ESI 242D and 707B - how useful?

[Rax]

Actually it is. The aluminium box, containing the 419A printed circuit boards, even has an HP serial number plate.

But no trace of (mercury) battery, correct?

[manganin]

But no trace of (mercury) battery, correct?

Sorry, but I can't remember. Must be ten years since last opened. (And I sold the whole bridge to a forum member earlier this year.)

[alm]

To make full use of its precision, the resistors (first four decades) of its rheostat need to be adjusted (over the last five decades they might well have drifted out of the adjustable range though).  For calibration of it's rheostat, ESI suggests a complete system, i.e. with a 2nd rheostat (there will be other ways to accomplish the same, but adjusting a resistor to better than 20ppm accuracy is no small task).  I'm just afraid that without a set of calibrated standard resistors in a temperature controlled oil or air bath, the 242D is a heavy paperweight.

For best accuracy, you can use the 240 and 802 to compare two resistors that are close in value. Or use the 925 as tare resistor in parallel with one of the resistors, so its contribution to the uncertainty becomes much less.

To measure relative differences between resistors you don't need a calibrated standard resistor. You can measure that resistor1 is resistor2 + 5 uOhm/Ohm very well without knowing the value of either resistor precisely.

[retroware]

As someone who recently acquired a R925D to complete (I used to only have a the non-adjustable 925) my 242D system I have a couple of comments:

First, there is no mercury battery in the 419 equivalent null detector. The battery was for nulling out the input voltage. This is not needed when used as a bridge detector.

The thing I find fascinating about all of this early metrology gear is how accurate a mostly passive only system can be. The 242D specs are 10ppm from 1 ohm to 10M. This is about as good as the 90 day spec of the 3458 (and better at the high end of the scale). Furthermore, it can be self calibrated with only a single 10K standard resistor (the same as the 3458).

Granted, for almost every application I can think of I would use my 3458 as opposed to the 242D. But as a collector of vintage metrology gear, I find this system fascinating (plus who doesn't love adjusting 40+ pots!).

[Rax]

All this is so helpful, thank you all. Some conclusions:

• The real deal breaker may actually be the most prosaic of them all - size. I have not actually seen the instrument - just pictures from the seller, who is a very respectable dealer of sorts from this neck of the woods - so I never quite gotten a real sense. But I think I have enough now to work with, and space is not something I have an excess of. Now, I don't really have an urgent and frequent need for this (but more on this below). Frankly, one of the reasons that may have worked is if I could utilize the different parts for their separate duty, so I could replace some of my existing units with these, gain functionality, while not sacrificing space.
• The lack of the battery suggests the 419A within this is missing functionality the standalone has - which I haven't thought through (still far from home, with very little time and no manuals handy to map this out) - so that's much of a gain either.

The overall question on "what's this good for in my activity" hits at the core of a bigger question for me, which I admit I don't have a lot of clarity. Heck, I have no answers to. Which is what do I hope to do with all this, which is essentially a (very serious) hobby. So, fair question, but I'll likely have a rather vague and wishful answer to it.

[donlisms]

I am also fascinated by these early techniques.  There is no question that is much more convenient to use a 3458a, but it seems like an awful lot of things have to go right for such an instrument to produce good answers.  I read of the struggles in repair and calibration and cost, and I know the 3458 is not for me.

I real a lot of the earlier papers from the NBS and others, Frank Wenner, James Thomas, Edwin Northrup, and others, building on the work of Jimmy Maxwell and so on.  It is interesting (to me) to see how they did so much with so little, and figured out many concepts that we still rely on (and often make oh-so-much-more complicated!).

Bridge techniques rely much more closely and simply on the fundamentals of physics.  The behavior of current flow, and the reliance on a relatively simple and self-verifiable null detector.  It's all the ratios.  Direct or only-slightly-indirect comparison of a standard resistor against the unknown, where the measurement is determined by a variable resistor whose accuracy and characteristics are improved by a factor of 10,000 by putting it in the middle of a much larger fixed resistance, that sort of thing.  Clever techniques based on the simple mechanisms available in the day.  Measurements made at ppm's without any amplifiers.  Without any amplifiers.  Verifiable ratio techniques throughout, and somewhere... one (or more) calibrated standard resistor(s) to provide the absolute reference(s).  Some (provable) Hamon buildups to scale the ratios up and down, verify linearity, and so on.

At last for resistance.  I suppose there are similar techniques for voltage measurement, but I don't pursue that at this juncture.

My impression of modern test gear is that measurements pretty much always rely on internal resistors, anyway.   They're inside there, and there is no normal hands-on with them; you may not think about them specifically, and just kinda trust that they're still right.  It's way easier, and I don't discount that, but to my thinking, there are more assumptions involved.  (And money!).  You don't have to (but probably will anyway!) make a hobby of making/keeping it all working right.  :-).

And you probably SHOULD have at least a couple of calibrated resistance standards around your lab, anyway, no?  So it seems like it sorta comes down to letting your meter make a comparison with its internal standards, or having your standards out on your bench (or in the 925D, which is not necessarily required).
--
It is 19.5 inches wide ((standard 19" rack), 20" tall (3sp + 3sp + 4sp) * 1.75 + filler strips), and 12.5" deep.  I don't know what that is in metric inches, but I'm pretty sure it's still... big.  It's liftable by one person, occasionally.  I am putting mine a rolling wire-shelf cart that fits it pretty snugly, and then there's space for other gizmos on the shelves, and I can move it away when I'm not playing with it.

I think I would only get one if I (1) really cared about fundamentally-implemented resistance measurements, enough to invest more time and effort in, for example, building some good Hamon standards, and working on calibration, and (2) had the space, or could make the space, and (3) had the money thing.

[Rax]

I think I would only get one if I (1) really cared about fundamentally-implemented resistance measurements, enough to invest more time and effort in, for example, building some good Hamon standards, and working on calibration, and (2) had the space, or could make the space, and (3) had the money thing.

What I have around already providing roughly what this offers (a Prema 6048 - so 8.5 digit capabilities -, a full set of ESI SR1s, a Fluke 5450, an ESI RV722, and a bunch of other such things) has the advantage of being economical on space. I actually have a very real problem of lack of space and have to consider getting rid of some of the units I have here. And hobby budget is obviously limited, so.

Some of this type of mechanical self-adjusting was replaced in newer models by all sorts of selfcal and int cal and other such ADC/data driven approaches. It's pretty interesting from a historical perspective, actually.