Most Stable and Quantifiable Heat Source
 

I am in the process of building a psuedo die simulator and I was wondering what would be the best heat source. I have seen alot of ppl reccomend the heater cartridges, but the ones that I have seen are all AC and i dont particularly like dealing with 120V AC.

I have also gotten reccomendations to use a TEC, but I have my doubts as to how accurate this method would be.

If anyone has any suggestion as to either side of this matter please say so.

You can use DC on an AC unit, just don't exceed the current rating.

Well it seems the tooth fair has solved my problems anyway.. lol.

but still its a good question.

u can put AC into the heater as well as DC? so it doesnt matter which you use? the only thing is that the heater is rated for 120v... 120v dc is just as nasty as AC.

actually DC is worse - AC current will knock you away from the source but DC will do the opposite, in fact just 50 mAmps (.05 Amps) can kill you

also i don't think that it's the same thing and is not safe to run DC on an AC circuit

size the cartridge for the current you want into the heater, 400W@120VAC work well for me
you will need a 100V 3A psu for 200W, ~75V 2A for 100W
use a shunt @ 30% of rating to measure current
cal everything

I will be using a multiple shunt system so that I can run it at full voltage the entire time with a range of 100 to 350watts. I see no reason to go over 350watts as it is not realistic. a prescott core with dual 6800's overclocked to the moon should get somewhere close to 350watts heat dump but nothing over it.

Even though all shunts will be at full voltage and their rated capacityies I will be using a clamp meter to measure the wattage and measuring the voltage on each just to make sure that everything is working properly and at the heat range i want.

so why on earth are you adding the GPU heat load to the CPU sink/source ?
you need to think on what you are setting up for

cal the shunt at its use temp
running a shunt at over 30% will cause progressively more self-heating -> changing the resistance

you should engage in conversation with Lee Garbutt
use his system as your guide if you do not wish to use professional equipment
a clamp-on meter ! wheeeeee

BTW, this is ALL dangerous, particularly for the uninformed
if you are not comfortable, and proficient, at working with electrical circuits -> DON'T DO SO

Excellent tips from Bill, as usual (thanks Bill!).

Yeah, you can use DC on an AC heater cartridge. As Bill stated, figure out the power that you intend to run it at, and check the max current rating of the heater, so that you're still running it safely. I didn't mention the voltage simply because I assumed that you were going to use DC and a relatively moderate power level, so you wouldn't be anywhere near 120 Vdc, which would be the max voltage at which you could safely run the heater.

Just pick a cartridge with a wattage rating a bit above your intended power run, with the other advice above.

I also didn't mention the other operating limit, temperature, because it's in the 400 dec C and above (far above, sometimes), and really doesn't apply here, unless you goof up (never run a heater without a cooling solution, doing so would damage it, and alter its properties, if it doesn't kill it, colateral damage aside).

McMasters.com page 454.

use a phenolic holder
if to run unattended:
install temp shutdowns
install loss of coolant shutdowns
think about all the different ways the system could fail, you will eventually experience most of them

coming in at 6AM and seeing a smoking die in a puddle of plastic is distressing

wait a minute
"I will be using a multiple shunt system so that I can run it at full voltage the entire time with a range of 100 to 350watts."

this is REALLY dumb, get an adjustable amp, I prefer linear but switchers will work

LOOK at Lee's setup
you seem hell bent on doing this in the most inaccurate manner possible

Bill the multiple shunts wasnt my idea. I will talk to the guy who is building it for me to see if i can just get it with one shunt that is really powerful.

The gpu and cpu is only for radiator testing as most users have their gpu's and cpu's in the waterloop. This wont be used for cpu or gpu waterblock testing. I just need a heatsource. I will prbably just atach a maze4 to it.


"cal the shunt at its use temp
running a shunt at over 30% will cause progressively more self-heating -> changing the resistance"

so you say not to run the shunt at over 30% of its rated voltage??? If that is the caes to get to 350watts I would need about a 1100 watt shunt.. EEK.

shunts are rated for current

stop building at once
you are wasting time, effort, and $s
you need to figure out what you are about re testing, and not really from me

there are several visible individuals doing testing who disclose their equipment
if you choose to do something different, without even understanding the BASIS for the equip selection, how do you think you will fare ?

I will leave you to Ben who is far ahead of you in terms of his understanding, and he has been at it some time
what you are seeking are reasons for not doing something different (dumb ?),
you need to test in a simple manner, then progress to more sophisticated setups
- if sophistication is not your bag, you'll find little solace for crappy test setups here

Its is the same thing since it makes no difference to a resistor what the waveform is, only the power delivered.

Both will hold you to a wire if you grab it (its not like your muscles can flex open and shut 60 times a second, so you'll stick anyway). If anything AC is maybe worse since it'll have a greater chance of arrhythmia and death by heart attack.

http://www.emedicine.com/emerg/topic162.htm

How are you planning to wire the shunt?

Well clearly Bill has had enough of my questions.

so Ben to get 350watts and below output from one shunt what exactly would i need to get.

The actual block used for this is already created by a more knowledgeable individual in this area in myself (he designs cooling systems and is an electrician) so I dont really have any questions on that matter. Just need help with the power source and measuring the output of it.

Bill do you have any links to such persons with these setups?

pleeeeze
Lee Garbutt - systemcooling
Joe Citarella - overclockers
'Roscal' - cooling-masters (have to read the articles)

an ounce of experience is worth a ton of "I think" by those w/o experience

"The actual block used for this is already created by a more knowledgeable individual in this area in myself (he designs cooling systems and is an electrician) so I dont really have any questions on that matter. Just need help with the power source and measuring the output of it."
if he is an electrician I would stick with him, up or down its on his head

I DO have a concern that the electrical part of this is beyond your present experience, keep to your buddy
I have 1KW psus that will take a direct short (which I have done !, blew up a switcher though), yours may not

Bill, I have several ppl that I can refer my questions to about electrical. stuff. Me EE skills are something that need to be improved on and I have no plans on going about building this thing on my own electrical knowledge.

EDIt: I wont be wiring the shunt, so I couldnt tell you. If it were me doing it i would use some solder, good low resistnace copper wire and some heatshrink.. but thats just me.

anyway thanks for the info on them. I read Lee's reviews alot, but was not aware he had a die simulator setup.

MR
sorry, but you are adrift

your 350W psu that will run flat out, dropping the 'excess' on a 'dropping resistor' will create ~500W from the psu + ~250W 'excess' that you will be dumping; thats 750W continous being dumped into your 'lab'

go get a 750W heater and run it in your lab,
but wait - there is more
you still have to dump the 100W in your coolant, how is that going to happen ?
to air ? -> got a constant temp air source ? (not if you're putting 850W into it, rest assured)

back up, get a feel for what your SYSTEM must do
that will give you an insight into HOW to do it

sigh.... testing is just about the most agrivating experience known to man.

So I ask you, is there a sensible way to get 350watts of heat into the loop other than connecting a prescott and two 6800ultras...

As to what I need the system to do. I need a copper block that will output a maximum of 350watts and have the output be measurable with DMM's and Ampmeters.

The heat dump into the room can be handled. I can duct the PSU so that it is dumping its heat out the window. As to the heater itself I will insulate it with closed cell foam to stop leakage through the sides of hte heater block.

I do not believe you have started testing yet, much to pass through first
350W ?
for rads use a water heater
for wbs use a resistor on a copper plate
if you want simple, keep it that way

hmm. well I was suggested this method I will present below.

take some 20pounds of water heat it to 150F and then run it through the radiator with a pump and measure the temperate at set times and also take count of how long it will take to get to 100F. Then this can be coverted to BTU over time. 20 pounds of water going from 150F to 100F is 100BTU's. So it would be a relatively simple calculation to get the time.

also first test how many degrees the water will drop per time from 150F in the bucket used along with the pump hooked up and the tubing used. This would be a constant that you pull out. for instance if it dropped 10F per hour then 200BTU's per hour.. just subtract htat from the BTU's measured for each radiator.

Bill pls let me know if you see any holes in this plan.

one question. do you know where to acquire a water heater that has a measured heat input into the water? Preferably one that doesnt cost an arm, leg and a baggie of gems..

ok
test the concept
with the thermometer of the accuracy that you intend on using, see how uniform is the temp in that 20# of water
- cannot be uniform w/o mixing (which adds heat), isn't this fun

MR
go to an engn prof and talk to him, someone is blowing smoke up your ass
I've said my bit

fun doenst beging to describe it.. so you stir the water it adds kinetic engergy.. gee how fun is that..

I was thinking the same thing about the uniformity of the water... but it will be pumped hte entire time which will move it arounda bit i think. sigh....

can you elaborate on the blowing smoke up my ass part.. I really dont want to start doing all this testing and then realize that I have a totally fubared test method. getting all the possible error points out of the setup before i start testing is the optimal solution imo.

Check eBay for 800 amp shunts; they're there, here's one:
http://cgi.ebay.com/ws/eBayISAPI.dll...tem=5766468739

Here's 1000 amps, with meter:
http://cgi.ebay.com/ws/eBayISAPI.dll...tem=7507355711

350 W is a lot of heat power. Are you sure that's what you want?

From what you're describing, you're looking to test a cooling solution (i.e. radiator), not a water block; is that right?

If that's the case, I'd recomend a custom made block; a solid copper piece, with integrated heater, inlet and outlet.
edit: ...or more simply a water heater, as Bill suggested...

thats a good idea ben.. the block that is being made would work well for that actually. its a round copper block that is 1.5inch in diameter. not sure on the length. but i could easly cross drill it and tap it for 1/2inch fittings.. that would work well.

350Watts its alot but that will be the max testing. I will do 100 to 350.. it will show how some rads max out and cant take the heat and where the limit is drawn for good cooling ability at each wattage rating. I would like to do 100-350 in 50 watt incriments.

EDIT: how the heck would i use one of those shunts....? it looks like it would just heat up the air no?

So this is for testing a cooling solution then?

Is this for yourself, or for site reviews, or for R&D/product development?

Did you understand the purpose of using DC versus AC?

There's more questions than answers in this thread.