Lab Worklog

[Ice Czar]

having been poked and prodded I'm starting a lab worklog
there will be little to no step by step stuff, just completed pieces
(odds and ends of fabrication)
and PICs of equipment and test setups

thought Id add a thank you here in the front, I owe most of what Ive been able to accomplish to Bill Adams (BillA), formerly owner of thermal-management-testing.com, then with Swiftech and now with CoolingWorks without whose advice I'd still be wasting money, of course all the guys at ProCooling both staff and forum members especially pHaestus and Joe Citarella's work both there and elsewhere and bigben2k. Bob Moyer with Autotest, Phil Hughes with Eesiflo, Kim Jensen with Dantec Dynamics, the folks at Keithley, and too many forum mates to list (you know who you are). And above all W1zzard.

Im going to start off with my latest score which hasn't even shipped yet
then wander around and take some PICs of other stuff tomorrow

my DegreeC (Cambridge) AccuSense ATM-24 w\ 24 x CAFS-220 sensors
an anemometer array (ATM = Airflow and Temperature Monitoring)
I should be able to build decent airflow maps and even determine some rudimentary turbulence levels,
not to mention thermal gradient maps for given cases, fan layouts and cooling products.

and the AccuTRAC software

dialing in altitude and barometric pressure

and about as far as I can go without anything to hook it to.

but should look like this soon.

will allow me to measure both the airflow rate and temperature of cases (24 spots per setup)
and is commonly employed for exactly that in the industry to verify Computation Fluid Dynamics Models
of electronic enclosures

a depressing number of these tools are currently beige
be forewarned, eventually I'll need to do something about that.

 

[Solaris17]

oooooooooooooo expensive toys.

 

[Ice Czar]

well I might as well post the list too I guess


(updated as of the last editing timestamp below)


Interface
National Instruments PCI-GPIB\LP NI-488.2 and cable
National Instruments ISA- GPIB-PCIIA IEEE-488.2
National Instruments PCI-DIO-96 w\ 2m 100 pin cable
(class status: need PIO breakout box)


Software
LABview 7.0
MATLAB & Simulink
DegreeC AccuTRAC
Keithley Runtime Environment
Autotest APG for Windows (3.0)
(class status: need to upgrade APG and like to add CFD modeling, need to get a grip on LABview and MATLAB so I can integrate all this together)


Measurement Hardware
Keithley 2700 Digital Multimeter & Data Acquisition System
Keithley 7701 32 channel Differential Multiplexer Module
Keithley 7708 40 channel Diff Mux Module w/ Automatic Cold Junction Compensation
Ametek 0-1000A Digital Recording Ammeter (AC)
Fluke 2176A 10 Channel Digital Thermometer w/ RTD
Digitec HT 5810 Thermometer
2 x Digidoc5 (hardware fan monitoring) and rheobus fan controls
(Class Status: more or less complete, need more Rheobus controls, adding another dedicated HP 34401A DMM for the UTS-325)


Sensors
10 J Type Thermocouples with shielded extension grade wire
10 T Type surface mount Thermocouples
2x 4 wire RTDs
Contact Temperature Probe (front panel of DMM) mystery tool being shipped
50x Thermistors (various makes and models)
10 calibrated J type panel connectors and plugs w\ strain relief in 2 handmade shielded quickswap modular panels
27 x IHA-25 F.W. Bell Current Sensors
EESIFLO Sonic Low Flow, Mass Flowmeter (pre-production prototype, temperature immune) being fabricated
Endress+ Hauser M-Point Mass Floweter
Setra 230 Wet/Wet Differential Pressure Transducer w/3-Valve Manifold
Keithley HA5050 50 Amp Shunt Sensor
American Aerospace Controls DC Current Sensor 913B-10-B
Micro Control Technology Calibration Module
(Class Status: need more RTDs and panel connectors, PCI\PCIe & AGP riser cards for the power consumption suite, need triple point cells (Hg \ H2O \ Ga) for in-lab calibration, need to have sensors professionally calibrated and their error curves plotted)


Electronic Enclosure Airflow and Temperature Test Suite
DegreeC Cambridge AccuSense ATM-24
24 x CAFS-220 sensors
AccuTRAC Software suite
(Class Status: Part one Complete on delivery, maybe get a dedicated RS232 card, part two build airflow resistance flowbench\wind tunnel to measure fans, as well as radiator and case resistance)


Power Supply Testing Hardware
Panasonic VP 5720A Oscilloscope (GPIB w\ digital storage)
Ithaco Lock On Amplifier
Autotest UTS-325 PSU ATE
w\ 5 x 150W programmable DC load modules (1A /uSec slew rate for transient step mode) a total of 900 Watts test capacity
2.1 KVA Programmable Variac, Differential wideband peak detector, 12 position, 4 wire switching matrix, line source measurement, load current measurement, universal timer counter and software suite.

(Class Status:The Autotest UTS-325 is a completely integrated automatic test system,
The test suite includes: AC Line Current, AC Inrush, DC Line Current, DC Inrush, DC Line Step, Efficiency, Vout Setpoint, Voltage, Frequency, Sequencing, Rise/Fall Time, Time Interval, Over/Under Shoot, Regulation Line, Regulation Line Frequency, Regulation Load, Cross Regulation, Regulation 2 Corner, Regulation 4 Corner, Regulation Dynamic, Dynamic Transient Response, Noise & Ripple, OverLoad Constant Current, OverLoad Constant Voltage, OverLoad Short Circuit, Short Circuit Recovery and Over Voltage Protection tests. Need to upgrade the EPROM and software, need dedicated DMM)


Sound Testing Hardware
Bruel & Kjaer 2203 Sound Level Meter with 1613 Octave Filter Set
Bruel & Kjaer 4131 condenser microphone
Manfrotto Tripod Adapter
(Class Status: need to build sound chamber)


Environmental Test Chamber Control & Measurement
PCon Multiloop PID control software
National Instruments PCI-DIO-96 w\ 2m 100 pin cable (listed above)
Omega CN77342-C2 Autotune PID Temperature Controller RS232
Rotronic HC321A Humidity & Temperature dewpoint Computer w\ Hygrometer
Kikusui PIA3200 GPIB Power Supply Control
Aquafine SP-1 UV Water Treatment Unit
100' 3\8" Anaconda Sealtite Isolated Ground Conduit
Aluminum Insulated Refrigerator door
3 Air conditioners
various blowers, pumps, insulation, sheet metal, metal stock, hardware, lumber and plumbing
(Class Status: need solid state relays, power supplies, UV replacement bulb,
incremental improvements will be made as time allows and need dictates, increasing cooling capacity ect.
Phase2 of the chamber calls for tight humidity and temperature control based on Saturation Thermodynamics)


Infrastructure
2 computers for DAQ and Control w\ 2U rackmount Cases
Dual Opteron Workstation for MATLAB\CFD Modeling (244's\K8W\FX3000\4GB)
3 pair Heavy Duty Rackrails (had others)
Bud Industries Rackmount 30U Enclosure
5 surge protectors
Olympus E20N
Lowel Tota halogen lights
Manfrotto 3251 heavy duty tripod and 3047 head
Welch Duo-Seal High Vacuum Pump Model 1400
CONDOR Linear DC Power Supply 5V@12A ±15V@3A 12V-15VDC
largely inaccurate mail scale (up to 75lbs)
ton of GPIB and RS232 cables
(Class Status: need centralized Online UPS approx 5KVA preferably Liebert, third computer for UTS-325,
fabricate vacuum chamber for thermal isolation, fabricate Calorimetry Tank)

plus odds and ends

 

[Solaris17]

i whanna come over your house and play

 

[Ice Czar]

you'll have to pay your own airfare

one of the "selling" points of all this is I'm not an EE or Test professional
my actual qualifications are that I was a spin doctor \ designer, that for some 22 years needed to quickly digest all the pertinent information of a given client then "translate" and order it into something the world could understand. Im very good at that, and if I can manage once to stuff something through my thick head I can explain it so anyone with more than 2 brain cells to rub together can get it.

Of course Ive been "studying" these fields now for several years (some more than others), and Ive yet to see the bottom
Right now I need to finance an Associate Student course I was accepted for at Bolton University (Design for Thermal Issues)

Ive always had an intuitive grasp of applied physics, especially thermodynamics, but that isn't the same as science, quantification not qualification is what is required, and for that you need exhaustive and repeatable measurement.

 

[Ice Czar]

handmade shielded thermocouple jack panels
these get cut into a case, and wired up to sensors via the back door (shown open)
and you plug the jack into the outside of the case, the plugs being directly wired to the Keithley 2700 DMM\DAQ specifically the 7708 card with cold junction compensation. (below)

have 2 more in the works
and still need to bandsaw a bit of aluminum for an increased standoff for these
(so they don't intrude very far into a case.)
will also make "feet" for the cover they are cut into so the case can be laid down without the panels contacting the surface.

A jack panel is on the lower left, they are a bear to get out once you have them in, but you can do it for some modularity
but basically I'll have 5 of the same kind in each panel anyway, easier to swap out the whole panel
shown is a J Type

Ill likely workup some sort of universal case cover these fit into which would let me access a case when its installed through a door in the cover and yet not disturb the sensor placement
however these are currently going into the back panel of my "test case" (Chenbro SpyderIII)
cant be whacking big holes in every case comes my way now

also make some sort of port for the ATM-24 above (which I can disconnect directly at the device), and likely shield those cables as well
you can see a bit O tin plated copper braid in the background there and some regular sheathing overbraid to keep it from contacting anything (its so pretty, but having the metal braid visible defeats the single point ground since it invariably touches something else grounded.)

here is a PIC of the 7708 with the thermocouples directly wired

just J types

 

[wtf8269]

Wow... I really wish I knew what all this stuff was lol. Whatever it is you're doing, it looks like you're doing a good job. You defenitely know what you're talking about.

 

[Ice Czar]

LOL I wouldn't exactly say that, I'm just stubborn not gifted


this is a PIC of the modular 120mm front fan mount (3 bays) with quick change perf metal panels, fan guards or filters and of course fans, not shown is the digidoc5 and rheobus that goes with it to "hardware" set and monitor the fanspeed (which is then passed on to a mobo header for software monitoring and logging as well hopefully)

here is an early PIC of it out of a case

the empty CDROM\RW cases are much shorter now
(I finally found a use for my Yamaha CDRW, it will be far more useful here than it ever was as a CDRW )
I'l be working on reducing the "planes" of the cases from interfering with the highly turbulent flow from the fans, being they act as artificial straighteners as well as add fixed mounts for CAFS-220 anemometer sensors, I'd like build a few more for this "test case" (lower front 80mm) as well as other test cases to come
including working on a setup to add into a case top and side in various positions.

this is of course to model "modded" cases not turnkey solution cases
but its an easy way to increase total case airflow when testing turnkey water cooling solutions
(Koolance you see in the back there)

 

[Ice Czar]

this is the rack with its sides off
(Bud Industries 2000 series 30U, that I rescued, was headed to a recycle bin)

from top to bottom
Panasonic VP 5720A Oscilloscope
which will actually be 2U higher as soon as my new rails arrive for the 2U Supermicro case seen leaning next to the rack.
It will contain a very old computer to talk to the UTS-235 which is the next item down
then there is the Keithley 2700 DMM DAQ (digital multimeter\data acquisition system) with a space next to it that will hopefully be filled by an HP-34401A DMM to interface with the UTS325
next another 2U Supermicro case with a Celeron 700MHz and the NI low profile 488.2 GBIP card which is attached to the 2700
thats the main thermal computer, the first being the "power" computer

I'm contemplating throwing my workstation in there (4U) which would then control both those computers do the heavy post calculating and graphing (MATLAB) as well as directly interfacing with the ATM-24 and the Environmental test chamber

need to synchronize the ATM runs to the DAQ, luckily AccuTRAC will timestamp to the computer's time and I'm not sure but I think I can do that with the Keithley DAQ as well, or at least get everything synchronized one way or the other.


now you can see what I meant by beige

that Autotest weighs about 200 lbs, it was quite fun getting it mounted
I cut two steel angle iron rails for it to sit on,
drilled and tapped them, then bolted it up and filed the heads partially down so the unit could just barely slide in
it gets a little extra space both top and bottom, and I'm considering a ground plane between it and the DMMs, though its already in a very heavy gauge steel case. I'm wondering if it would benefit from "enhanced" cooling, but worry about EMI.
It shipped very complete, with only a single screw missing (top cover) and a nut that holds the door lock altogether, at least as far as I can tell at this point. Came out of an AT&T lab or production line, and is virtually identical to the same unit available today except an EPROM firmware chip and the software version.

 

[Ice Czar]

this was the first multiple thermocouple log run
with a little cold water and lighter action on the thermocouples

thats in the Keithley runtime environment with the 7708

 

[Solaris17]

wow

 

[Thermopylae_480]

So... What are you planning to do with this lab? Besides do thermodynamic testing... You are creating a buisness aren't you?

 

[Ice Czar]

No not really, it's a lab to do reviews on an "unprofessional" or at best semi-professional basis.
I suffer from no delusions when it come to what real certified labs do
a good friend is an ISO certification consultant, to get traceable certs and calibration for my pile of antiques would exceed their list value by 3 to 4 times maybe, and their used value by over 10 times as a rough guess. I can get them "pretty" close to standards, but NIST and ISO doesn't work on "pretty close". It really depends on what device is under test, if I'm trying to get a general picture of what is happening to airflow and temperatures in a case and I'm running the whole shooting match through a slow temperature gradient from 5C to 40C (41F > 104F) to replicate the worse possible user environment, the end user doesn't care if the last couple of decimal places in a 6.5 digit DMM are correct, but a company looking for an ISO certification would. Real accuracy would of course be more important if your comparing “photo finishes” like you currently see in many waterblocks, of generally the same caliber performance.

I'm trying to develop in-lab calibration capabilities that are reasonably accurate, but for instance to cert the ATM-24 is $2100
Id like to have a highly accurate quartz thermometer that stays very stable for long periods and in lab "cells", the triple point of water, the melting poiunt of mercury and the melting point of gallium, so I can calibrate to those, and with the quartz plot error curves for thermocouples and RTDs. Then it might make sense to have the quartz calibrated and certified. But everything needs to be "calibrated" to a reasonable accuracy and real certifications are simply beyond my means as well as generally pointless since I have no customers or personal "professional" qualifications.

Each review I do will mention this level of scientfic "uncertainty", from both the lack of cert to the error factors inherent in each of the measurement processes, but ultimately I'll take the necessary steps to get "close enough", but not really that much further, unless I could get some corporate assistance.

For instance Eesiflow is going to both build me and calibrate a massflow sensor, its a sonic "low flow" temperature immune sensor, and my lab makes for a unique "application note" they can cite. Its a pre-production prototype and they like the idea of the exposure. So they offered me a very steep discount. Its still one of the more expensive outlays considering its overall utility value in the lab (being strictly for watercooling as opposed to say the DAQ), but considering the offer one I can't refuse. As soon as I get it, I'll probably round up as many pumps as possible just to start generating some publicity for them, even if the rest of the test suite isnt online, though I hope to incorporate it into in case testing as well as benchtop testing.

Any "money" being made would be from doing "reviews" or otherwise website oriented guides and developed databases.
Of course when I have clients come by to pickup their computers they are rather impressed

The lab is designed to be largely automated so that I can do a large volume of reviews without the historically intensive labor involved. It should eventually pay for its operation and overhead and slowly regain the investment in materials, if not the time Ive spent. Ive actually considered forming it like a non-profit business, sort of like a PBS, its really a media company with a very specific educational objective.

The primary one is to "review" whatever is under test, the secondary one is science \ electrical engineering \ thermodynamic \ metrology \ control engineering \ reliability prediction for dummies

and being a dummy Im eminently qualified

 

[Thermopylae_480]

Sounds great. Even with somewhat antiquitated hardware, your data measurements should be reliable enough for end users. I don't know how to calibrate that kind of equipment, but if you have to get, and maintain, water at its triple point to calibrate your thermometer you will be in for a difficult time. What kind of meteorological work do you plan to use your equipment for?

 

[Ice Czar]

a triple point cell for water is actually one of the easier DIY projects
properly insulated and chilled it should last a week without having to redo it.

Tackling The Triple Point Shawn Carlson Scientfic American Jan 99
build your own triple point cell, a sustainable calibration point of 0.01 C
keep in mind I actually have a vacuum pump too.

Calibrating With Cold Shawn Carlson Scientfic American Dec 2000
based on the freezing point of mercury calibration point is –34.8 degrees C
also includes a proceedure to compensate for a boiling point calibration of 100C
an important point at my elevation

I have assorted dewars on my watch list almost constantly, and they are rapidly climbing the must have list,
but right now they are still behind getting the PSU test suite up and running, I have sourced 99.9999% pure mercury and gallium as well as located a pyrex glassblower that can give me a hand. Strangely rubber stoppers and rubber stopper borers are higher on my list than dewars, I need them for making ports into the chamber.

and metrology wasnt a mispelling
but to answer the question the indoor kind or more accurately the environmental chamber kind
but that is the phase 2 environmental chamber Im still tweaking the phase 1 chamber
I do have a cool dew point computer though, its getting fitted into a Stainless Steel cabinet\case\former locking towel dispenser

the over riding concern right now is to integrate the ATM-24 and test chamber and create a "test" case review
so W1zzard can pimp case and turnkey cooling solutions, I need to be doing reviews to get some more capital

in the meantime Ive been making due with the old NIST shaved "Ice point" proceedure and just a boiling point
using distilled water and ice rather than deionized

 

[Thermopylae_480]

Sounds like you've thought of everything, that's pretty cool. I completely didn't catch metrology, don't think I've ever heard of it before. I just assumed it was a mis-type of meteorology .

 

[CjStaal]

So are you going to be doing reviews for TPU?

 

[Ice Czar]

That's the idea

had my Supermicro 2U rackrails show up today, will update the rack picture on page one tommorow.

 

[FLY3R]

Dam, your pretty Knarly dude!!! Well i can't wait to see your results, also taks some pics of when you start testing. Thumbs up.

 

[bigben2k]

Bump for an update

I've also recently acquired some equipment from Bill, but now I'm in a position where I'm moving, which is going to set me back at least 2 months, on everything.

Otherwise the acquisitions continue... 3 chillers, 1 heating water bath, 6 Fluke temp readers, 5 temp probes (TC, RTD), TC wiring, 3 scanners for the temp readers above, two translators for the same above, and the list goes on...

 

[CjStaal]

Wait, are you Ice_Czar?

 

[bigben2k]

Lol! No, I represent NordicHardware.com , but in the process of building a testbench, I founded the WBTA (Water Block Testing Alliance - wbta.us), of which Ice Czar is a member.