Fluke 720A - condition assessment

[Rax]

I have a 720A at my bench which doesn't seem to allow for self-calibration. Namely, 2-24 cannot be adjusted for a null.

To do this, I am using 20V from a 5440A, and nulling with a Keithley 155. The BRIDGE BALANCE pot cannot adjust for a zero, though some sort of low point can be observed. It's just far from zero.

Welcome thoughts and input.

[RandallMcRee]

Hey Rax,

You would probably get better suggestions if you list the available equipment you have for testing.

Knowing that you have a 5540 I think you could do some good INL testing just by feeding 10v into the KVD and looking for deviations to the 5440 using your Solartron.

The KVD cal is pretty much just for the first two decades. (I have a unit with a similar problem, btw).

Randall

[Dr. Frank]

to both of you:
I have published a comprehensive article about analysing and re adjusting a 720A.
You only need to search for this article.
Rax, your equipment is just fine.

it's here:
https://www.eevblog.com/forum/metrology/fluke-720a-kelvin-varley-divider-repair-calibration-linearity-validation/msg3897077/#msg3897077

[Rax]

Rax, your equipment is just fine.

Frank,
The article seems to indicate that a 720A which cannot be zeroed requires internal calibration. You state p.4-14 through 4-18, which would be the entire calibration procedure (4-19 CALIBRATION and 4-25 Bridge Calibration), and not just the latter, as your article states, but maybe I'm misreading it?

That stretches it beyond what I'd regard as "being fine" (this is an instrument I don't yet own, btw), being that it's an "open case" calibration procedure. But maybe you're making a different point.

[Rax]

Frank,
The article seems to indicate that a 720A which cannot be zeroed requires internal calibration. You state p.4-14 through 4-18, which would be the entire calibration procedure (4-19 CALIBRATION and 4-25 Bridge Calibration, and not just the latter, as your article states, but maybe I'm misreading it?).

I think that's a typo in the article. It seems to me the procedure required in case the bridge cannot be nulled is 4-25 Bridge Calibration, which starts at p.4-15.

[Rax]

I find it very strange that the manual says nothing at 2-24. SELF-CALIBRATION PROCEDURE about what to do if one cannot zero it. At least, I don't find clarity in the language.

The two statements regarding the fact that one needs to perform 4-25 in case nulling is impossible I am able to find are at:

• p.4-14 (which is within the Divider Calibration, so I'd not assume it's applicable)
• p.4-15, but this one also ties it to the same, and it additionally says that "[...] the procedure is required only if a null could not be obtained during Divider Calibration [...]"

[Rax]

You only need to search for this article.

I was aware of the article. If you look at it, you'll see I thanked you for the work and authoring it a long time ago (I read it then). I just didn't know (or recall) it discusses this very particular topic.

Maybe just me, but sometimes I (re)encounter a problem from a completely distinct vantage point - personal involvement for one - and I'd not (at least immediately) reconnect with the prior experience of it. I read your article many months ago from the vantage point of the 5440 you use in it - which I had - but not at all from that of the 720A - which I did not.

Many thanks for pointing out it discusses "open-case" tweaking the null of the bridge.

[Dr. Frank]

You only need to search for this article.

I was aware of the article. If you look at it, you'll see I thanked you for the work and authoring it a long time ago (I read it then). I just didn't know (or recall) it discusses this very particular topic.

Maybe just me, but sometimes I (re)encounter a problem from a completely distinct vantage point - personal involvement for one - and I'd not (at least immediately) reconnect with the prior experience of it. I read your article many months ago from the vantage point of the 5440 you use in it - which I had - but not at all from that of the 720A - which I did not.

Many thanks for pointing out it discusses "open-case" tweaking the null of the bridge.

Hello Rax,
however it is, I truly hope, that this basic, internal bridge alignment will solve your problem, i.e. that it's not necessary to look after resistors who might have drifted too far.
Frank

[Rax]

Done. Thank you, Frank.

The R203 control was much more "jumpy" than I thought it'd be (there's a very narrow "range of adjustment" in the middle of the run of R203). Not very smooth as far as indication goes - so I ran it a few times from side to side, as per manual - but I'm happy with where it landed.

[Rax]

Inputting 10V into it and trying to get out ratios of it outputs weird voltages out (nonsensical). I probably need to run through the gamut checking it out at different settings - so I'd hopefully isolate the trouble to something specific and maybe fixable - but I'm either doing something wrong, or this has experienced some serious abuse.

I pulled my RV722 for a sanity check and, if anything, I'm getting even closer to spec readings from if than last I tinkered with it.

I may have to reopen it and see what could be wrong. For whatever that's worth, it looked very clean inside. I mean, if there's anything in the world of electronics I'd expect to be clean inside, this is it.

[alm]

The R203 control was much more "jumpy" than I thought it'd be (there's a very narrow "range of adjustment" in the middle of the run of R203). Not very smooth as far as indication goes - so I ran it a few times from side to side, as per manual - but I'm happy with where it landed.

As the manual says, R203 is a coarse and fine pot into one. So a, small rotation range is the fine adjust and if you go outside that range, you hit the coarse adjust.

Of course it's also possible that the wipers aren't making good contact. You can pop off the lid at the front of the pot if you want to clean it. It's press-fit, at least the one I saw was.

Inputting 10V into it and trying to get out ratios of it outputs weird voltages out (nonsensical).

Stupid question maybe, but did you set both the internal and external function switches back to operating and use the correct binding posts for wiring? The F720A is somewhat sensitive to input impedance / loading, but 10 MOhm should give no gross error.

[Rax]

As the manual says, R203 is a coarse and fine pot into one. So a, small rotation range is the fine adjust and if you go outside that range, you hit the coarse adjust.

I didn't quite understand what they meant with that. I assumed there may be two coaxial pots but upon looking closer that was not the case. Thank you for clarifying.

Of course it's also possible that the wipers aren't making good contact.

They seem to specifically provide direction on how to address if this is the case, so I assumed the "jumpiness" of the control is a sign of this issue. The fine vs. coarse action of the pot makes a lot more sense. I did run the pot through the entire range a few times, which can't hurt.

Stupid question maybe, but did you set both the internal and external function switches back to operating and use the correct binding posts for wiring?

I don't think there's stupid questions with this, alm (and you're at the bottom of the list of those I'd expect that from). If I'd think that, I'd also think no one fixing or generally tinkering with these things has ever done anything untoward (or... fine: stupid), and I know that not to be the case. At least yours truly has done their (thankfully, small) share. Lack of infallibility is why most of us are here, and the demiurges can hang out some other place.

Now, along those lines, as far as I can tell, the manual states the bridge balance adjustment is done with the internal switch in OPR position. Which is also where the unit needs to be in regular operation (right?). In fact, I removed both panels (top and bottom), to have access to the internal switch, only to realize I just needed to remove the top for what I was looking to accomplish. Of course, I'd not have known the internal switch is already in OPR position unless I removed the bottom panel to see that.

I should also say that, after the bridge balance procedure completed successfully, self-calibration seemed to not need any adjustments (except for one control, I think it was S2?).

[Rax]

Did some more testing this morning. Readings are hard to make sense of (to me), but there's a pattern. This is kinda good, kinda bad, as it seems to point to something specific being broken, but also seems to confirm there's something broken (and not just "user error"  ).

If I input 10V from the 5440A and I adjust the decades from .0000000 to .5000000, I get the following.

• .0000000: 1.8181818V
• .1000000: 2.7272727V
• .2000000: 3.6363635
• .3000000: 4.5454545V
• .4000000: 5.4545470V
• .5000000: 6.3636374V

I didn't rebalance the bridge before doing this (I doubt that would fundamentally change the above readings) and the temperature is maybe a couple of degrees at the bench lower than then (20.0C currently).

I don't really know what to make of these results.

[alm]

This is so far off that it's definitely not down to adjustment of any sort. My money is on either pilot error or a major defect like an open connection. Please show a picture of the front panel showing all the connections and switch settings, and the setting of the internal and external function switches. And explain to which terminals the Fluke 5440A and DMM are connected.

[The Soulman]

Bad switch not connecting to the bottom of the resistor ladder. 

[Rax]

Please show a picture of the front panel showing all the connections and switch settings, and the setting of the internal and external function switches. And explain to which terminals the Fluke 5440A and DMM are connected.

I forgot to attach it to the previous post as I intended. Enclosed here.

[The Soulman]

[alm]

Connections look fine. Assuming the internal function switch is also set to operate, my guess would indeed be an open connection either on the switch contact or the wiring to the switches. Try rotating all switches through all settings a couple of times, and do a visual inspection to see if you see any loose wires.