Muirhead KVD and Internal Corrosion - Ignore it or fix it?

[grizewald]

I'm looking for some advice from the expert restorers here on the forum.

A while ago, I picked up a Muirhead K-175-E Kelvin Varley divider. When I bought it, it looked a bit sad and neglected.



After cleaning it externally and restoring the outside to its former glory, it also turned out to still be very accurate. Here it is dividing down a 10V reference.



The inside is a different story though. The divider had previously lived at RAF Sealand on the northern coast of Wales and it would seem that the salty sea air had penetrated the divider and started some rather nasty looking corrosion. Here's a few pics from the inside.







It now lives far away from salty sea air, but I wonder if the corrosion is likely to worsen.
Should I leave it well alone or should I try to remove the corrosion to prevent it from getting worse?
If removing the corrosion is a good idea, does anyone have any suggestions as to how I should best remove it?

The corrosion hasn't affected the resistance wire at all, or the silver switch contacts, but other metals and the solder in particular seem to have been quite badly affected. I'd hate to see this classic and trustworthy piece of test equipment succumb to the damage that has started, so any good advice would be much appreciated.

[Gyro]

The corrosion hasn't affected the resistance wire at all, or the silver switch contacts, but other metals and the solder in particular seem to have been quite badly affected. I'd hate to see this classic and trustworthy piece of test equipment succumb to the damage that has started, so any good advice would be much appreciated.

That just goes to show the quality of those old Muirhead switches.    I think the contacts are self-lubricating Silver-Graphite btw.

You are wise to be cautious of cleaning. I would avoid adding moisture at all costs, my approach would be to gently brush and vacuum off any accessibly deposits. If they are removed dry then they can't cause any further harm in the presence of moisture. I certainly wouldn't do anything like trying to reflow any of the soldered connections.

For any corroded metalwork, such as the pressure arm of the switch in your photo, again I would remove surface deposits dry, and then apply a thin smear of Vaseline to the surface to prevent any further corrosion.

I notice that yours is one of the earlier wooden cased models. The lacquer should have mostly prevented salt absorption on the outside. The wood should have long since dried out, but give it a good brush inside.

I've attached the brief spec page from the catalogue in case you don't have it.

[grizewald]

Thanks Gyro, that's good advice. I'll see what some small brushes can do, starting with toothbrush soft.. The wooden case is in great condition. All it needed was a damp microfibre cloth and some label remover for the inevitable tape and stickers. The original lacquer finish still shines like the day it was made.

Like yourself, I'm kind of a fan of Muirhead test equipment...



Not pictured is a D-402-A Weston cell which is still in perfect working order despite being nearly 53 years old and surviving being posted from the UK to Sweden!

I keep looking to see if I can complete the whole set of the 825 series resistance decades. Not only do they look great, they're really accurate if you don't get a fried one.

[Gyro]

You've got a really nice collection there - even if you have plundered the UK. They shine up very well and the engrave-and-fill lettering is easy to freshen up, unlike silkscreen.

I would go really easy on the brushing, certainly nothing harder than a soft toothbrush, and even then, avoid the higher value resistors, which should still be well lacquered in any case.

Yes I love the construction quality of old Muirhead stuff. I envy you the KVD, I only have the 4 dial D-801-D Rayleigh divider. It exceeds it's spec by a very good margin though... https://www.eevblog.com/forum/testgear/kelvin-varley-reistor-porn/msg773770/#msg773770. I was lucky with that one as they tend to get pressed into service and abused as resistance boxes, less of a risk with a KVD.

I also have a K-375-C bare saturated cell that I still need to mount in an Aluminium block for temperature stability (I had forgotten that project on my list). Their saturated cells tend to last pretty much forever - life is mostly based on overall capacity consumed rather than calendar age. There are a few threads on here about them. I have ambitions to find a D-402-B so that I can do differential comparisons between the two cells. One day!

[tggzzz]

Not pictured is a D-402-A Weston cell which is still in perfect working order despite being nearly 53 years old and surviving being posted from the UK to Sweden!

Pah! That's modern...





... and note the date: 19-1-49, i.e. 71 years old and still in spec

FFI: https://entertaininghacks.wordpress.com/2016/12/28/a-weston-standard-cell-an-introduction-to-voltnuttery/

[tggzzz]

I have ambitions to find a D-402-B so that I can do differential comparisons between the two cells. One day!

Around 40uV for mine

[grizewald]

You've got me there tggzzz! Mine really is modern by comparison.



The cell was made on the 9th of March 1967



It's interesting to see that they used way more cadmium sulphate crystals in my cell than yours.



In your cells, the layer of crystals stops at the constriction in the lower parts of the tube whereas in mine, they go nearly all the way up to the connection between the two tubes. You would have thought that all those crystals would keep the mercury at the bottom of the negative side where it should be, but as you can see, there are three distinct balls of mercury on top of the crystals.

The Weston cell still beats modern voltage references when it comes to noise. I have two 10V references, one based on the LM399 and one based on the LT1021. The LM399 reference displays around 15μV of noise and the LT1021 about 20μV. If the Weston cell has any noise, it's less than I can measure. The lack of noise is what proved to me how stable my 7075 is and what made me decide to get it properly adjusted and calibrated.



1.018480V is 70μV high for the ambient temperature but the reading on the 7075 didn't change by even 1μV over 1 hour.

[tggzzz]

The Weston cell still beats modern voltage references when it comes to noise. I have two 10V references, one based on the LM399 and one based on the LT1021. The LM399 reference displays around 15μV of noise and the LT1021 about 20μV. If the Weston cell has any noise, it's less than I can measure. The lack of noise is what proved to me how stable my 7075 is and what made me decide to get it properly adjusted and calibrated.

Precisely.

One of my ancient zeners shows popcorn noise, another doesn't. Here are two graphs over a day, the first of a Weston cell, the second the better zener. Note the different vertical scales, 10uV/div and 1uV+10uV/div! If I have calculated it correctly, the Weston's noise is around 0.2uV and the zener's around 1.5uV, but the zener is 10V output.

But, as you can see, the Weston cell can also be used as a thermometer!

[grizewald]

Once I have the 7075 calibrated, I'll be hacking together something to plug into the BCD control interface which I got with the meter so that I can do some proper logging. An Arduino Mega 2560 has just enough I/O pins for the job.