Thanks for the info. I will try to repair the switch next week with epoxy and hoping for the best, while searching on ebay for a similar one.
Also: Got the KVD running again after carefully rotating the shaft with the brushes in the S8-function switch to the OPR-function with long nose pliers and soldering it back in for a sanity check. It performs beautifully, apart from the 7. decade which seems to be partly corroded on the switch and therefore gives false readings in some positions.
I could successfully repair the S8-switch by cleaning and then glueing (epoxy, "Uhu Endfest 300") the metal-thread-shaft from the switch-front to the remains of the plastic-shaft, which rotate the bushings. The switch now performs really nice and hopefully lasts as long as i use the KVD (fingers crossed).
Since the plastic shaft was badly damaged and very short, i didnt see any possibilty to repair the switch by means of a shaft coupler, therefore i used the glue as a last resort. If the switch breaks again at some point, it might be not repairable anymore since one cant open it due to the glueing (if it doesnt break at the glued shaft-transition).
Time to come back to this poor abandoned project.
What is better than fixing Fluke 720A??
Fixing two of them! Second one got delivered today.
Somebody vandalized poor KVD, and gutted kelvin-varley potentiometer for bridge balance (lucky for me, I have one spare!) and Fluke factory trim low ohm (4-6 ohm) wirewounds for decade A calibration.
Perhaps after some time, previous owner got hit by regrets and remorse, so they included a bag with removed wirewounds in the package. All but missing one
Lucky for me again, i have 10+pcs of
these too.
I guess idea is to play some tetris to find the proper ones to go proper spots...
TBD...
And after that you can use both KVDs to rebuild the setup from the Jim Williams AN86 with 3x 3458A and "spend
many delightful hours “surfing the Kelvin”".
Put a 2002 in 3458A emulation mode. No one will ever know.
Put a 2002 in 3458A emulation mode. No one will ever know.
I doubt that
The 2002 isn't nearly as linear as the 3458A. It is more in the league of a 34401A (talking about linearity only)
The 2002 isn't nearly as linear as the 3458A. It is more in the league of a 34401A (talking about linearity only)
Maybe 2002 isn't as linear as HP, but saying K2002 is in league of 34401a is understatement at least
Here's what I got from today's test on one of the meters.
If we stay over 15% of the FS, then
total system (MFC+DMM) linearity is <0.2ppm.
Hi Tin,
to what value are your ppm's referenced to? 20V?
I attached the meaurement of my HP 34401A after ADC adjustment. Deviation is referenced to 10V. The measurement attached is measured against my Fluke 5440B. A measurement against a 3458A looks quite similar.
This kind of linearity is what I measured over and over again on many 34401A's. Dr. Frank showed similar results here in the forum.
It seems to me that the Multislope III is much better in linearity than the (faster) Multislope IV (beginning with 34410A to the actual units).
Edit: Here I measured the Keithley 2002 against two 3458As. For me it looks like the K2002 is in the league of a 34401A (we are talking about linearity only!)
https://www.eevblog.com/forum/metrology/dmm-linearity-comparison/msg1354870/#msg1354870
Reference is to calibrator output value, not range. I can try your script once I have meters free (this test was part of my calibration check report procedure, its not targeted to pure INL tests anyway).
I bought a 720a from a legacy in unknown condition and the shafts are broken, too. Also ALL plastic standoffs are broken, as lowimpedance wrote.
...
One interesting mechanical failure which I suspect most of these oldies suffer from is the acrylic standoffs craze then crack and fall apart. see pic.
...
What material are these plastic standoffs? Can I use PMMA/Acryl tubes for this?
@TiN:
Which material and technology did you use for the black shaft? (It seems not the "normal" FDM-3D printing. It looks like ABS with injection molding technique...)
I can see, that your standoffs of the faulty wired switch are also made of a different white material. Do you know, what it is?
Taking of shafts.
I'm replacing acrylic bad ones with plastic from my 720A parts box. Took few quick measurements to make 3D CAD in case some poor voltnut want to 3D-print some, as I don't have access to this advanced technology.
....
Looks like some thermoset material. FDM printed thermoplastic material would probably be to soft and could break pretty easy. Maybe you can use extruded PMMA, if you plan to machine them. If you plan to print them, you can use SLS parts that have been infiltrated or use DLP printed resin parts.
-branadic-
Replacement black shaft is not 3D printed , it is spare part from other 720A.
Now I used 4/10mm PMMA (acrylic plasic) tube for the standoffs (see the picture). The original screw sockets fit very well
For the broken shafts: a few drops of dichlormethane make them strong again (so it seems it´s really PMMA!)
But then I measured the input resistance: 80k-ohm. I´m afraid of a broken decade A switch
to be continued...
Hello all - I've been watching this thread with interest as I have a sick 720A here with a broken shaft coupler. I bought some replacement couplers from Fluke recently - I will post a picture and P/N in a follow-up later.
I was wondering - can anyone recommend a procedure for removing/replacing the couplers for example in decade B or C? Is everyone taking the dials off? I'd like to minimize how far I have to disassemble the unit.
OK, much delayed but as promised current Fluke P/N for these shafts is
701979. I've attached a photo of the order envelope.
Also attached is a photo showing the 720A mid-replacement on one of the decade switches. I marked the dial position with some nail polish but a sharpie would work as well. I then loosened the two dial set-screws and unscrewed the metal coupler thing mounted to the bulkhead. This let me slide the metal shaft out the front just enough to play Operation and get the old broken acrylic shaft out and get the new black plastic (Nylon?!) one in. Buttoned it all up and the switch works great.
One caution is that some of the new black shafts have molding artifacts and may not slide freely into the multideck switches. I'll probably have to trim the worst ones before installation.
Hope it helps someone.
Bumping an old but very helpful thread ... I am now also an owner of a non functional 720A.
I'm seeing very slightly high input resistance (100.007k) when set to all zeros, but it jumps up to 100.033k when decade C is set to anything other than zero.
Output is "roughly ok" (a bit low on higher ratios) with decade A and B, but decade C on anything other than zero causes a jump about 12x what it should be and then no variability between settings, the remaining decades seem to be twice as sensitive as they should be (e.g. 20uV instead of 10uV.)
I need to run through a proper check, but from my quick dabbling late last night I think I was mostly inline with the output resistance table from the first page, but I saw around 40k output resistance once decade C was non-zero (with all else on 0.)
So my feeling at this point is that S3 is faulty, probably going partially open circuit when non-zero (and then the 40k would be the shunt?)
It does actually look like this decade has been removed or replaced before, it has similar paper labels on all of the wires (although these are printed rather than hand written) -- I'll post some photos this evening.
I'm a bit daunted by the prospect of removing this and trying to fix the wafers ... any advice on how to go about this?
Thanks.
An open contract at the switch sound like a reasonable failure mode. However, with an open contact the lower decades should not have any effect other than changing the resistance to the output a little.
Anyway an open contract should be measurable with a DMM even in circuit, at least in one of the first decades (could be tricky with very low resistance ones at the end).
If the decade was replaced before, there is a slight chance this replacement was a failed repair and the connections may be wrong around this decade.
Thanks Kleinstein ... on further investigation I think you are completely right. The switch does look to be functional.
It looks like they have swapped over the two red/white wires with the black one... TiN's original and a photo of mine attached.
There is a lot of slack on the black wire, and on TiN's one it actually goes around the back which would make more sense.
I'll have a thorough look for other wiring problems first, then I'll see if that does the trick.
EDIT -- yep, that fixed the problem. All seems absolutely spot on now!
Thanks for the pointers, and I'm really pleased I didn't start trying to take that switch out!
Just for completeness...
I had actually asked both Fluke (UK) and a reseller for pricing and availability for the C deck switch when I thought that might be the problem.
Fluke have been incredibly responsive and helpful which I thought was worth highlighting since seems to conflict with some other reports on this blog, that said, the cost of a new assembly (they didn't provide just the wafer set cost) was £4,165 presumably plus VAT!
They also suggested that a calibration of a 720A would be £1500.
Thought it might be of interest.
Hello to all Voltnuts !!
I am a very happy owner of a Fluke 720a KVD since some Days too.
Faulty A Decade Switch seams to be very common Problem, so my KVD also got that Problem.
On my KVD i measured 103 Ohms between Pins 19 and 20 of the Oilcan. I checked the Resistor inside the Oilcan by removing the Wire going to the Rotary Switch (Decade A). The Resistor shows the expected 9,9... KOhm and i found out, that the Rotary Switch of Decade A has a short between two Pins (grey and whites Wires).
All other Decades are OK.
Ill test the Switch today and i'll post some more Details soon. Hope that i can help others to repair this nice little Poti too.
Does any body know how to replace the Decade A Switch with a new Switch ? Which Brand of Switch will fit in there ? Maybe some Grayhill Stuff ?
I found one Provider named "
https://www.multi-tech-industries.com" on the Web. On that Homepage i saw rotary Switches (Model 820) that looked like the ones in the Fluke 720a.
CU
Stefan
