Options on making this load tester

[_HL4E_HalfLife_]

I picked up 2 crankcase heater cartridges that were in large semi-hermetic compressors and i want to turn them into a load tester for all my phase change/chiller systems and finally give u guys some loaded temps from all the systems i've made over the past few years.

I'd like ur guys opinions on the best way to turn these 2 heater cartridges into a single load tester.

http://img444.imageshack.us/img444/9520/heatercartridgent6.jpg (http://imageshack.us)

http://img297.imageshack.us/img297/6357/heatercartridge1ya1.jpg (http://imageshack.us)

Here is a drawing of how i was going to make it but i thought i'd get ur guys opinions first before i went ahead and made it.

http://img100.imageshack.us/img100/7884/loadtestideaug7.jpg (http://imageshack.us)

[sacha35]

I would put them in side ways into a sold copper block rather than though the bottom, also how will you be adjusting the amount of current to these heater cartridges and measuring the amount of watts they are putting out.

[[XC] gomeler]

Those are physically a bit large for a DD load-tester, going to need a large bar of copper to fully encase them but you'll get the best results that way. I myself just bought some wirebound resistors and a variac and I'm hoping to be able to go from 0 watts to ~350 watts with these particular resistors. Hope I don't blow any of them up by running them out of spec :p: Get yourself a multimeter that measures current or a kill-a-watt, either will do.

[_HL4E_HalfLife_]

It says right on them there 85watts each. I just plan on using a 3 way switch so can either select off,1 HC,or both HC's and in combination with that i may use a dimmer switch to adjust the heat output.

My CPU's rated at 86watts so just having 1 of these HC's on will be perfect for simulating the heat output of my CPU under load.

Sacha35 - I don't have a solid block of copper so instead im gonna use some concreate patch mix to fill in around the 2 HC's. I did something like this in another project so i'll post a pic of what i mean.

http://img156.imageshack.us/img156/4648/concretearoundevapplatevd7.jpg (http://imageshack.us)

[wdrzal]

since they are resistive heaters you can boost the voltage to gain wattage.

forget the concret,at least your design. what dia. x length ???? what your power in your home 60 hz or 50 hz then I,ll know the voltage available without a transformer.


go to this site below and enter 169.412 under resistance and 120 under volts and hit calculate. read watts

next enter same resistance 169.412 but 220 volts then calculate then read watts.

http://www.chromalox.com/technical/ohmscalculator/

[[XC] 2long4u]

What was that thing you made and what did you use it for?

[Ssilencer]

http://img518.imageshack.us/img518/5632/dado5cj4.jpg

[_HL4E_HalfLife_]

since they are resistive heaters you can boost the voltage to gain wattage.

forget the concret,at least your design. what dia. x length ???? what your power in your home 60 hz or 50 hz then I,ll know the voltage available without a transformer.


go to this site below and enter 169.412 under resistance and 120 under volts and hit calculate. read watts

next enter same resistance 169.412 but 220 volts then calculate then read watts.

http://www.chromalox.com/technical/ohmscalculator/



Wdrzal im suprised that u don't know what the hz is here in Canada:stick: it's 60hz and 120v. The size of the HC is 4" long and has a diameter of 2/3 of an inch.

Ps. it wouldn't be 220v, 120+120=240.

Ps.Ps. I have no idea what u mean by "forget the concret,at least your design"

[wdrzal]

concrete is a fairly good insulator ,you need to conduct the heat from all around to copper heater or a good conductor other wise you will burn out the element from over heating. put under phase it will handle the higher voltage you really only need 250 to 300 watts .so 220 or 240 volts is not needed ,changing the voltage is how you will control and measure watt output. lower voltage = less heat=longer element life. temperature will change resistance.

information in avatars is not relevant to me so I don't look at them .especially where you live or post counts because its not worth wasting brain cells on because the info is not verified, and I don't care about post counts or where you live.;)

I try to be a equal opportunity annoyance :p: :p: :p:

[_HL4E_HalfLife_]

concrete is a fairly good insulator ,you need to conduct the heat from all around to copper heater or a good conductor other wise you will burn out the element from over heating. put under phase it will handle the higher voltage you really only need 250 to 300 watts .so 220 or 240 volts is not needed ,changing the voltage is how you will control and measure watt output. lower voltage = less heat=longer element life. temperature will change resistance.

information in avatars is not relevant to me so I don't look at them .especially where you live or post counts because its not worth wasting brain cells on because the info is not verified, and I don't care about post counts or where you live.;)

I try to be a equal opportunity annoyance :p: :p: :p:

If concreate conducts heat well won't it conduct cold just as good??

I dono why u guys keep thinking we need load testers that put out 250 to 300watts when no cpu that i know of will put out that much heat, Heck im using a p4 661 3.6 and at stock speeds its output is only 86watts. I got it overclocked to 4086Mhz and thats still only 98watts on a Pentium4!!! Even the QX6800 is only 130watts thats a far cry from 250watts.

[wdrzal]

who said concrete in a good conductor,

I said insulator

you may want to gander at this sites list of common materials conduction . higher number is best ...........lower is worst


http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html

copper is 401 may want to look at cement and concrete products.

[[XC] gomeler]

If concreate conducts heat well won't it conduct cold just as good??

I dono why u guys keep thinking we need load testers that put out 250 to 300watts when no cpu that i know of will put out that much heat, Heck im using a p4 661 3.6 and at stock speeds its output is only 86watts. I got it overclocked to 4086Mhz and thats still only 98watts on a Pentium4!!! Even the QX6800 is only 130watts thats a far cry from 250watts.

You kidding me? I don't know how much overclocking you do but 86 watts is nothing.. take for example my e6400 which is rated from Intel for 65 watts. @ stock speeds the entire system(in rig) consumes 143 watts running full CPU load/no GPU load. Bumping the clocks up to 4GHz and the entire system consumes 227 watts @ full CPU load. 65 watts(stock value) + 84 watts(difference between stock system and OCed system) = 149 watts. Even if it's only pulling 40 watts @ stock it's still sucking down 124 watts. Now double the cores and voila, there's you 250 watts. This is only with 2 cores at 4GHz also, now imagine 4.5 GHz, easily get up to 280+ watts. These results were also recorded with abnormally low voltages, 1.25V @ stock and 1.475V@ 4GHz(wonder chip :slobber: ). This is why we need loadtesters capable of 300 watts.

[_HL4E_HalfLife_]

who said concrete in a good conductor,

I said insulator

you may want to gander at this sites list of common materials conduction . higher number is best ...........lower is worst


http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html

copper is 401 may want to look at cement and concrete products.

The concreate patch mix seems to work really well on my Cold Plate Cascade that mix is what's in contact with the copper lines and the plate itself and it gets cold pretty darn quick.

If u don't think i should use the mix than by all means walt lets hear ur idea.:stick:

[_HL4E_HalfLife_]

gomeler;2185240']You kidding me? I don't know how much overclocking you do but 86 watts is nothing.. take for example my e6400 which is rated from Intel for 65 watts. @ stock speeds the entire system(in rig) consumes 143 watts running full CPU load/no GPU load. Bumping the clocks up to 4GHz and the entire system consumes 227 watts @ full CPU load. 65 watts(stock value) + 84 watts(difference between stock system and OCed system) = 149 watts. Even if it's only pulling 40 watts @ stock it's still sucking down 124 watts. Now double the cores and voila, there's you 250 watts. This is only with 2 cores at 4GHz also, now imagine 4.5 GHz, easily get up to 280+ watts. These results were also recorded with abnormally low voltages, 1.25V @ stock and 1.475V@ 4GHz(wonder chip :slobber: ). This is why we need loadtesters capable of 300 watts.

I'm sorry but don't C2D's already have 2 cores??? cuz if they do than doubling the cores would make it a quad core no? With ur single C2D CPU ur still only drawing 149watts like u said.

I'm tired now so im not sure if this is making any sence or not...

[[XC] gomeler]

By doubling the cores I was showing what it'd be like with a quad-core, re-read it in the morning and it should make sense. While dual-cores would have trouble breaking 200 watts loaded, quad-cores easily can hit 200 watts.

[sacha35]

The concreate patch mix is an insulator as wdrzal said, if you want this to work correctly then the only way is to put the resistive heaters in copper as this will transfer the heat they will put out more evenly and give you a more accurate dummy load.

As for what chips put out in the way of watts/heat, we are not talking about stock values as we only make or use Phase systems to see how far we can push these chips we have, this is why the you want to be able to set a given dummy load for a range of chips,so we can then estimate how well it will keep these chips cool, a 300watt draw can be reached on some chips these days, so if you are looking to make units for a range of chips then it would be better to have a load tester that can be varied up-to at least 300+ watts.

Don't take this the wrong way but I will stress though, as you are working with mains voltages you need to take all the precaution's possible to ensure that the device you make is safe and not just fit for its purpose, as electricity can Kill, also if there are problems with the load tester and you have a phase system connected to it and any leakage of current will be transfered though to the phase unit, making this live and also causing damage.

[Gray Mole]

Personally, I would try to find smaller cartridges.

You've got a lot of 'ground to cover' and that's not exactly ideal.

That said, if you're going to use them, try to apply the heatload in as concentrated a way as possible to the area you wish to heat.

I wouldn't use concrete, because it conducts too well to be an insulator, and conducts too poorly to be a good conductor.

I would build it exactly as you've stated in your first design pic shown, with a few changes.

You should really get a coldplate, just a small 5cm square, 5-10mm thick plate in copper.

Then, use a heat conductive epoxy between the ends of those cartridges and the coldplate, or better, machine the coldplate so the cartridge ends actually fit into the plate slightly and epoxy into place.

Once they are there, fabricate a 'box' to cover the rest of the cartridges, and use insulative expanding foam to cover the rest, if you have access to high temp stuff (just in case)

That would be quite similar to how you're laid out now, give you a copper surface for heat transfer to the evap being tested, and also prevent any heat from going outside of where you want it.

It's pretty much exactly how I made mine, but with flat ALU resistors.

I do like the loads you have though :) I think that if you're going to be stuck with a '2 stage' tester, then 86w and 86 x 2 = 172w is pretty good.

It would be ideal if you had one more, for a 86 x 3 = 258w load as well, for 'stock' load, 'high oc' load, and 'quad' load choices.

Fitting 3 of those monsters might prove awkward, but I think it's a decent setup.

Would just be nice if you could get more compact designed cartridges, as the heatload will be somewhat slow to respond.

Cheers

Gray

[_HL4E_HalfLife_]

Personally, I would try to find smaller cartridges.

You've got a lot of 'ground to cover' and that's not exactly ideal.

That said, if you're going to use them, try to apply the heatload in as concentrated a way as possible to the area you wish to heat.

I wouldn't use concrete, because it conducts too well to be an insulator, and conducts too poorly to be a good conductor.

I would build it exactly as you've stated in your first design pic shown, with a few changes.

You should really get a coldplate, just a small 5cm square, 5-10mm thick plate in copper.

Then, use a heat conductive epoxy between the ends of those cartridges and the coldplate, or better, machine the coldplate so the cartridge ends actually fit into the plate slightly and epoxy into place.

Once they are there, fabricate a 'box' to cover the rest of the cartridges, and use insulative expanding foam to cover the rest, if you have access to high temp stuff (just in case)

That would be quite similar to how you're laid out now, give you a copper surface for heat transfer to the evap being tested, and also prevent any heat from going outside of where you want it.

It's pretty much exactly how I made mine, but with flat ALU resistors.

I do like the loads you have though :) I think that if you're going to be stuck with a '2 stage' tester, then 86w and 86 x 2 = 172w is pretty good.

It would be ideal if you had one more, for a 86 x 3 = 258w load as well, for 'stock' load, 'high oc' load, and 'quad' load choices.

Fitting 3 of those monsters might prove awkward, but I think it's a decent setup.

Would just be nice if you could get more compact designed cartridges, as the heatload will be somewhat slow to respond.

Cheers

Gray

I can get another cartridge of the same rating but i can't get anything smaller than these unfortunitly. :(

[esdee]

i have a couple of cartridge heaters attached to my load tester, ill bring to my room from my lab to picture it for you tonight

Ssilencer i really really don't like your load tester, i m sorry for being that honest but with this gigantic mass you will have too much heat stored on the lower part of the copper thats not in contact with the evap/block

[Ssilencer]

i have a couple of cartridge heaters attached to my load tester, ill get it up to picture it for you tonight

Ssilencer i really really don't like your load tester, i m sorry for being that honest but with this gigantic mass you will have too much heat stored on the lower part of the copper thats not in contact with the evap/block

Well, I appreciate and respect your opinion.
Being that all is heavily insulated, heat can only escape from that big chunk of copper by the little dice on the top.
It works fine on waterblocks, I didn't used that one for evaps.

For evaps I did some testings with aluminum and yes, being that aluminum is not as good as copper to transport heat my current heat load tester for SS is not good at all as it is storing too much heat on the aluminum cilinder, but anyway, makes a brutal load (that I have to compare against a real cpu as soon as I can) and it is better than nothing in the meantime.

http://img518.imageshack.us/img518/7532/dummyload004mediumdp7.jpg
http://img518.imageshack.us/img518/8646/dummyload006mediumxv6.jpg

Anyway, it was a prototype for the independant part of our own load tester controller.

[esdee]

http://img246.imageshack.us/img246/3645/dsc7921mb6.th.jpg (http://img246.imageshack.us/my.php?image=dsc7921mb6.jpg)
http://img246.imageshack.us/img246/182/dsc7922hh1.th.jpg (http://img246.imageshack.us/my.php?image=dsc7922hh1.jpg)
http://img296.imageshack.us/img296/7105/dsc7930ar0.th.jpg (http://img296.imageshack.us/my.php?image=dsc7930ar0.jpg)
http://img296.imageshack.us/img296/5316/dsc7932zb5.th.jpg (http://img296.imageshack.us/my.php?image=dsc7932zb5.jpg)
http://img296.imageshack.us/img296/8400/dsc7934on8.th.jpg (http://img296.imageshack.us/my.php?image=dsc7934on8.jpg)


this is how i made my load tester minimum copper for fast heat up, and higher temps.
the heaters can be inserted in the holes applying a gentle force
the side that contacts the cpu is the left on the fifth image,
the distance between the heater and the evap/block is really close to the size of an CPU HSP,
i usually place 1x130w on the middle and 2x40w on the sides to emulate better the fact that the core is in the center and thus the most heat comes from the center
i have another copper piece with 4 holes for heaters that i use to emulate quad core which have more even spread heat across the HSP (2x(same wattage on the middle) and 2x(same wattage on the sides)

totally i have 5x90w, 5x130w and 5x40w heaters:D



Well, I appreciate and respect your opinion.
Being that all is heavily insulated, heat can only escape from that big chunk of copper by the little dice on the top.
It works fine on waterblocks, I didn't used that one for evaps.


yes heat can only escape by the point you say on the top, but if you place a thermocouple on the bottom of your loadtester im 100% sure your temp will be nowhere near like ambient so that you can assume all heat is desipated for the top:)

[Ssilencer]

http://img246.imageshack.us/img246/3645/dsc7921mb6.th.jpg (http://img246.imageshack.us/my.php?image=dsc7921mb6.jpg)
http://img246.imageshack.us/img246/182/dsc7922hh1.th.jpg (http://img246.imageshack.us/my.php?image=dsc7922hh1.jpg)
http://img296.imageshack.us/img296/7105/dsc7930ar0.th.jpg (http://img296.imageshack.us/my.php?image=dsc7930ar0.jpg)
http://img296.imageshack.us/img296/5316/dsc7932zb5.th.jpg (http://img296.imageshack.us/my.php?image=dsc7932zb5.jpg)
http://img296.imageshack.us/img296/8400/dsc7934on8.th.jpg (http://img296.imageshack.us/my.php?image=dsc7934on8.jpg)


this is how i made my load tester minimum copper for fast heat up, and higher temps.
the heaters can be inserted in the holes applying a gentle force
the side that contacts the cpu is the left on the fifth image,
the distance between the heater and the evap/block is really close to the size of an CPU HSP,
i usually place 1x130w on the middle and 2x40w on the sides to emulate better the fact that the core is in the center and thus the most heat comes from the center
i have another copper piece with 4 holes for heaters that i use to emulate quad core which have more even spread heat across the HSP (2x(same wattage on the middle) and 2x(same wattage on the sides)

totally i have 5x90w, 5x130w and 5x40w heaters:D




yes heat can only escape by the point you say on the top, but if you place a thermocouple on the bottom of your loadtester im 100% sure your temp will be nowhere near like ambient so that you can assume all heat is desipated for the top:)

It is impossible to get that kind of heaters here, they are really small.:toast:


yes heat can only escape by the point you say on the top, but if you place a thermocouple on the bottom of your loadtester im 100% sure your temp will be nowhere near like ambient so that you can assume all heat is desipated for the top:)

Sorry I don't understand that part

[gosmeyer]

HL4E HalfLife,

There is not a lot of common ground when it comes to load testers imo.
It just seems to me there are so many variables to look at, type/size of material for load plate.
Type/size(physical) for heater, insulation type and on & on.:confused:
Basically you want the least amount of mass with the least amount of leakage producing the desired amount of heat.
If it were myself starting from scratch, my starting point (if nothing else) would be based on the suggestions by those who
have already been there done that. With that said, I do encourage thinking outside the box. :rolleyes: It just creates a greater knowledge base.

Good luck,
Eric

[gosmeyer]

esdee
That nice. no more no less

[esdee]

It is impossible to get that kind of heaters here, they are really small.:toast:



Sorry I don't understand that part

yeah, language barrier...

if you place a thermocouple at the bottom of your tester you will see that it's a lot higher than ambient. that is some heat generated by the heaters than doesnt make it close enough to a place that can be affected by the evap/block temp

[wdrzal]

I,m going to explain in simple terms:
Think this way take a piece of solid copper 1-1/2 round, cut 1 piece to 1" in length,cut another 4" long. Now hold one in each hand and heat both ends equally. which one are you going to drop first????? answer:the short one.................why????

well copper is the second best conductor of heat but even the best is not perfect.

"A good conductor is a poor insulator" as is "A Bad conductor is a good insulator"
now of course you dropped the short one first, heres why,

First think of home insulation, there is a R value, just more thickness of the same material increases the R value(insulation). How fast heat can move from point A to B.


Second since you dropped one before the other, There is Time involved.

In the thermodynamic equations which are(quite complex for modeling heat flow thru irregular shapes,like copper with heaters in the middle) there are engineering equivalents to consider equal to the R (insulation factor,thickness) & T for the (time factor) for heat flow thru solids. plus many more factors those are just 2 of the simple ones.

Hope this helps

[Ssilencer]

I,m going to explain in simple terms:
Think this way take a piece of solid copper 1-1/2 round, cut 1 piece to 1" in length,cut another 4" long. Now hold one in each hand and heat both ends equally. which one are you going to drop first????? answer:the short one.................why????

well copper is the second best conductor of heat but even the best is not perfect.

"A good conductor is a poor insulator" as is "A Bad conductor is a good insulator"
now of course you dropped the short one first, heres why,

First think of home insulation, there is a R value, just more thickness of the same material increases the R value(insulation). How fast heat can move from point A to B.


Second since you dropped one before the other, There is Time involved.

In the thermodynamic equations which are(quite complex for modeling heat flow thru irregular shapes,like copper with heaters in the middle) there are engineering equivalents to consider equal to the R (insulation factor,thickness) & T for the (time factor) for heat flow thru solids. plus many more factors those are just 2 of the simple ones.

Hope this helps

I understand, but which is the hurry?
This is not a evap block, it is a heat element.
The copper is heavily insulated, 4mm asbestos plus 10mm polyurethane foam over it, also the copper piece cannot keep getting heat for ever, and the only part that has to flow is the little dice in the top.
The copper piece is as big as the cartridge heaters, not bigger.
Also, it`s purpose is to give a simulated heat over the block or evap that will be contrasted against other blocks or evaps, so, it is good enough for this.

[_HL4E_HalfLife_]

Frankly im out of ideas to turn these 2 HC's into a load tester:( There's only 1 place around here that i can think of that might have a solid block of copper, that would be the place u can take scraps of copper,alum,etc to for recycling and get some cash.

[[XC] gomeler]

onlinemetals.com but it might be expensive for the size you are looking for. Look over their barstock and roundstock sections, they'll have want you need.

[_HL4E_HalfLife_]

gomeler;2192285']onlinemetals.com but it might be expensive for the size you are looking for. Look over their barstock and roundstock sections, they'll have want you need.

I don't buy online.

[Jort]

I don't buy online.


Time to change then, today many people buy things online without a risk.
Nol is an ebay master :p:

[[XC] gomeler]

Agreed, you are shooting yourself in the foot if you don't utilize online retailers for a few things. I live in Atlanta and I wouldn't relish the concept of having to find local suppliers of all the parts I need.

[_HL4E_HalfLife_]

Time to change then, today many people buy things online without a risk.
Nol is an ebay master :p:

Easier said than done when ur on a fixed income.:slap: