The Cooling Solution

[bigben2k]

Here is where we discuss how to maintain the coolant temperature.

Off hand, we could use a chiller, or we could slap together some rad/fan combo, to which we could add a circuit to control the fan speed, based on the coolant temp. There may end up being some temp fluctuations, but carefully tuned, it might work.

Opinions? Thoughts?

[jaydee]

This is a tough issue. You can either fork over a small fortune for a self controling water temp controller and not worry about ambient air temp (got one at work to control the temp of the laser tube, about $8,000), or go with a huge or multiple rads with a thermaly controled fans and "hope" it keeps the water temp conistant. Problem is huge temp swings would kill you and you would still have to maintain a fairly steady ambient temp.

Is there less expensive "chillers" out there? The one I use at work is used to keep the 80watt laser tube at whatever you set the dial to. Now it is set to 23C and it stays at 23C no matter what the room temp is. Room temp swings about 20F from morning to after noon.

[nikhsub1]

To properly test the WB's, a rad setup will not be accurate at all. You NEED the water temp to be exactly the same all the time. For example, you need a chiller that can sustain a 25C water temp out of the chiller unit. This takes the guesswork out of what water temps are. With this comes a host of other issues. The need for accurate water temp measurement is one for example. IMO it would be a waste of time to attempt this project without the consistant water temps that a rad setup most likely won't give you.

[cybrsamurai]

Do radiators cool at a predictable rate with variations in ambient temp?

For instance in Alaska its 60F in Cali its 110F would it look like this: Alaska the water temp is 64F in California its 114F

Or would it be more random based on the ability of the equipment in a specific range of heat: Alaska the water temp is 61F in California its 122F


This is probably a gross simplification but I would suspect it would be closer to the latter. If this is the case it would be a requirement to chill the water with something and keep the system as isolated as possible to negate properties of lower and higher environment temps.

[bigben2k]

Some excellent points.

Even a chiller may have fluctuations, it really depends on each unit.

Rads do have a predictable response, but it depends on more than one element: airflow rate, coolant flow rate and temperature differential.

So from my perspective, it would be technically possible, although admitedly very difficult, to use a rad fan combo. The largest issues with it, is that the circuit may need to have a pretty quick response time, and that response needs to fall "in tune" with the rad/fan combo's response.

So to build something like this, might require a very accurate and quick response temp probe, just to tune the thing!


Either way, the "cooling solution" must be able to maintain the same temperature, and handle a heatload of more than 70 Watts. (Are we going to test 100W?)

So I'll ask the questions now:
-What kind of temperature fluctuations are acceptable (ex: speed, shape, form)?

-What rate of fluctuation would be acceptable (ex: +/- 0.1 deg C, with +/- 0.1 deg C/second)?

-If there's a fluctuation, but the average is still maintained, are we still OK?

-What exactly is the effect of having a fluctuation?


Then I've got my own questions:
-What is the effect of "relative humidity" on the testbed?

-Could it affect a circuit, such as described above?

-Alternatively, could we use a mixing valve, to mix hot and cold water, to get to that temperature? (answer: it depends on the response time of the valve)

-Are we testing at ambient +5 deg C, regardless of the room temp, or should we require a specific room temp, or a specific range of room temps? What's the difference?

[nikhsub1]

I know there are inexpensive chillers (can't think of the name off hand) that would fit the bill perfectly. I don't think humidity has any effect on non-living objects (heat index) although I could be wrong but I don't think so. I think 100W should be the tested wattage, prcocessors are just getting hotter and hotter and 100W makes it a nice round number and easy to calculate the differences in various C/W's.

Ben, wouldn't the 'effect of fluctuation' be fluctuating results? I would think a fluctuation of .1C would be OK, but hopefully no more.

As far as equipment goes, has anyone asked BillA for recommendations of decent equipment with affordable prices? I know we can look at all his equipment, but we know it is all rather expensive too.

[bigben2k]

That's where we come in. A variance in the water temp would indeed result in fluctuations in the results. Obviously that's going to depend on the block design, so the question then becomes: what fluctuations are acceptable?

Do we want to eliminate the fluctuations, or have a minimal amount of it? Can't we just set the level of aceptable fluctuations based on the error margin of the rig? (which would have its minimum error margin)

As we know, the best block out there right now is the Cascade, so it could be very likely that it would be most prone to fluctuations than any other block.

But I don't have any testing data to answer that. Anyone?

[nikhsub1]

Ben, the way I see it, is that even the Cascade would only fluctuate as much as the water temp fluctuated, all else being equal. So a .5C rise in water temp, again everything else equal, you should see a .5C rise in the die temp. I would think that would be true for all blocks though, some may take longer to register the rise/fall in the water temp, but at some point it should be linear. Of course, I could have my head up me bum too...

[winewood]

I don't understand how an accurate water temperature measurement reading at each increment of readings wouldnt do the same thing as a chiller. It makes the math harder, but really, is it worth it? All your doing is 'almost' eliminating' a variable, but you still cant discount it all together anyway... can you?

Sorry, but I need some help to get up to your guys thought processes.

[bigben2k]

I understand what you're saying nikhsub1, but my question is: how fast would that difference show up on the CPU?

If it takes say, 60 seconds for the CPU temp to drop by 0.5 deg C, just like the coolant, then as long as my cooling solution doesn't vary by more than 0.5 deg C per minute, I won't see a difference, and I have that parameter for the cooling solution.

I just used the Cascade as an example, because if it's the best performing block, then it's more than likely to allow that temp change over the shortest period of time. This is where direct die cooling comes in handy, because it can give us some pretty interesting data, even if the application isn't practical.

On the other hand, if we use a heat die, and it's made of copper, then that fluctuation may be more dramatic.

But again, I have no test data.:shrug:

(BTW, congrats on the silver Cascade )

[nikhsub1]

I think the only way to remove this variable is to have ultra-accurate temp monitoring and complete temperature stability in the water to the block. Isn't the theory that if you want within .1C accuracy you need a measurement increment of .01C? I just am clueless as to how much these kind of instruments cost.

[Aardil]

ok I may be way off base (wouldnt be the first time) but I think the water temp can be solved fairly easy. Please dont laugh, here is my thought:
A large Res with a simple cjiller should keep the water temps in check.
Let me explain, I have a 125 gallon salt water tank, I use a small chiller to keep the temp down so my salt fish and sea life stay happy. Even with a hi output pump ( tank is running a little giant pushing 600 GPH, the temp stay within a degree or 2F of constant. this is with five 6 foot high UV output floresent bulbs shining in the tank for a minimum of 10 hours per day.

So since we seem to only be using about half the GPH pumps in WC and with a 30 gallon tank with chiller I would think the temp variation would be almost nothing.
Just trying to help
FLAME AWAY
Aardil<---- Runs for cover

[pHaestus]

You will find that water temp variations are not trivial to deal with. To get 2.5-3GPM through a flowmeter, radiator, waterblocks, filters (you SHOULD use one to protect the flowmeter from teflon bits and other junk), and valves and crosses and plumbing will take a powerful pump. Your powerful pump will make the water temps increase as you throttle it. Even with a chiller you'll probably have to adjust the set point downwards to keep temperatures constant over the whole flow range.

I have had some luck using a rheostat and fans on a radiator. I check out equilibrium water temperature at the lowest flow rate first with fans wide open. I then back the fans off a bit so that I have a bit of margin for error and use the water temp at that point (probably 80% power or so?) as my "set temp" for the water. As I increase flow rates, I can lower fan speed to arrive at the same final water temp. It takes a ton of fiddling and is not user friendly at all. I have my testing equipment in the basement room on a table that doesnt face any external walls, and I have a couple of house fans blowing onto the testing area at all times.

It is basically the opposite of "automated" but it can work with practice. I bought a PID controller thinking I could make it PWM the fans using a type t thermistor in the water as it's input (compare it to set temp and PWM as needed) but I haven't gotten it to work yet.

If you are serious about getting a chiller then search for "recirculating water bath" and get one with heating and cooling capabilities. You will be no better off than I am with the rheostat unless you get a digital one with good temp regulation. The one I have at work for reaction kinetics is +/- 0.1C from 5 to 60C and was a vwr brand. Haake and Lauda make good units but were more expensive. I think I paid $2000CAD for my unit but it was listed for much much more. I added it onto a sizable order to get a break

[Aardil]

PH,
I understand what you are saying and while I am in complete agreement with you, my point is with a rather large res ( say 30 to 40 gallons) precooled to 25C, I would think you should be able to complete the test of said block before a noticable rise in temp that the chiller cant overcome. While I doubt you could run it for a 24 hours test, I think it would keep up for a couple hours at least.
as far as the pump induced heat I agree it would need to have an adjustment to set the temp at start according to flow rate and pump induced heat.
Like I said I may be wrong, but thats what I see.
Aardil

[pHaestus]

I have been playing with 1 hour per point at 0.5, 1, 1.5, 2, 2.5 gpm. That's 6 hours by the time all the fiddling with flow rate adjustment and fan speed is done. I am limited to Saturday and Sunday tests as a result (I stay up laaaate those nights anyway). Automation would be a worthwhile topic for your test design. Costs big bucks to do right though (ask Bill how he likes Labview too).

[Aardil]

ok lets take this one step farther,
my chiller is in a wet dry filter system.
The wet dry would act something like a bong set up removing heat , then passing through the filtering system (IF you remove the charcoal bags and leave only the foam filter wich is as wide as the tank and 3 inches square thus low restrition ) and the chiller taking care of the rest of the heat before returning to the pump and block.
the wet dry would need a few mods to do this but not on the scale of thousands of dollars.
While I know salt water fish and reef tanks are not quite as critical to temp as Wc block testing, it is important enough that it has to be done and 15 dregees below ambient ( around 60 degrees F ) is what they try to achieve.
I know I am probably way off base and you can tell me to STFU anytime, I know alot more about salt aquariums than I do water cooling PC's. just trying to apply same basic technology since that is what I know.
I know I am reaching but it may be that extra little bit to make the temp stable.
Aardil

[bigben2k]

(a prepared reply!)

Nikhsub1: A chiller is going to be expensive, at least that's my prespective on it. Jaydee116 hinted at $8'000, and I think that most chillers are going to run in a similar range. You may still be right though.

The reason that I push this point, is because I look at it from the perspective of the WBTA, which is to set testing standards, and not so much to impose the type of equipment needed, to some extent.

So if we can define those parameters, then that leaves the test bench builder the option of going with whatever solution he sees fit, as long as it meets the standard.

Isn't the theory that if you want within .1C accuracy you need a measurement increment of .01C?

All I know for sure, is that all the errors are cumulative, so the less we have at each point, the more accurate the end result is going to be. We can take this one up in the accuracy thread <here>


Winewood: yes, you right, but the temperature still needs to be maintained.

To recap:
We went from "it must be a chiller because it's perfect" to "the temp needs to remain rock solid".

A chiller may still induce temperature fluctuations. So the question became: how much of a fluctuation is acceptable?

Then I suggested that the acceptable level of fluctuation is going to depend on how a block would respond to it. So I pointed out that if we knew how fast a Cascade responded to a temperature fluctuation, we could keep the fluctuation under that, and set the standard.

Then I suggested that those fluctuations may be more dramatic on a heat die, because of its nature (being possibly copper, instead of silicone). In other words, even if a CPU didn't see the fluctuation, because it went by so fast, a copper heat die might still catch it.


Moving along...

There's two things I have to mention here:
1-It's possible, and I've thought of this, that we may need some ultra accurate measurement tools, not for testing, but for building the test bench.

2-There might have to be an acceptable level of temp fluctuations, as measured at the heat die. Then again, it might be possible to keep it at zero fluctuation, even if the coolant temp fluctuates by some temperature range, over an X period of time. If we target zero fluctuations, which I think (off hand) may be possible, as measured at the heat die, then we 'd have to define the timing or frequency of those temperature measurements.

I feel like I'm just running myself in a circle here. The answer may be absurdly simple: keep the die temp steady.

The problem is that if I only measure the die temp say, every second, I may be missing a big fluctuation. So I guess what I'm eventually asking is, how often (what frequency) do we measure the die temp?

So I'm back to: it depends on how much of a fluctuation you expect from the block, which in turn, leads back to how much of a fluctuation the cooling solution is going to induce.

Aargh!

Ok, let's try this: the magnitude of the fluctuation of the temp measurement at the heat die, has to be zero. In order to make sure that it is indeed zero, we need to sample that temp within time interval "X". Time interval "X" depends on the block, and the cooling solution.

Given that the heat die temp probe will return a result of say 35.7 deg C, that the room temp is a steady 25.0 deg C, all with an error margin of +/- 0.1, then the coolant temp has to remain between 24.95 and 25.05, on average.

(I'll think about it some more. As pHaestus said: it's not trivial)


If we use a rad/fan combo as the cooling solution, where the speed of the fan is controlled by a simple rheostat which is set once so that the coolant is at 25.0 deg C, then the room temp cannot fluctuate, otherwise the coolant temp is going to fluctuate too.

Unfortunately, most AC or a furnace react within a margin of 2 or 3 deg C (that's so the units don't come on and off constantly). In short, the ambient temp can be expected to fluctuate. This is why a chiller is really handy, because it handles that very easily.

In a rad/fan combo, there needs to be an additional circuit that controls the fan speed, to compensate for that temperature swing. IF it's driven by the water temp alone, then the circuit is going to allow some fluctuation, in order to react. If it's based on the air temp, then it can hold a temperature steady, given that the circuit is designed in such a way that it knows exactly how much faster a fan must turn, but it still must know the water temp, as part of that scheme, and it cannot allow the temp to fluctuate by more than 0.05 deg C, which is half the error margin of the heat die temp probe.

Possible, but not easy.

More ranting later.

[pHaestus]

You need to start by making the following statement be set in stone:

"All temperature readings reported are under steady-state conditions"

This is a requirement!

Now, how can one arrive at steady-state? Over the length of the data collection then water, air, and die temperatures cannot change appreciably. Not so hard for one point on a C/W vs GPM curve; can get to be a big pita for 5-6 over several hours.

You have the cart before the horse though. The first step is to set a water temperature. Everyone will need to use the same one, you realize (right?). Assuming your die simuator is perfectly insulated (heh) then I don't think room temperature directly affects the results. The difference between room and water temperature though could have an effect on the observed delta T. Can the equipment being used pick this up? Not sure.

[jaydee]

Originally posted by bigben2k
(a prepared reply!)

Nikhsub1: A chiller is going to be expensive, at least that's my prespective on it. Jaydee116 hinted at $8'000, and I think that most chillers are going to run in a similar range. You may still be right though.

That's for a very very good one. The laser requires a perfect temp to maintain a quality beam. If the laser tube is allowed to warm up or cool down during the engraving it can produce bad results. It is a very good chiller. I am not so sure you need one that big for this job as it is cooling an entire laser tube which puts out well over 100watts of heat. Also this thing is huge, It is about 3 feet long and 2 feet high and 1.5 feet wide. Not light either. Guess one could browse google for info.

[pHaestus]

look here

I'd look at ebay and labx. Mine was about $2000CAD new (1200US or so) but I think that is about half retail price at VWR. No I can't take it home and use it; it's in constant service in the lab.

//edit: long url

[bigben2k]

Originally posted by pHaestus
You need to start by making the following statement be set in stone:

"All temperature readings reported are under steady-state conditions"

This is a requirement!

Now, how can one arrive at steady-state? Over the length of the data collection then water, air, and die temperatures cannot change appreciably. Not so hard for one point on a C/W vs GPM curve; can get to be a big pita for 5-6 over several hours.

You have the cart before the horse though. The first step is to set a water temperature. Everyone will need to use the same one, you realize (right?). Assuming your die simuator is perfectly insulated (heh) then I don't think room temperature directly affects the results. The difference between room and water temperature though could have an effect on the observed delta T. Can the equipment being used pick this up? Not sure.

Well put.

I have to agree that the die temp measurement really needs to be steady, at least within the error margin of the temp measurement. (i.e. it can't fluctuate)

I also agree that the water temp should be maintained at 25.0 deg C, as Bill has been doing. It's a fair representation of an average cooling solution (some people go as low as ~1 deg C (delta T between water and air), some people use a BIM with a small fan ), as 5 degrees above an ambient of 20 deg C.

I think that in general, anyone should be able to reach temp measurements within +/- 0.1 deg C, so let's use that assumption, for now (it's not a given, I just need a basis for discussion).

So now I'm down to (where I should be) how we measure the performance. At flow rate X (still needs to be defined), we're going to report a C/W. W is going to come from the measurement at the heat die and C is the temp delta (aka delta T) between the die and the water.

So obviously, we don't want fluctuations in the water temp either. So I'm back to my earlier statement:

Given that the heat die temp probe will return a result of say 35.7 deg C ... all with an error margin of +/- 0.1, then the coolant temp has to remain between 24.95 and 25.05, on average.

Regardless, we want a steady water temp, at least within 0.1 deg C (the accuracy of the water temp probe).

And if we have +/- 0.1 deg C at the water and at the die, the resulting delta T is going to have an error of +/- 0.2 deg C.

An example: over a result of 0.20 C/W (not typical!), and given a 70 Watt source, the C/W is accurate to +/- 0.003 C/W. Then we add the error from the heat die power measurement, say 2%, and the accuracy is at +/- 0.007. Then there are other errors that will need to be factored in (all of this is theoretical!).


The thing is, I have no problem running a log of the temperature readings, but I don't know what sampling rate to use. The cooling solution is going to be a source for fluctuations, and that's what's really going to define that sampling rate, no?


I really like your idea of a PID merged with a PWM, for a rad fan combo. That's the cooling solution that I, personally, would like to focus on.



JayDee116: do you have any specs on that chiller? We're looking for "expected fluctuations", both in temp range, and over a period of time.

[pHaestus]

In a 22C room, my Little Giant pump kept the loop's water at ~ 26-27C with fans wide open and flow rate not throttled (~2.5-3gpm). 25C is not going to be attainable for water temps for testers using a radiator and fans. Might want to go with 30C instead (still well within acceptable temps for pumps and tubing).

[bigben2k]

Tell me again why that temperature needs to be set at a specific point for everyone?

[jaydee]

Originally posted by bigben2k




JayDee116: do you have any specs on that chiller? We're looking for "expected fluctuations", both in temp range, and over a period of time.

Unfortunatly no. It is a prepackaged deal with the manufacture of the laser. So I assumed they decided we really didn't need to know the capabilities of is as we shouldn't be messing with it.

Best I can give you is a brand name (Poly Science) and it is one of two series:

http://www.polyscience.com/indust/14to1chillb.html

or

http://www.polyscience.com/indust/14to1chill.html

[bigben2k]

Found it (Thanks JayDee116!)

On page 4 of the "Operator's manual" (here, PDF, 14 pages), it states:

General Specifications to all chillers

-Temperature setability +/- 0.1 deg C
-Temperature uniformity +/- 0.5 deg C
...

So the temp may fluctuate up and down by as much as 0.5 degrees C.