Re: Giving my test equipment away?
 

well I can tell you my proposed specs at the moment for your review:

Iwaki 30 rzt
He 120.3
Minco heater http://www.minco.com/order/country.a...%3fPN%3dHR5293
Swissflow flow meter sf800
Digitec 5810 Thermometer (ordered)
Dalec RTD thermometer (soon to be ordered), for calibration and a little more accuracy then the Digitec
fluke 8050
4 PRTs 100 ohm
1 thin film RTD ohm for Omega
2 YSI probes
barton 247c Dp meter (0-600" wc or about 0-20 psi)

Accuracy they state:
DP .75%
flow 1%
Temperature about .3 C for the digitec and about .1 C for the dalec

Re: Giving my test equipment away?
 

Yeah, you're right, Bill. I'm being a little reactionary - I got the impression pH was considering giving up the whole LabVIEW setup and throwing in the towel. I don't want to see that. At least without some assurance it will get used correctly.

Re: Giving my test equipment away?
 

haha Bob

When Joe was talking to me about the labview system I started kinda explaining the configuration and all the ins and outs and then just ended up with "nobody else is likely to make it go".

I had 2 things on my "to do" list today though: Aquarium needed upgrades and cleaning and I was going to try and get my hobby room in a condition that I could play around in there.

If I put Radmin on my labview PC would I be able to give you a login and a phone call sometime? I'm sort of confused about a couple of things right now that SHOULD be easy I think.

Re: Giving my test equipment away?
 

After the high-and-mighty lecture I threw down? Of course I'll help out. It's the least I can do.

I was just worried that we'd get someone who was so worried about the spec's, that when you asked why they believed it was 3.0000 volts, they'd reply, "cause the meter said so." No understanding of how the measurement actually was taken or what it meant in terms of the physics. Bill will chastise me for too many "negative waves".

Cheers

Re: Giving my test equipment away?
 

bobo
British Sub Fleet ?
(yea, I'm slow)
the ocean as a sink ? good stuff - heat pipes to hull plates, no biggie (lol)
- I can do MUCH better with condensers though

I am starting another profession and encountered this rather aggresive doc, thought you might like
http://www.scs.uiuc.edu/chem_eng/cou...lationRot3.doc

Josh
get a grip, those paper #s don't mean jack in a system; system cals are what define the capability

Re: Giving my test equipment away?
 

Oh I know that simulating the cpu is nearly an impossible task and even if my test equipment is accurate it does not mean it simulate a cpu properly or effectively. That is the task that anyone as a tester struggles with. I am putting out with those numbers are not how effective my proposed bench will be but that what it is simulating will accurately measured. The best that we can hope for in any bench is a good comparison that offers enough accuracy to differentiate between blocks. Just look at systemcooling's review of the new apogee block. His test predicted that the storm would far out rank the apogee but it did not beat it out in system testing. This shows me two things, one that the die sim is a good comparison but cpu testing is also needed and that our tests do not account for everything. For now I am going to shut up as without my bench up and running, no one is right but arguing a unprovable point.

Re: Giving my test equipment away?
 

I’m not precluded yet but if I get the job I maybe or I may have to have my posts check and things which would make talking about things not worth the effort especially when I have such expensive toys to pay with. I’ve heard the training budget there would get me what I need in terms of experience and my chartership :) .

Ooo chem. Eng. Now that it is geeky and fairly difficult. Chem eng degrees also have the highest leaving range of any prevision. At imperial the leaving wage of chem. Eng easily beats anything at oxford or Cambridge. Go to the oil industry can easily get $100k+ leaving wage. The doc seems par for the course on topics like that where copying can be blatant and defeat the point.

Can’t do heatpipes to the walls as the walls shrink and thermal effects are not good in something of that size plus the orientation issues with heatpipes could be a problem. The boxes are not hard to cool (custom dsps) but they have to operate under conditions which mean custom design.
http://www.detica.com/indexed/CaseSt...rroyalnavy.htm
Resistant to acid mist and power shower style wetness I’m guessing kinda kills anything normal from a cooling perspective. The box is roughly rack mount size.

Re: Giving my test equipment away?
 

good luck re the job, sounds interesting and challenging - the right 2

yea, I am clearly beyond my depth attempting PFDs but the PIDs are not a terrible reach
spoke with a friend yesterday who was a ChemE major, only to hear that it was so difficult that he backed off to a regular Chem degree; he has the smarts but not the perseverance I suppose
no dummies survive with that major

Re: Giving my test equipment away?
 

Chem Eng is hard but you get good money from it in the end. The investment banks luv chem eng graduates because they stick at it. I also presume because not many people like it and it is hard so those that go through it. Engineering degrees in the UK are completely different to the USA. They bear almost no relation in the way we do things.

Ah Detica is not the only one option there are many others so nothing is determined yet, been very bad about applying. All are challenging and interesting:

Deutsch bank, energy derivatives trading – no time for life but im a big playa can afford some expensive stuff. May gold plate my PC. May die from stress (believe me this is not a joke I know ppl who have seen it happen) or I will get rich.
50 cent would be proud.
Qinetiq – Ultra high spec military engineering the equivalent of working for Lockheed martin skunkworks / department of defence in the USA. Do most of British military research. That will probably also be confidential don’t know if I can post if working in the same area.
Engineering consultancy – No practical experience and tell ppl what they already know but money and travel the world for free. You can earn serious cash working in Iraq for example if your willing to take the risk.
RCA Integrated Design Course - £18k a year so nasty but the most prestigious Design course in the world. Hard as nails to get into and if I graduate I will never have a boring job again. Former graduates include Jonathan Ives at Apple. Mr Dyson of vacuum cleaner fame. http://www.designcreate.info/ , some nice ideas in there.
Watercooling for non-scientists testing rig as a final year project? :P
PhD/ Masters at Stanford, MIT. Might want somewhere warmer and I need to cover up my bad grades from not revising/ doing tutorial sheets though. I was thinking Caltech but my housemate says they’re shit apparently, maybe Berkley I don’t know about these American schools and I would need a good reference and some cash I think.

While I remember there are novel testing methods that could be done as well. Know one here has considered having how water connected to a cold cpu. 2nd law irreversibility’s may play a part but it is certainly a way of getting rid of the cpu problems. It is a method at least but I’m not sure how well it would work.

Been having a look at direct die cooling with FC – 77 . I mention this not because it’s a good way of cooling the cpu it’s not (see below cos I know some of you will be interested) but the experiments there had a long term variability of 0.4%. Of that 0.3% was from source (cpu die) and they are using similar die sims as the DIY guys around here.

Cooling wise this doesn’t work cos while FC 77 is good it is not as good as water. For one it is 1.7 times the density of water and has a 1/10 of the conductivity (jet impingement performance is roughly equal to thermal conductivity). For 5lpm and a single jet nozzle you get h = 4000 W/m^2 which gives a W/c of 0.8 (dc althon64) which is nowhere good enough. I think I could get it better maybe up to about h=20000 W/c using multi jets maybe try crossflow as well but still not good enough. There maybe some benefit as the resistance of the thermal paste is removed so cooling of hot spots could be more effective. Will have a look at ways of making it better in the literature as well.

FC 77 is flourinert – water with fluorine instead of hydrogen. Low freezing point of about -70c. non toxic but I doubt anything would grow. Clear and just like water as well so you can run an lcd in it , but cost about $250 a litre. Definitly pimp look Ma ive put a hair dryer and a hand in a bucket of water.

Re: Giving my test equipment away?
 

Hi Josh,

You mentioned earlier that you didn't plan to use a "chiller" and you have a He 120.3 listed in your equipment list - how do you plan to control the water temp going into the WB? For me, scoring a Haake chiller was the turning point of setting up a WB test system. Temp, flow, dP are all important but you need some way to control the WB inlet temp as well (IMHO a raddy isn't going to give you the accuracy or control you need /are looking for).
Good luck with your project and let me know if I can be of any help. I look forward to having some one else in the pool... :)

Re: Giving my test equipment away?
 

I have worked as a Engineering Tech for 22 yrs in a R & D Lab, for a big oil company (Arco Oil & Gas, acquired by BP). After working with PHD's of all kind, chem, petro, mech, materials, geologist, most were terrible in the lab. Geologist were in their own world. They had no skills when it came to setting up equipment and running test. Though the chemest were pretty good. I think I only had one that could duplicate his results, though wrong due to the way he set it up. They were good when it came to processing, interpreting, data even thought I did the most of it. A good lab tech is worth his weight in gold. Most have a broad knowledge of intstrumentation. The more automated, less the error. I watch as our lab went from gages to full automation. I was one of the few that did my own programming, as well as build and operate equipment. A good tech is a perfectionist, and accepts nothing less. Recognizes weekness and make the required changes. The equipment has a MAJOR affect in the integrity and accuracy of the data. The biggest mistake is resolution and responce time. In todays world everything is computerized. The world of Acquisition and Control is no different. I use to use National Instrument LabView. Most of the other systems copied them. I ran test that ran for months down to day or less. My biggest task was scalling back, Oil field projects can be rather large. But the chance for error increases as you down size. Resolution and Repetability is every thing, but if the instrumentaion does not respond quick enough, the data is flawed. Its nice to work for a company with Deep Pockets $$$$$$.

Doing the math is the easy part. Getting accurate data is hard.

I will get off my box.

Give the guy a chance and the tools needed to do the work. Then make a judgement.

Re: Giving my test equipment away?
 

I know that not having a chiller is a major problem but right now I cant afford a $3000 chiller. Secondly I will probably make a chiller using peltiers when I get around to it. As it does not take that much power (peltier wise) to cool to room temp (I'm thinking 120 to 150 watts of cooling power) but that hard part is generating a control circuit that will be able to maintain a stable temp. This I will look into a much later time but I am hoping to build a "chiller" that can hold outlet a +-.5c. Blue I recognize that response time is very important but in this case it is a little less so. This is simply because the temps need time to stabilize (which are much greater the the response time) so this gives time for the machines to "catch up" but yes it is something I am worried about. Also I know that repeatability and drift are insanely important and they are the biggest problem for this bench. Well I should say drift is, if I am claiming that my bench is a good comparison over time then my instruments can not drift and need to be recalibrated often. To say the least I have been thinking a lot about all the afore mentioned topics and trying to find the cost effective middle ground.


Just a small update on equipment bought:
digitec 5810
Swissflow flow meter
HP multimeter (I cant remember the model but I will probably replace it)
Regulated 30v 10a power supply
Delta 0-100" Dp meter (I know the scale is to small but for now it will do)
Cole parmer 8502 thermometer
Cole parmer thermistors

Also blue I so jealous I would love to have a large budget. Also Robo I am jealous of your bench it is what I hope mine will turn out like.

Re: Giving my test equipment away?
 

Responce would be needed if you are monitoring temp vs time. SS covered probles have a big time lag responce. Which is used for tank and inline probes. If you have an option use exposed sensors. If not get grounded for faster responce, un-grounded for less noise, and slower responce. RTD are highly accurate used them mostly for references.

Flowmeters. I use to have to measure rates down to 3cc/hour. Can you say sloooooow, liguid and gas. Where the gas was toxic. I alwas liked the lamilar flow, felt it was more accurate than the others. But its alway nice having tranduces in the 0-30" and 0-6"WC range, and a Big Budget. Made of HC276 wetted materials with a overburden pressure at 1500psi..... I think they cost $1200USD each in 1990, the last time I purchaed them.
I used Eurotherm Temp Controlers with scr's. With these I could do my own tuning for responce curves. I was able to have temp contol deviation of less 0.5F.

Since I worked with Hazardous materials all of the time. They (Arco) built a building specially to handle it. Restricted Access, tighter than our data center. I had single pass air (scrubbed), explosion proof rooms/bunkers. Designed to with stand a cubic foot of gas @ 30,000psi, catastropic failure = 5 sticks of TNT with no damage. I was paid to take thing to failure some times. Fun when you get to do that. A lot of the work I did, I watch come to market 5+ years later.

Re: Giving my test equipment away?
 

Yeah thankfully time is one variable that we get to remove in this case well atleast temp wise.

Re: Giving my test equipment away?
 

you are not listening Josh, but thats ok 'cause you will learn the same when you start testing
if you cannot CONTROL your water temp you will chase your tail for months
what is this $3000 song ?, there is a Haake A82 on eBay for $200 twice now; it will be back

in any case get setup and start, just hold the 'data' 'till its repeatable; no small task
how do you intend to cal your system temps ?
http://forums.procooling.com/vbb/showthread.php?t=12386

Re: Giving my test equipment away?
 

Bill's right; eBay has chillers, be patient. Plenty of Thermoteks right now, but averaging $500 (they used to go for as little as $150).

I've explored the alternatives around a chiller, and they're not pretty;
if you're not able to sustain a specific temp (as a chiller would), then the only option left is to try to maintain a steady temp slightly above room temp (i.e. 25 deg C). So you integrate a thermal probe with a PID temp controller (preferably a smart controller), to vary the radiator fan speed... long, complicated, and will only work with the right controller.

Thanks for the comments blue68f100; you're spot on about resolution and response time.

Re: Giving my test equipment away?
 

Ok I will buy a chiller or build one but the bench will have one.

Re: Giving my test equipment away?
 

I have had 3 Haake A82s, ±0.02°C is the cyclic variation (more or less at the DUT depending)
good luck on a DIY effort

EDIT
had the range

Re: Giving my test equipment away?
 

well now looking at the prices I do not see it being worth the effort. Although it wouldn't take much besides a very good control ciruit.

Re: Giving my test equipment away?
 

I pass

Re: Giving my test equipment away?
 

yeah I will too but I wish I could build one.

Re: Giving my test equipment away?
 

even if you could build one it is likely that the variance of that one component would make having all the expensive other equipment abit expensive.

Easiest DIY approach off the top of my head is to use tap water or a very large tank. Can get constant temps that way but can't varry input temps. Could try water heat exchanger using house water supply (comes through your cold tap and goes back into your cold water system in your house) should be able to rig something up that it can varry temps. Would be alot better at cooling the flow than a rad.

Re: Giving my test equipment away?
 

Good Luck Josh

having personally invested some $7000 to date and countless hours
Id highly recommend you listen to your uncle BillA



can I nominate this for the understatement of the month award?

Re: Giving my test equipment away?
 

hey IC, good to hear from you
an update ?

Re: Giving my test equipment away?
 

aquired Labview
the Agilent 6050A needs repair (ouch)
been battling an idiot that never sent me my NI SCB100 breakout box
and Im about to do my first flawed review :p
ready or not

back to the front page today however
have to be a witty monkey and spin the news