigeyes: Is it the cold-block fittings that are damp or is it the tubing from the HC unit to the cold-block and fittings that are damp? :sigh!: Those white plastic clamps could leak if they are stressed by the dangling tubing weight of the cold-block and possibly they are not tightened fully. :!: Question: Is there any green marker/paint dots on the metal clamps on the cold-block? They may not be there since this is an appearance item. Inside the HC unit, the markings verify that each clamp is secure and leak tested. It is a very rare thing that the cold-block metal clamps would have a leak at those fittings. I may not get back to the forum until Monday. :D: Stev
Link to comment
Share on other sites

well i also had the p4 bracket clamp thing together with the cold block and both the top side of the cold block as well as top side of the p4 bracket clamp were moist - wiping it with a paper towel ya can see the greenish tint of the water the tubing from 25-40cm from the cold block end was also slightly moist after i wiped it with a paper towel - strange as no water made it past the funnel to the res inside the HC unit. The connector end of the back of HC with plastic clamps are fine no leaks at that end though. No greenish market dots on the metal clamps holding the tubing to the cold block end. I checked again now and there doesn't seem to be any more soaking of the paper on the floor... i'll let it run until tomorrow to see Might need to spray some PCB lacquer to water proof my board as a precaution heh
Link to comment
Share on other sites

well after 8-10hrs leak testing seems fine so decided to install HC200 on my AI7/3.4C :) Took me a while to figure out where to place HC200 unit so decided to use a chair besides the mobile utility trolley pc setup :) I used Ceramique thermal paste since it's non-conductive. I did a test mount, and to my surprise when i tried to remove the cold block to see the cpu imprint on the cold block to see how well it contacted, i found out that the cold block was stuck to the cpu! Eventually pulled the cpu out of the cpu socket stuck to the cold block :eek: I find this happens more when using aluminium heatsinks such as my Igloo 4300 100% aluminium fin HSF as well. Luckily there was no damaged cpu pins .. phew! Reapplied some ceramique paste and reinstalled the cold block. Booted pc and I love the temps I'm seeing :) --------------------------------------- Results with room temp at 28.5C - 28.8C P4 3.4C default 1.55v vcore in bios Abit AI7 16beta 1 bios 2x 512MB KHX3200/512 80GB WD 800JB 550W Antec TrueControl PSU Win2k Pro SP4 [b]HC200 cooled[/b] At boot in bios cpu idle = 39-41C system idle = 29C pwm idle = 34-35C At boot HC200 display = 28.5C - 29C In Windows cpu idle = 39-41C system idle = 29-30C pwm idle = 34-35C 2x Prime95 v23.8 cpu load = 45C system load = 31C pwm load = 46C HC200 Display under prime load = 31C - 31.5C ------------------ [b]Conclusion:[/b] 1. On average at stock P4 3.4C speed and vcore, HC200 water cooling reduced my cpu idle and load temps by 4-6C compared to my previous air cooled setup of Swiftech MCX478-V with 80x38mm Delta 80cfm fan on high speed. :cool: 2. HC200 displayed idle and load temps are ~10C and 14C, respectively lower than actual reported bios/board temps at stock 3.4C speeds and vcore. Need to do more testing though and I'll post some pics later on :)
Link to comment
Share on other sites

Hydrocool 200 oc'd results * Results with room temp increased to 30C :rolleyes: P4 3.4C @220fsb 1.725v vcore in bios (1.66-1.68v idle real vcore) Abit AI7 16beta 1 bios 2x 512MB KHX3200/512 80GB WD 800JB 550W Antec TrueControl PSU Win2k Pro SP4 * according to [url]http://newstuff.orcon.net.nz/wCalc.html[/url] 3.4C@3743mhz is outputting 114.83W of heat HC200 cooled At boot HC200 display = 30.5C In Windows cpu idle = 41-45C fluctuating system idle = 30C pwm idle = 36-37C 2x Prime95 v23.8 cpu load = 49C system load = 32C pwm load = 56C HC Display @ load = 33C C/W idle = (45-30)/114.83W = 0.1301 C/W load = (49-30)/114.83W = 0.1586
Link to comment
Share on other sites

Looking good m8! Heres mine, im using the shinetzu stuff on my non Ex hydrocool model which I installed the other day. Used nearly the entire tube of shinetzu. Abit-max3 P4 3.2 @ 3.8 1.70 volts (default being 1.55) Temps with my room being quite warm. Idle CPU: 35 Idle Case :34 Water: 25 After 19 hours of prime95 (very heavy load) CPU Temp: 45 Load Case: 38 Water: 30-32 Previously on air (Jet4 aero cooling max rpm) my load temps were over 60 under prime. So all together very pleased, like you. One interesting thing, on the pipes that plug into the unit itself, one of the O rings was missing! I only disocvered this after firing the unit up and therefore a small puddle appeared on the carpet (twas green!). I went down the DIY stores and managed to find a replacement O ring (3/8") and used that, topped the unit up a little and now all is fine - no leaks. Some lazy employee at corsair needs to apologise to my wife about her green carpet though :P
Link to comment
Share on other sites

you using Win2k or WinXP with that setup ? as i'm using Win2k and it's said it reports warmer temps than WinXP.. ?? With Hydrocool 200, so far been able to gain 5-8fsb over air cooling :) [IMG]http://www.fileshosts.com/pentium4/D1/abit/AI7/results/3.4C/HC200/Corsair/XMS3500C2/256/NB800/230-230-2226-TAADD-1.8-3.2-1.65/pcmark2002.jpg[/IMG] idle temps [IMG]http://www.fileshosts.com/pentium4/D1/abit/AI7/results/3.4C/HC200/Corsair/XMS3500C2/256/NB800/230-230-2226-TAADD-1.8-3.2-1.65/cpuz_idle.jpg[/IMG] [IMG]http://www.fileshosts.com/pentium4/D1/abit/AI7/results/3.4C/HC200/Corsair/XMS3500C2/256/NB800/230-230-2226-TAADD-1.8-3.2-1.65/9600pro_486-364_15295.jpg[/IMG]
Link to comment
Share on other sites

I'm using WinXP m8, win2k does not support hyper threading so is therefore not a good OS for a p4 cpu. The temperatures should be the same however as they are read directly from the probe's (maybe higher in XP as there is a load on the 2nd logical CPU). I'm on 3800 now rock solid, with air I would only use 3700 due to temps. I'm on 1.7v now and not too confident going above that, going to rest a bit before pushing for 4gig in a week or so (i rekon gonna need 1.75v). Here's my system and temps. [img]http://www.edgeris.boltblue.com/pc.JPG[/img] and the temps, this is over a 29 hour period running prime constantly and various games, it did reach 48 at one point but only for like a second, average load was 44-46. [img]http://www.edgeris.boltblue.com/mbm.JPG[/img]
Link to comment
Share on other sites

nice temps actually win2k pro sp4 does support HT :) here's where i am at with AI7 springdale board and 3.4C@3915mhz [URL=http://www.fileshosts.com/pentium4/D1/abit/AI7/results/3.4C/HC200/Corsair/XMS3500C2/256/NB800/230-230-2226-TAADD-1.8-3.2-1.65/pcmark2002.jpg]pcmark2002[/URL] [URL=http://www.fileshosts.com/pentium4/D1/abit/AI7/results/3.4C/HC200/Corsair/XMS3500C2/256/NB800/230-230-2226-TAADD-1.8-3.2-1.65/cpuz_idle.jpg]cpuz idle temp[/URL]
Link to comment
Share on other sites

here's some pics for ya! [url]http://www.fileshosts.com/pentium4/D1/abit/AI7/HC200_AI7/[/url] according to [url]http://newstuff.orcon.net.nz/wCalc.html[/url] 3.4C @ 3915mhz @ 1.75v vcore with 30C case and 50C cpu load temps is around 130W heat output and 0.153 C/W :)
Link to comment
Share on other sites

Nice m8 seems our CPU's are pretty similar, I'm on 1.7v at 3800 so guessing at least 1.725 for 3900. You had a crack at the big 4.0? I tried it briefly a few days ago when I first got the HC, was making memtest ok but errors in test 4. Saying that I only went as high as 1.725v so with a little more who knows :) Going to leave my rig at 3800 for a while though, it's rock solid (as u can see from prime screeny) and I need a stable PC for work this week!
Link to comment
Share on other sites

It's a 3.2 mate D1 (30 caps) stepping SL7WG I think. Mine has similar behaviour at 4ghz tho I've only tried lightly. We need to acheive it though, its cruel being so close but with no cigar - I'm determened to get there eventually :)
Link to comment
Share on other sites

[quote name='edger']It's a 3.2 mate D1 (30 caps) stepping SL7WG I think. Mine has similar behaviour at 4ghz tho I've only tried lightly. We need to acheive it though, its cruel being so close but with no cigar - I'm determened to get there eventually :)[/QUOTE] ah a 3.2C :) for >4ghz i'll shove the cpu into my Mach2 :D But too lazy to go through the whole mach2 install process :o
Link to comment
Share on other sites

[QUOTE]As for the thermal grease, AS5 is conductive. We do not warranty the HC or the MOBO if users use this and the processor dies. We use the ShinEtsu-MicroSi non-conductive thermal grease that both Intel and AMD spec on OEM applications. Being only a degree off in performance isn't going to matter much vs. the AS5. Better to be safe than sorry.[/QUOTE] Hey Stev, what's the problem with AS5 being conductive? I thought the manufacturer says it's not (though it is pure silver)? You mean if there's a water leak it will be a problem? I can't imagine a problem otherwise.
Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...