RS925D

[cobychuck]

I started working an RS925D this week and I am having trouble with the setup.  Since the procedure was changed the .001 thru the 1 ohm dials are checked on the 3458A.  The readings I was getting were out of tolerance so I ran it by my section supervisor.  He thinks the problem could lie in the settings for the 3458A, but so far we have had no success in getting an in tolerance reading.  I read through appendix A for the adjustment procedure, but it does not appear to cover the lower ranges of the RS925D.  If there is anyone that knows if there is a setup or specific setting that will correct my problem please let me know.

[WestCoastCal]

3458A, FUNC OHMF, INBUF ON, NPLC 100, NRDGS 1, TRIG SGL, OCOMP ON
Datron 1281 or Fluke 8508A TRUE_OHMS, AUTO (range), RESL7
We set the UUT to 10 Ohm, measure that, increment the 0.01 Ohm step, measure that subtract the 10 Ohm measurement.  Recheck the 10 Ohm reference.  etc. etc. All this is done over the bus with instructions to change the UUT setting.  We have found it painstakingly slow to do it this way as each reading takes a while and we discard the first 2, BUT we feel pretty confident that we are getting good measurment results.

[Smokey]

Our RS925D Did the same thing. We fixed our problem simply by opening it up, cleaning and excersizing all the switches really good and then it passed with no problem.

[Grizzly]

We set the UUT to 10 Ohm, measure that, increment the 0. 01 Ohm step, measure that subtract the 10 Ohm measurement.   Recheck the 10 Ohm reference.   etc.  etc.

I was hoping you could go a little more in depth on your set up West.   When you set up to 10 Ohm, do you set to 9.  99XX or 10.  0100.   Either way it seems almost like doing it the old with with the 242E or the 6625 where you'd calibrate your lower decades using a larger decade to facilitate the measurement.   In any case I wonder if that satisfies the intent of the new procedure, which appears to call for isolating the .  01 Ohm/step decade entirely so as to take into account internal switching/wiring and so forth.   My question is this; if you use the 10 Ohm reference then aren't you using the .  01 Ohm/step decade in the reference and thusly attributing any deviation of said decade to your reference? 10 Ohm reference aside, if you connect your RS925D set to .  0100, to your 3458A (I do like your DMM set up) does it meet specs throughout the decade?

[pmelgeek]

We encountered the same issue with both our RS925D's.  Our Quality Dept deemed using the extra 10 ohms as too big of a deviation, so we ended up taking the cheap way out and limited the range.  The lab I left about a year ago had one of those go down on an AFMETCAL audit for this as well.  If this problem every gets completely pinned down, it would make a good newsletter article for certain. 

[CalibratorJ]

I'm not an expert in DC/Low Freq by no means, but I would definitely want to get my 3458A back to the manufacturer's 24 hour specs. I looked over the T.O. and it didn't mention using the 24 hour specs, but, it couldn't hurt to do the characterization of the 3458, if you guys can. (Not sure the AF can as you are basically recal'ing it)

I know I am battling a similar problem in our lab with 5520s atm- but in that case it was just the characterization of the resistors that led to the issues.....

Opening it up and cleaning all of the contacts and exercising the dials definitely will not hurt either.

Just my 2 cents- not sure what you guys can and can't do.

[Grizzly]

I cleaned the internal switching and wiring with a contact cleaner called DeoxIT (Craig Labs, Inc. ) and it did meet specs, but as a precaution I shortened the interval to six months to make sure it would hold.  It didn't hold; six months later, my RS925D is OOT.

[Hawaii596]

I like Deoxit... great stuff.

I had an RS925D that was a used equipment dealer special (I'll leave the name of the company off.  I didn' t buy it).

I had low resistance problems.  This was at a former lab.  I always thought it was ingenious how by doing away with a zero setting, you could actually dial in the fully compensated resistance.

I believe my problem was an OOT 90 Ohms on the x10 Ohm/step range.  It was just barely out on the high side.  I had gotten a quote from the new OEM (IET Labs, I think) for I believe $3500 or so to fix it.  I ended up doing strange things like doubling up a couple of leads to lower residual, and other things like that. 

I agree it was odd that all the higher ranges had trimpots, but nothing on those lower ranges.  I spent about a day on mine getting it back in spec.  They're nice decade boxes, but a pain when they go OOT.

[WestCoastCal]

I'm pretty slammed right now, so will try to give a short answer that's correct.  Basically, if I remember correctly when I wrote the procedure for decade resistors, the majority were specified as ±n % change from zero setting, with their "zero" resistance a typical specification.
Additionally, we had 3458A's and Datron/Wavetek 1281's.  (We have been transitioning over to 8508A's and they have a 2 Ohm range so I need to look into that)

The lowest resistance range of the 3458A is 10 Ohm.  The fullscale range error of the DMM for instruments that have a 0.01 and/or 0.1 Ohm/Step decades can be a significant portion of the allowable calibration tolerance.  The 3458A full scale error for it's 10 Ohm range is ±50 µOhm.

Because of this full scale range error, the performance test is performed as follows:

The resistance is set to 10 Ohm on the 10 Ohm/Step decade.  The Multimeter will measure this value, this value will then be mathematically subtracted from each measurement below 2 Ohm.

[cobychuck]

We discovered that the .001 dial had slipped so that there was actually a small amount of resistance applied when it was on 00.  The contact was also bent out a little bit so that the only thing keeping it in contact with the rail it rested on was pressure from the case being screwed on.