Electro Instruments (Non Linear Systems related?) DVM

[Cubdriver]

This arrived today from the 'bay.  It was sold as a Non Linear Systems DVM, but its actual markings show the manufacturer as being a company called Electro Instruments Inc.  It certainly LOOKS like the early NLS meters, with edge lit stacked plastic numerals for the display.  I haven't found a date code on anything yet, so the best I can guess at this point is that it's late 50s to perhaps early 60s.

Here are a few preliminary photos of it as it was received; the packaging left a LOT to be desired and I'm amazed that it got to me relatively unscathed.  I think this one is going to be quite the challenge to get up and running.  If anyone recognizes it or has any hints/links regarding its pedigree, I'm most interested to hear them.  I'm going to go do some disassembly and cleaning of it now; I'll post more detailed photos once I've taken them.  In the meantime, here are some taken of it with the iPhone.

Overall view, undergoing inspection by Model 3 Lab Cat:


A close up of the front panel lit by a strong side light to make indented lettering visible - no NLS logo, but Electro Instruments instead:


And the rear panel - the serial/model number tag is illegible (thus far; may try the microscope later), but a model and serial number are written in magic marker on the rear panel model 1040, S/N 3186 - these are what I'm going by unless anyone can say "No, it's actually an EI model <fill in the blank>:


And finally a general pic of the interior will have to do for now as I realize I haven't taken an overall pic of the open chassis:


More to follow as I take them, and these and others are at:
https://pmanning.smugmug.com/Electronics/Electro-Instruments-NLS-edge

-Pat

[med6753]

It looks like it WAS literally used as a boat anchor.  That's going to be some challenge to get running again.

[Vgkid]

Good luck with that repair   

[Seekonk]

I have an old NLS somewhere.  The only thing it was good for was charging a cap at the input and listening to it for about 5 minutes as it discharged.  Sounded like someone putting a thousand quarters into an old Coke machine.

[PaulAm]

That will be a challenging restoration, but at least I see your assistant is on top of things.

[edpalmer42]

How much did you charge them to take it off their hands? 

[retiredcaps]

How much did you charge them to take it off their hands? 

Close to $100 USD for the unit.

http://www.ebay.com/itm/282088885809

[Cubdriver]

That will be a challenging restoration, but at least I see your assistant is on top of things.

   Yeah, it's a little rough.  I have torn it down further, and now have the relay/display section by itself on the bench.  The electronics themselves don't appear to be in terrible shape, but I haven't yet begun checking out the individual components yet.  On the bright side, it's not packed to the gills with things like the typical HP stuff I play with is.  On the bad side, I can usually get a manual for said packed-to-the-gills HP stuff.  I've found nothing other than a bit of history on the company as far as this one goes.  Apparently Electro Instruments was formed by a couple of guys who left NLS to build their own meter.  I guess that explains the striking similarity to the NLS units.  My fingers are crossed that finding some NLS schematics might give some insight into the workings of this one.

Left side - power supply and unknown board (at least partly power supply, I'm assuming, based on the regulator tube and filter capacitors) - tubes are 0A2 and 0B2 regulators, (2) 5751 dual triodes and a 6136, which is a high reliability 'computer' version of the venerable 6AU6:


Right side - amplifier PCB - tubes are (2) 0A2 gas regulators, (2) 2D21 thyratrons, (2) 5696 thyratrons, (2) 12AX7A dual triodes and finally a 5751 dual triode:


-Pat

[oldway]

The cat search mice and it is likely to find a few 

[med6753]

Obviously the filter caps in the power supply need to go. What is not so obvious is that those two blue selenium rectifier stacks need to go too. Which can be a little problematic because selenium rectifiers have a greater forward voltage drop than silicon diodes. You'll need to experiment with series dropping resistors to bring the B+ in spec.   

[Cubdriver]

Yeah, it's probably a safe bet that those filter caps are toast after all this time.  I'll take a crack at re-forming them but seriously doubt it's going to happen.  Figuring out the ESR of the selenium bridges will then be the next challenge.

-Pat

[Kleinstein]

The selenium bridges might still work, at least one of the diodes might allow a measurement of the resistance. There should be reasonable formula / data to calculate the resistance / drop from the number and size (in mm) of the diodes.

As this will be more a historic instrument than anything practically useful,  one should keep much of the original, even if better part exist. It's more like restoration of an old-timer than a regular repair. So if the selenium rectifier is still working one should keep it. If in doubt with breakdown voltage, maybe add modern diodes in series. For the caps, I would at least keep the case, and maybe fit new one inside.

The circuit is not than complicated - so reverse engineering should be possible.

[Cubdriver]

The selenium bridges might still work, at least one of the diodes might allow a measurement of the resistance. There should be reasonable formula / data to calculate the resistance / drop from the number and size (in mm) of the diodes.

As this will be more a historic instrument than anything practically useful,  one should keep much of the original, even if better part exist. It's more like restoration of an old-timer than a regular repair. So if the selenium rectifier is still working one should keep it. If in doubt with breakdown voltage, maybe add modern diodes in series.

I am actually inclined to leave them in place and in use if they seem to be functional.  I have a Sams book from the late 50s on metallic rectifiers; it was written by Leonard Crow.  I need to do some reading on them.

For the caps, I would at least keep the case, and maybe fit new one inside.

The circuit is not than complicated - so reverse engineering should be possible.

The caps, if bad, will be re-stuffed.  I've had good success doing that thus far:
https://pmanning.smugmug.com/Electronics/Restuffing-can-electrolytics

I'll likely continue with the cleaning and disassembly tonight, with more pictures to come.

-Pat

[med6753]

Selenium rectifiers have a finite lifespan and do age over time, even when not used. The result is increased voltage drop and possible overheating. And the magic smoke that they release is toxic.

I suggest reading this....

http://w3hwj.com/index_files/RBSelenium2.pdf

[Cubdriver]

Thanks for the link.  It sounds now very likely that they'll be going bye-bye.  If they're functional, I can measure the voltage drop to more easily size the required series resistor.  I appreciate the additional info!

-Pat

[AF6LJ]

Good luck with that, it has 1960s written all over it.

[Cubdriver]

Yes, definitely late 50s to mid 60s vintage.  I still haven't spotted anything with a definite date code (other than tubes, which may have been changed - one had a 1957 date on it, while others are from the mid 60s), so I have no certain answer to its age.

Here are some more photos of the relay/display section.

The relay/display assembly removed from the chassis:


In this image, it is inverted and the bottom part of the enclosure has been removed.  The sound deadening foam has long ago hardened and begun to crumble, and I'm sure the fact that two of the stepper relays have broken their shock mounts and were bouncing around in there helped to further pulverize it.  What a mess!:


Yum, foam residue!     Thankfully, it's dry and sweeps/vacuums up pretty easily.


This is what remained on the inside of the enclosure.  It looks like some kind of adhesive was painted on the surface of the cover, and then the foam sheets glued in place.


-Pat

[Cubdriver]

The stepper relays were covered with foam dust, but seem to have by and large cleaned up nicely with a light brushing/vacuuming.  I'm still planning to blow them out with low pressure compressed air (don't want to bugger up the contacts), but haven't dragged it out to the garage to do so yet.  I have an old book about the stepper relays on order; it should arrive in a week or so.

Relay coil before cleaning:


The same relay coil after being brushed off/vacuumed:


And now the contacts before cleaning:


And after:


And views of the complete relay, before and after:



These are the connections to the lamps that illuminate the lexan digit plates:



-Pat

[AF6LJ]

The potential for that working is quite high.
/The Cat Lady wonders what they were using for a voltage reference.

[Cubdriver]

Well, the Cat Lady piqued my curiosity and prompted me to do a little digging.

I grabbed the manual for one of the early NLS meters from BAMA and saw the standard cell called out, realizing that it resembled something on my meter than I hadn't looked closely at thus far.  There is a small glass standard cell hiding behind a protective metal cover on the back side of the smaller circuit board on what I've been calling the power supply side of the meter.  Apparently this board also provides the voltage reference.  I can't imagine that it's much good at this point in time, likely being at least 50 years old and having been stored under lord only knows what conditions before I got it, but here are some photos.  I suppose I should read up on and learn about standard cells, seeing as the museum I volunteer at has things from the Eppley standard cell laboratory.

An overview of the side with the cell in the corner:


View of cell behind protective cover:


'Front' of cell:


'Back' of cell:


It still has liquid in it, but I'm unsure if there's any output and until learning more I'm not inclined to try to measure it considering that from what I know they put out practically NO current, and can be damaged if overloaded.  If by some miracle it's still functional, I don't want to trash it by being ham fisted and improperly trying to measure it.

-Pat

[Gyro]

Wow, a DVM with a Standard Cell in it! 

It must have got a fair amount of shaking in operation though.

[AF6LJ]

Yah it is a Weston Cell.

[SeanB]

Probably still works, though it will have drifted a lot by now. Easy to measure the voltage with a modern DVM, especially one with a 2V range which is over 11m in input resistance.  Does look like the one seal has corroded though, letting out the magic mercury filled acid a little

[Cubdriver]

Wow, a DVM with a Standard Cell in it! 

It must have got a fair amount of shaking in operation though.

Yeah, I suspect it might have gotten a teeny tiny bit jostled during shipping - check out reply #62 in the 'How do people pack delicate instruments properly?' thread in this repair section...   

-Pat

[Cubdriver]

Probably still works, though it will have drifted a lot by now. Easy to measure the voltage with a modern DVM, especially one with a 2V range which is over 11m in input resistance.  Does look like the one seal has corroded though, letting out the magic mercury filled acid a little

I'll have to check it out later.

-Pat

[SeanB]

I came across an old Heathkit bench meter a whle ago that used a mercury button cell as a reference, but the cell had long before corroded away, and some previous owner had replaced it with the next best thing, a single LR44 cell, which of course was also dead and corroded.

[Cubdriver]

I came across an old Heathkit bench meter a whle ago that used a mercury button cell as a reference, but the cell had long before corroded away, and some previous owner had replaced it with the next best thing, a single LR44 cell, which of course was also dead and corroded.

I recently got an HP 740B DC Standard/Differential Voltmeter and it has a dead mercury battery in it as a reference.  I need to figure out what to do to replace it at some point.

-Pat

[SeanB]

LR44 and a shottky diode is a quick and dirty substitute, but if there is room put in a TLV431 and adjust it to the right voltage. 1.35V is within the adjustment range, though if you want a better one just take a LM10 and use it, which will give a buffered reference with slightly better performance, or at least improved PSSR.

http://www.ti.com/lit/ds/symlink/tlv431.pdf

Just remember not to use the TL431, very easy to confuse the 2.

[Cubdriver]

The mention of the reference got me searching on mercury battery substitutes, and I saw SR44 silver oxide button cells suggested as a possible alternative (with a series diode to drop the voltage) on several camera sites - apparently this chemistry has a relatively flat discharge curve similar to that of the old mercury cells.  However, seeing as the 740B is an old HP boat anchor, the installation of a TLV431 or LM10 and their associated support circuitry should not be a huge issue space-wise.  Thanks for the suggestions - more things to dig into.  (in my 'spare' time!!)

-Pat

[bob91343]

Here is a pic of mine.