View Full Version : Cheap testing
Techmasta
09-04-2004, 06:17 PM
Ok guys, I've read all the threads where you bash reviews for inadequate testing methods. I have the opportunity to review waterblocks in the future and I want to do it right. The thing is I don't want to spend $1,000's of dollars on a system like pH's. I need that cheapest method with the cheapest equipment that produces better accuracy and repeatability than a thermal probe.
And I have another question, what are you going to do when waterblocks aren't made for Socket A any more?
bigben2k
09-04-2004, 07:03 PM
Adapt.
You know where I am, don't make me make you recruit #2 today., ok? ;)
It's possible to do it cheaply, but there are consequences to "going on the cheap": you'll have a hard time telling *for sure* which block performs better.
Today's WC products are getting better, and their performances vary so little.
Are you looking to test your own blocks, or start reviewing blocks for a site?
Techmasta
09-04-2004, 07:15 PM
I will be reviewing commercial blocks for a website.
bigben2k
09-04-2004, 07:24 PM
You have PM.
jaydee
09-04-2004, 10:52 PM
Ok guys, I've read all the threads where you bash reviews for inadequate testing methods. I have the opportunity to review waterblocks in the future and I want to do it right. The thing is I don't want to spend $1,000's of dollars on a system like pH's. I need that cheapest method with the cheapest equipment that produces better accuracy and repeatability than a thermal probe.
I want a BMW quality car for $5,000 new. Won't hold my breath though.
And I have another question, what are you going to do when waterblocks aren't made for Socket A any more?
Die sim.
BalefireX
09-04-2004, 11:56 PM
I want a BMW quality car for $5,000 new. Won't hold my breath though.
Well lets not run him off just yet :)
He is asking to try to be part of the solution rather than the problem, and that must be a good thing.
Obviously precision costs money, but there is no reason that you can't do an adequate (ie, good enough for basic comparisons) job of testing without spending thousands.
I would think the cheapest solution for testing would be a watercooling system with a powerful pump, a reasonably accurate flowmeter (most expensive bit), and a ball valve to control flow rates. Then do the motherboard mod to read directly from the thermistor for more accuracy, do multiple mounts of each waterblock to minimize mounting error, and graph temperature vs flow rate. Make sure to write up all the components in your system and your methodology so that others can see what you've done.
This is not perfect by any means, but its a hell of a lot better than the typical tester who slaps a block on once, runs a burn program and records the MBM temps, then calls it a day.
Etacovda
09-04-2004, 11:56 PM
http://forums.procooling.com/vbb/showthread.php?t=9772
The CF and DOW thing could be for you. Its nice and cheap, and has an ok accuracy. Multiple pump pressures are a must though.
jaydee
09-05-2004, 09:00 AM
Well lets not run him off just yet :)
He is asking to try to be part of the solution rather than the problem, and that must be a good thing.
Not unless he is technically qulaified for the job and not into spending the required $$$'s to do it decently. All he will be doing is adding to the problem period!
vBulletin® v3.7.4, Copyright ©2000-2024, Jelsoft Enterprises Ltd.