I’m sitting on a few darn-near-ancient twisted-pair network cable certifiers that are 15 years or so outside of factory calibration. (Fluke/Wavetek). I know a number of third parties will calibrate them – but really, I’m more interested in learning how to do that myself, of course 
Does anyone know anything at all about the calibration of this sort of equipment, or about calibration of Fluke or Wavetek test equipment in general?
(and if anyone wants me to bring a couple of these down for a gander at them cracked open, let me know and we’ll arrange something.)
Really you need traceable calibrated equipment to calibrate other equipment. But the instructions should be in the maintenance manual or something like that.
You’d think, right? But I can’t find maintenance or service manuals anywhere for these things. (The long-form user guides are no help.)
My goals aren’t quite as lofty as “official” calibration, for which certainly the test equipment used to do it would itself have to be calibrated and certified. One of these units is far enough out of tolerance so that its self-calibration fails 80-90% of the time (but does pass sometimes). To know just enough to be able to “tweak” it back to where it self-calibrates regularly would be handy. But again - finding documentation on that seems just about impossible. I’m sure those service manuals are proprietary, but these are old devices - and again, there are certainly a number of third-party services willing to calibrate these (for about $500 a pop). I figure the information must be around somewhere!
Have you tried just calling the service number? Sometimes you can talk them into emailing stuff like that for free…
First how ACCURATE do you want this. Accuracy generally translates into $$$$
The first thing you will need to get is a certified standard that is traceable to NIST -National Institute of Standards and Technology. It will state what it it’s readout is (ohms, inches, weight, whatever) and what it’s tolerance band is ± .01 ohm, ± .0001", whatever the unit of measurement is.
Then you check and adjust against the standard. Then maintain the standard’s certification, or not if you think it is stable and good for several years. Then it is just reference standard.
I wouldn’t know how to do anything electrical, but say Machine Shop wants to make a simple in house standard for calpipers. We could get a piece of say 2" wide 1" thick by 18" long piece of tool steel, it’s hard/stiff and wear resistant.
Then we do our absolute best to align it in the HAAS or send out. Have notches machined through with as smooth and pararllel a surface as possible at specified intervals. Then we send it out to a lab and have then measure the distances from the end to an edge and provide us the measurement to within .0001".
To check our calipers, it won’t matter what the actual distance of our machined bar is, we may have been trying to do 1.000", 3.000", 6.000", 10.000" 12.000" and 16.000" The measurements will come slightly over or under, but they will be a precise measurement (the reason to go to calibration lab is they have the ability to measure to at least .000025" (25 millionths of an inch).
When we check/calibrate against this, lets say our 4.000’ is actually 4.0018", if my measuring tool is good to ±.001" as long as we get a measurement that is within ±.001" of 4.0018 (after adjustments if needed) then we know the tool is good. If we are measuring to .0001" but can only get repeatability to within .0002" then we would note that on the tool even if it has a readout in .0001".
Anything we use at the Space would in industry be marked REFERENCE ONLY meaning it wouldn’t be allowed for product acceptance purposes.
The most accurate measuring devices I’ve used were a leakage meter that measured He loss in a vacuum chamber - it could detect several CC’s per year. Flatness: optical flats that were good to about 30 nanometers or 10 millionths of an inch, super micrometer: 00001". Ohms, all I know is as you approached the meter you affected the “electrical field”, humidity had to be is certain range, you waited till it stabilized, zeroed it, then did the measurement (was for detonators for explosive bolts used for rocket separations).
Or you buy standards and pay a hefty premium.
Yeah, tried the service number… No luck. Part of the issue was that for a lot of these testers, the companies that produced them went through numerous mergers and acquisitions in the '90s and '00s, with some resources from the older company lost each time - that added to the fact that really, the acquiring companies had no vested interest in maintaining those devices vs. upgrading you to their coolest latest thing. So I get very polite, professional CS reps wanting to help but having literally nothing to work from regarding these old devices.
(A HUGE shout out there goes to Ideal - not that I probably need to brag on them here or anything - but I had a question on a device owned by Test-Um, bought by JDSU, then bought by Ideal, and they literally bounced the question off of three different national CS departments trying to find an answer. We gave up after a couple of days, but then a week later I got a response from Ideal GmbH - Germany! - about an unpublished software update that solved my problem, that they dug up from some dusty hard drive somewhere. VERY impressed.)
Far as I can figure it, my next best option is to find someone that works for one of the third-party certifiers and offer them enough good booze and/or steak for them to talk…
A good question. I don’t need it accurate as much as I need it precise. (Case in point: the issue I mentioned with one device’s self-calibration)
But it is some sort of electrical/electronic calibration, it would seem; reference to standards for physical properties is fortunately not something I likely have to contend with.
It can give a precise measurement that isn’t accurate. E.g. Every time you measure an item it says 34.0145 units, but it really is 35.0145 units. Metrology uses “precision” and “accuracy” in very defined ways.
Just google “Calibration Labs” or “Calibration standards” for whatever the unit is your measuring and the range.
The other el cheapo way … find someone with a calibrated meter, have them check with their meter, then you check. Do several different things (I’d suggest a minimum of 5 over a range) so you can see if the difference is linear. You can adjust your meter to match or just note the offset.
Have you tried Davis Inotek? They have/had a local office the last time I used them.
Right. I use “accurate” and “precise” above in that same, formal sense. Precision is prioritized in this particular case because the meter I mentioned fails self-calibration - meaning it can’t negotiate a consistent set of results with its remote unit via a known interface. A precision, not an accuracy, problem. Factory “calibration” addresses both precision and accuracy. But it is the precision component I am presently most interested in.
The thing about Cat5/Cat6/Cat6a twisted-pair cable certifiers is, their calibration is hardly as simple as comparing to a known-good and then tweaking the results to match. “Tweaking” pots to get, for example, a proper PS-ELFEXT calculation, would be nearly impossible. And the meters, at least without some service-menu trickery, don’t provide such raw data. (Which is why a service manual would be useful - at least then I’d know whether there even is a service menu, and how to get to it!)
For that matter, having cracked a few of these open, there aren’t awful many pots… No surprise, really, considering their complexity. Most have a secondary, soldered-in LiPo battery whose probably exclusive use is maintaining the unit calibration data, in firmware. So I’m guessing even the calibration process itself requires that same “service-menu trickery.”
@TBJK, thank you! I wasn’t aware they were local. Looks like they have an office in Lewisville. Once I officially give up on this thing and put it back together (which may be sooner rather than later), I’ll give them a shot.
Turns out Davis just has a local sales office in town. They ship all their “real” work out to Illinois. Dang.
If anybody knows of a local resource for higher-order test instrument calibration/repair, the info would be appreciated.
I miss the Tek trailer
Trescal in Irving is a possibility. They do have folks local who calibrate equipment.