Stuff that needs cleared up on temps and a few other things.

[jaydee]

I am going on another road trip tomorrow (9-4-2002) and may not be able to get on the net. So If I can't I would like to hear some good answers to the following statement,comments, questions when I get back. Unfortuantly I pissed off unregistered enough to put me on his ignore list and his input may have been usefull but oh well.

So here is goes.
1) Temps??? I here from a few individuals consistantly that we cannot accuaratly measure temps to their liking. Well instead of throwing a bunch of stuck up attitute towards us please explain in detail what it will take for use to come up with the proper temp monitoring device that you will "accept" and not whine about. Give us schimatics, parts lists or where to buy a current model and then tell us in detail how to properly attach this device to get acceptable results. Like I said I really don't give a rats ass about 1-2C but if it is the only way I am going to be able to be apart of this forum without being harrases then I will do it! Hell it will be a good project anyway. And please don't give me the google is your friend BS. I want a simple direct answer, not a long winded one or links to previous long winded answers. If you guys are as good as you claim then it should be a rather simple answer.

2)Copper vrs. AL. I think I have finally understood myv65 and others on how Copper and AL react. And it seems to fall right into place with all the testing results I have done. I seen myv65 lay it out before but it is just to technical still for us not so quick to catch on people. I would like to see it written in a more novice friendly form. Maybe with a few illistrations to help. Something that can be quickly linked to when needed. I don't want to see arguing on which is better or more practical.

3)(mainly for the peeps that don't make their own blocks) Overclocking. Explain to me why it is more important to spend a bunch of money on misc. water blocks to gain a few C to OC higher when it should be far more efficent to just make the water colder with a chiller of some sort. I mean if OC'ing is why you went with water cooling then should this not be what you want to do as it can gain you better results than just strait water cooling even with the perfect block.

Well thats all I can think of right now.

[pHaestus]

1) To be useful, good numbers for flow rate and wb inlet and outlet are needed at the minimum. Drilling out baseplates and getting a bp temp would be suggested as well. Dunno how useful CPU temps are yet; playing around still with better gear.

Here's what I use currently:

For water temps (wb in and out) and air temps (rad in and out) I am using a Digitec 5810 thermometer and YSI 700 series linear thermistors. Not perfect, but two of the water probes read exactly the same (to 0.01C) over the range of water used.

For CPU temps I use a Maxim 6655EV System from dallas semiconductor. I bought it because the homemade readers are limited to the resolution of MBM (1C) while the maxim software is 0.125C (over || port). It took some fiddling to get it working acceptably.

For wb testing, I am primarily relying upon the delta T across the waterblock at this time as I still don't know how relevant the diode temperature of the CPU is. I am pretty confident a superior block will put more heat into the water though.

Flow rates are basically "the master variable" for wb testing, and I use a GPI paddlewheel type digital flowmeter. It has a substantial pressure drop, and I compensate with a beefy pump (Little Giant 3E-12NYS)

All this was ebay/surplus/leftovers from BillA. Not sure about the YSI 70x thermistors, but there are some cheap digitecs on ebay now.

re: 1-2C error. We have gone over this before; start propagating the error forward in the calculations and you will see why better numbers are needed (Uncertainty proportional to sq root of the sum of squares).

2) Maybe we could get Bill to put some of his tests with Swiftech bps of differing thickness and material? I think he has changed his testing methods since the testing was done, but he might be convinced

3) Also cheaper just to buy a faster CPU (at least with AMDs). Stock CPUs run 60C no prob; why not just stick with whatever free MHz you can get with the retail heatsink and thermal pad?

[jaydee]

Originally posted by pHaestus
[b]1) To be useful, good numbers for flow rate and wb inlet and outlet are needed at the minimum. Drilling out baseplates and getting a bp temp would be suggested as well. Dunno how useful CPU temps are yet; playing around still with better gear.

Here's what I use currently:

For water temps (wb in and out) and air temps (rad in and out) I am using a Digitec 5810 thermometer and YSI 700 series linear thermistors. Not perfect, but two of the water probes read exactly the same (to 0.01C) over the range of water used.

For CPU temps I use a Maxim 6655EV System from dallas semiconductor. I bought it because the homemade readers are limited to the resolution of MBM (1C) while the maxim software is 0.125C (over || port). It took some fiddling to get it working acceptably.

For wb testing, I am primarily relying upon the delta T across the waterblock at this time as I still don't know how relevant the diode temperature of the CPU is. I am pretty confident a superior block will put more heat into the water though.

Flow rates are basically "the master variable" for wb testing, and I use a GPI paddlewheel type digital flowmeter. It has a substantial pressure drop, and I compensate with a beefy pump (Little Giant 3E-12NYS)

All this was ebay/surplus/leftovers from BillA. Not sure about the YSI 70x thermistors, but there are some cheap digitecs on ebay now.

Ok, not exactly what I am looking for. I will try to make this more clear. I am not looking for temp reading accuaracy for testing water blocks on a professional level. I am talking about a senario where someone has a setup that just want to get a decent CPU temp. Something that does not require flow meters and all that. Something the average water cooling user can use to take temps without all you guys saying it is not right. Otherwise there is absolutly no point in anyone telling their temps at all because you all will say they are not measured right which in some cases is clearly obvious they are off but in some it isn't so clear that they are right or wrong even if they seem off base. Also I would much rather have a digital display show the temps as opposed to mbm on the computer. Is that possible?

re: 1-2C error. We have gone over this before; start propagating the error forward in the calculations and you will see why better numbers are needed (Uncertainty proportional to sq root of the sum of squares).

Yes, I know this, but to the average user who just has one setup it is really imaterial and these are the people I am trying to refer to as I am one of them. Maybe this miscommunication is what been causing all the arguments lately. Because if you guys want to make this a strictly professional forum then you will not get many posters.

2) Maybe we could get Bill to put some of his tests with Swiftech bps of differing thickness and material? I think he has changed his testing methods since the testing was done, but he might be convinced

Well I am not sure if that is what I ment either. I would like to see it explained in plain idiot english what happens when heat or cold is applied to Copper and AL. Like a good example of why copper spreads the heat betterthan AL.

3) Also cheaper just to buy a faster CPU (at least with AMDs). Stock CPUs run 60C no prob; why not just stick with whatever free MHz you can get with the retail heatsink and thermal pad?

Well this was a real constructive answer that has zero to do with the question.

I can see why things get so miscommunicated around here and ending up in disaster.

[Joe]

To everyone.. just my take on all this banter lately…

I have noticed a lot lately that there seems to be no patients for any "noob" who doesn’t know cooling inside and out. This could also because people claim numbers or qualities of products they know nothing about. It’s the same in every industry. I see it in Automotive, IT, Media, and now cooling; there is always someone who claims something as fact when it’s really fiction. Most of the time its simply because the person doesn’t know enough of the technology to validate the comments he’s posting. That on its own is bad, but not as bad as people who see this and choose to tear the person apart and not offer any suggestion to really help the kid.

Now if comments come from someone established, and who SHOULD know what’s going on before posting facts and numbers. There is less tolerance with that because to become established and to want to be seen as that in the field, you need to research before you post. But even at that, helping people understand the technology is much better than rubbing their nose in it. People aren’t dogs, rubbing their nose in the problem and humiliating them wont fix it. What it will do is just build a wall between that person and someone who may really know what’s going on, but is just too arrogant to spell it out. Its also much more simple for people to act like assholes than to be helpful.

We are here to learn right? What I would like to see is more constructive debate, or criticism, and less arrogant banter between people who all think they are right and wont come down to a level of communication where people can actually learn.

Tolerance. It may take little patients or time to spell some stuff out... and may go against that whole ego thing... but in the end I think it will be much more productive.

Also... NOT EVERYONE NEEDS TO AGREE!. There are multiple schools of thought. Some folks are concerned about different aspects than others. Some are engineers, some are purely performance driven and could care less about engineering. Not everyone has a Ph.D... Not everyone wants one

I dunno, I am getting off the soap box... now you kids play nice.

I also may be wrong.. it happens..

[pHaestus]

Jaydee to comparatively test waterblocks then flow rate measurement is NOT optional. This goes back to the whole nature of empirical observations. Perhaps I should clarify:

Empirical: the theoretical calculations are not available to do things purely from first principles (ab initio) and so experimental data is required and a model is built that fits this data.

If you were to test a lot of waterblocks without measuring and controlling flow rates then you will find performance differences that are dominated by the flow rate of the block. Does this mean block a is better than block b? No it just means that for block b that perhaps a different pump is required. Gotta be careful with empirical stuff or you'll get burned (see "copper takes in heat better but aluminum gives up heat better")


re: CPU temps. Working on it for AMD; Nevin at AS had some interesting pdfs from Intel (check my worklog) regarding variability of CPU die temps on P4s

Also I would much rather have a digital display show the temps as opposed to mbm on the computer. Is that possible?

Sure. The DIgitecs aren't actually integrated into MBM . For a prefab solution the Crystalfontz 633 with a probe epoxied under the center of a ceramic CPU is pretty good; make sure either as much airflow as possible or absolutely none around the socket. You could do the same with a Fluke 2190 series thermometer with a Type T tc or hell even a compunurse since 2-3C error doesn't bother you at all.

re: number of posters and professional testing. Yes the majority of overclockers are excluded from producing quality data. This doesn't trouble me; part of having a college degree in a technical field is an elitist attitude. They don't let you graduate without one That doesn't exclude overclockers/enthusiasts from the discussions; it could easily work as there is an interesting question posed and the techincal people work to address it. It breaks down when the friendly nature of the forum is lost though.

re: water chiller. You were skirting the issue of wb performance by adding a chiller to the mix to achieve a higher overclock. I skirted the issue by just adding a faster CPU. Same difference; skirting the issue in either case.

[jaydee]

Originally posted by pHaestus
Jaydee to comparatively test waterblocks then flow rate measurement is NOT optional. This goes back to the whole nature of empirical observations. Perhaps I should clarify:

Empirical: the theoretical calculations are not available to do things purely from first principles (ab initio) and so experimental data is required and a model is built that fits this data.

If you were to test a lot of waterblocks without measuring and controlling flow rates then you will find performance differences that are dominated by the flow rate of the block. Does this mean block a is better than block b? No it just means that for block b that perhaps a different pump is required. Gotta be careful with empirical stuff or you'll get burned (see "copper takes in heat better but aluminum gives up heat better")

Yes, I know this and I am not arguing what you are saying asI agree. But this does not apply to what I am trying to get answered. I am talking about someone that has a water cooled system already hooked up and just want to see his temps more accuratly so he will not get nailed when he tries to bragg his system off.

re: CPU temps. Working on it for AMD; Nevin at AS had some interesting pdfs from Intel (check my worklog) regarding variability of CPU die temps on P4s

I read that already. Interesting stuff!

Sure. The DIgitecs aren't actually integrated into MBM . For a prefab solution the Crystalfontz 633 with a probe epoxied under the center of a ceramic CPU is pretty good; make sure either as much airflow as possible or absolutely none around the socket. You could do the same with a Fluke 2190 series thermometer with a Type T tc or hell even a compunurse since 2-3C error doesn't bother you at all.

This is definatly what I am looking for! I would like to get more around .5 accuracy though just to say I can. So would the Crystalfontz 633 or the Fluke 2190 do this?

re: water chiller. You were skirting the issue of wb performance by adding a chiller to the mix to achieve a higher overclock. I skirted the issue by just adding a faster CPU. Same difference; skirting the issue in either case.

You know damn well that real overclockers are not going to buy a faster CPU. No matter how impractical it is it just doesn't work that way and you know it! So your answer was taken sarcastically by me. If it wasn't ment to be I am sorry.

Also if your answer was serious then pretend you already have the fastest CPU you can buy and want to continue. Remember the question is ment for overclockers not water block makers.

[pHaestus]

My opinion is that for the average overclocker the best thing is to take a good radiator air intake temp, water (wb inlet) temp, and CPU diode temp (with some care to how the diode reader is made). Intel chips seem to have some problems regarding the diode, and AMD diode readers need to be built carefully. I have an Epox 8K3A (supposed to read diode on XPs) and still haven't put an XP in it to test it out (in the mail). I evaluated the GTower case with the included digidoc and a diode reader; the important thing for overclocker is how changes in case airflow, tubing layout, components, and such affect performance. However, you can't get at questions such as "which is better: Cu or Al?" or "what parameters affect wb performance most?" with such equipment. Always consider what the goal is first.

CrystalFontz 633 is 0.5C accuracy, 0.125C res. With their newest version of software you can make the LCD scroll things like "Accuracy si for teh Loosers!" or whatever you like (mine occasionally has (oYo) roll across the screen; spices up staring at decimal places a bit) . A little overpriced, but for a temp junkie it is recommended (32 temps!)

Be aware it is still powered by the computer and connects to a serial port (or USB with adapter).

Fluke 2190 is better (0.2-0.3C accuracy) no ascii boobies though.
Check ebay for this one.

Yes sarcasm. I would add a pelt if I wanted substantially lower temps, and choose block accordingly (based upon pelt cooling).