Muirhead STD and REF cells - all catalogue info uploaded !. (AYCE)

[lowimpedance]

 Okay since I had a request to post some info on the Muirhead K375 cell , and as I had the old catalogues out of the filing cabinent for what it appears to be the first time since the late sixties  , I have copied all there was in the folder pertaining to wet cell standards and attached here for everyone's enjoyment/archive.

 Muirhead made quite an array of different instrumentation from standards lab gear to general components and gear. The small number of instruments I have come accross are of the highest quality.
 I wonder just how many examples of their gear still exist?

 There is some quite good info locked away in these old catalogues, (and this also applies to many other equipment manufacturers that have long since vanished leaving only old gear sitting around in a cupboard as a memento of times past!).

[lowimpedance]

More !

[lowimpedance]

still more !!

[lowimpedance]

Part 1 of a "Muirhead Technique" journal of instrument engineering Vol 1 no 3 .
Quite an interesting read.


Sorry I had to split into 2 parts to fit on the forum (and the original was a non std paper size which was a real PITA)

[lowimpedance]

and lastly Pt 2..

[Vgkid]

Thanks for these.An interesting read, even though they won't help me. I dont plan on having any Eppley cells . 
Do you happen to have any manuals for L&N gear?

[lowimpedance]

Thanks for these.An interesting read, even though they won't help me. I dont plan on having any Eppley cells . 
Do you happen to have any manuals for L&N gear?

Catalogues or equipment manuals?

[Gyro]

Thanks very much lowimpedance, very interesting reading and the effort is much appreciated. I suspect that you probably have one of the few archives still in accessible captivity... though I think Woodchips may have some stuff too.

Edit: The K-375-C appears to be a saturated cell. I know that saturated cells are longer lived but I wonder how long is long in this context, the datasheet only mentions that the cell will remain within spec as long as the current drawn does not exceed 1 Coulomb during the required life of the cell.

[Vgkid]

Thanks for these.An interesting read, even though they won't help me. I dont plan on having any Eppley cells . 
Do you happen to have any manuals for L&N gear?

Catalogues or equipment manuals?

both, just list what you have if you want.

[lowimpedance]

Thanks for these.An interesting read, even though they won't help me. I dont plan on having any Eppley cells . 
Do you happen to have any manuals for L&N gear?

Catalogues or equipment manuals?

both, just list what you have if you want.

 I have some catalogues to hand, but the equipment is a bit more scattered and that is what has not been thrown away !. So I need to go 'hunting' for any manuals that have outlived the actual instrument by hiding in filing cabinets etc. Therefore its better to ask for a specific bit of kit and then if it rings a bell I can see if it still exists.
 I do have quite a range of the old catalogues from various manufacturers of old time 'standards lab' equipment so I have something that would be of interest I can scan the odd page or two if the info has not already been posted.

[Gyro]

I'm pleased to say that my Muirhead K-375-C Standard cell arrived intact today.

The cell really is tiny at just over 40mm long by 12mm diameter. I can see why they describe it as PCB mountable in the datasheet (which lowimpedance kindly provided in the thread above). It has probably been shaken up a bit in the post, but Muirhead describe it transportable in any position. It came, still sealed in it's bag with un-striped leads. I worried initially what the black bits loose inside the bag were until I realised that they were the remains of the rubber band that tied the leads.

I've carried out an initial voltage measurement using my test setup of Geller SVR-T connected to my home built 10:1 Hamon divider and measuring the voltage differential with my Datron 1041M on 10mV range (10G input). The voltage has settled at 1.018732V at 22.8'C which seems within spec (though at the high end).

I'm still a bit confused about whether this is a saturated or unsaturated cell. The only clue in the datasheet is where it references temperature coefficient where it says: "EMF as per internationally agreed formula for 0.05N acid electrolyte saturated cells" (-40uV/'C).

Here's where I got it if anyone else is interested in a play:

http://www.ebay.co.uk/itm/STANDARD-CELL-Battery-MUIRHEAD-K-375-C-C1970-/322251421354?hash=item4b07ae76aa:g:QT0AAOSwe7BWzxMl

Edit: The Muirhead Journal... pages that lowimpedance uploaded are also related to this particular cell model. Thanks again. 

Edit2: Just looking over the journal pages again, it is a saturated cell.

[Assafl]

Okay - so this thread made me get the Muirhead I had seen at a local scrap yard and try to revive it. It is a K-231-A Temperature controlled standard cell assembled in 1967 (on 5-10 I assume May 10).

It had two main problems: 1. a disconnected wire; 2. intermittent pot. Unfortunately all the wires were frayed and a short fried the 9.5V zener so I replaced it with a pair of 5.1v and 4.3v matched to measure 9.5v. 

@lowimpedance - much appreciate the datasheet for the unit. Really helped me understand that the cell was on spec and what the intent was for the design. Do you have additional info about the K-231-A model? Especially a calibration guide/service manual would really help.

As I am working on getting it back to shape - I took pictures and reverse engineered the circuit. The soldering is really horrible but all the components seem original - so I guess mine was soldered by an intern. BTW - Looking at the parts resellers from 1967 era magazines - this was one expensive device - just the zener diodes were over 10 English pounds each (in 1967 pounds...).

Here it is (for posterity):

Edit: There is an error in the calibration process. Adjust the black pot (inner oven thermistor) at a low ambient temp and the red (outside temp calibration) at the higher ambient temp. Repeat. It helps to push down on the thermometer so it pushes against the bottom bakelite and the response is faster. When temperature there is stable to 0.05C the (much heavier) cell temp is rock solid.

[Assafl]

So the K-231-A now holds temperature within +-0.02C. Voltage I measure is 1.017688V on the recently calibrated (measured - not adjusted) HP34401A. At 1V range the offset at the calibration report was about +10uV.

According to the datasheet, the voltage at 37.5C should be 1.017675V. I guess to that one should multiply by the 9ppm of the NIST 1990 change in the volt (the spec was from 1966). Hence the expected voltage should be about 1.017685V.

Now the spec further states 64uV per C. The last calibrated in 2013 thermometer is precise to 0.01 but accurate to +- 0.05C. So I expect a max of 3.2uV offset.

If my reasoning is correct - seems like the cell is still within calibration. Nifty for a device made 50 years ago in 1967.

 

[manganin]

As I am working on getting it back to shape - I took pictures and reverse engineered the circuit.

Very interesting. Could you please post more photos of the mechanical/thermal construction? Especially the owen mass, insulation and locating the two thermistors.

[Assafl]

I apologize for not taking more pictures but now that I have the device assembled and the oven control calibrated I don't want to disrupt the cell any more (It is stabilizing now). So I'll have to answer with what pictures I have taken to facilitate reassembly (as wires tended to break at the solder joints) accompanied bu a textual description. I hope it helps...

First 2 pictures (don't know why the images are rotated) - this is the cell "baking" - cell temp is about 37.2C. The brownish wire coming out of the thermal well is a PT1000 RTD.





Oven is a stainless steel cylinder about 5cm diameter and about 9-10cm long. It has a connection panel below it. In the picture you can see the outer jacket with the expanded polystyrene foam insulation. The insulation cover all of the oven - with cutouts for the wires, thermometer well, copper connections, and wires.



Under the bottom cover a Mullard OC28 serves as the heater. See diagram in previous post. It is connected to a steel panel that has a bakelite standard cell and PCB holder. It is barely visible under the board in the next picture.



In the picture the jacket is also visible in the background. If you look closely at the bottom bracket you'll see that thermistor well for the GT25 thermistor (the red wire going into the well is one of the thermistor wires). The thermistor measures the heater temp - not the cell temp.



The top of the cell is held by a bracket that also houses the two trimmer potentiometers (Red and Black). The cutout in the bracket is the shape of the cell with a small indent to hold a thermometer against the cell.





Not pictured:
I do not have a picture of the cell - but it is encapsulated in hard heatsink compound, wrapped in a stainless jacket and then with something like a PVC jacket over it. It is held snugly by the top and bottom brackets.

Outer thermistor - an indent is molded in the Styrofoam and the GT24 is snugly connected to the outer aluminum jacket by a bracket and 2 screws.

At the opposite side (and bottom - probably trying to get this heat source as far away as possible from the thermistor and cell) in a square indent in the styrofoam - is located the tiny box with the resistors and zener.

A very nice device. Sort of like a poor man's LTZ1000 or LM399 - albeit I am not sure these were cheap when purchased back in 1966....

[Assafl]

I took a picture to show where the external thermistor is located. It is connected inside the aluminum sleeve using the two screws visible in the photo and a bracket.

(Good grief - why are all the photos rotated??? Is it the iPhone? Outlook? the Forum?)

[VintageNut]

Excellent work resurrecting this ancient voltage standard. Congratulations. Anybody who frequents this metrology forum would be proud to have have it running in their lab.

These cells are good because they do not degrade over time but require temperature control.

Does your analysis of the manufacture date measurement compared to today's measurement include the change in the definition of the volt as reported by Dr. Frank in his #1000 post?

[Assafl]

Excellent work resurrecting this ancient voltage standard. Congratulations. Anybody who frequents this metrology forum would be proud to have have it running in their lab.

These cells are good because they do not degrade over time but require temperature control.

Does your analysis of the manufacture date measurement compared to today's measurement include the change in the definition of the volt as reported by Dr. Frank in his #1000 post?

If it were not for this forum this device would sit at the junkyard eventually to be discarded. Guy had no idea what it was. Frank's and your posts reminded me that I saw that device and eventually I went to pick it up. He got greedy and demanded 50$. I think it was a great deal.

Well, it looks good but I am not sure I have enough data or capability to actually check that (I am not a volt-nut hence everything I write below must be wrong. I am an engineer so hate making mistakes way more than I hate being told about them SO Please point out my mistakes!):

1. The main problem is that it did not have the original certificate that would tell me what the original voltage at 37.5V was. The spec sheet does have a voltage listed there but a generic one:1.017675V @ 37.5C and 64uV / C. That should probably be lower about 2ppm per the two changes to the volt (1927 & 1990) so about 1.017673V?

2. My thermometers are not 0.02C accurate. One (thermistor) was - but is too much affected by ambient temps. I am now using a B class Sensorox PT1000 which isn't that accurate. It measure 37.22C. So 0.28C off the target of 37.5C. That should account for about +17-18uV off...

3. My HP34401A reads (6 digits slow, >10G) 1.017698V. Now I know my HP34401A is off about 10uV (it was just back from calibration). So I am actually about 1.017688V or so. (Note the expected uncertainty from the lab rather crude at 0.0000103...)

Given the uncertainty in day 0 voltage, the temp measurement and the cal lab reference for 1V - I do not think I can validate Frank's numbers. On the positive side - the numbers do work out within the 1ppm / year so I can't disprove it as well.

A question - I'll leave it to rest at 37.5C for a week or so - should I leave it always? Wouldn't the lifetime get shortened at a higher temp?

 

[VintageNut]

I believe that the cell is intended to be left on all of the time. If you turn it on and off frequently, I would suspect that there could be a hysteresis issue.

Dr. Frank would be the person to chime in about the change in the definition of the volt over time and how it relates to your cell.

If your device was manufactured in 1967, then there are definitely changes in the definition of the volt which should be included in the comparison of the datasheet to measurements of the device today.

My Eppley cell was manufactured during May 1975. There is a museum in Connecticut USA that has many Eppley manufacturing documents. I paid USD $5 for a copy of the manufacturing document. My cell was one of a batch of 50 pieces. Sadly there is no calibration information. The cell has been on my equipment rack since July 2015 and is aging/decaying at a very consistent rate during that time. It has some faint writing on paper stickers that help with calculating long term aging/decay.  Since I received the cell, it has decayed about 2.4 uV per month.  I paid something like $20 plus shipping.

Your purchase for $50 was a bargain. The original cost in 1967 was most likely more than a hobbyist could afford.

I have a Keithley 515 resistance bridge built in the 1960s that was $2200 in the 1968 catalog. Google says that is $15,000 in today's money. I paid something like $300 for it .

Some of the very old equipment that was state-of-the-art when manufactured is worth owning. The gear can still perform well. One of my two KVDs was built in the 1960s and it works very well.

Once again, excellent job of bringing the cell to working condition. Cheers!

[HighVoltage]

Thanks for a great thread

I just received a new, never used, Muirhead cell from 1971, that was originally seald.
- Came with certificate
- Came with original box

And it is in perfect specs.

Value 20.1.1971:   1.01864 V
Value 13.11.2016: 1.01861 V

I just need to build a good housing for it.

[Assafl]

I believe that the cell is intended to be left on all of the time. If you turn it on and off frequently, I would suspect that there could be a hysteresis issue.

Dr. Frank would be the person to chime in about the change in the definition of the volt over time and how it relates to your cell.

I do Hope Dr. Frank will chime in. I would hate to cook the cell if it shortens its life (albeit not cooking it means running the compensation calculation).

If your device was manufactured in 1967, then there are definitely changes in the definition of the volt which should be included in the comparison of the datasheet to measurements of the device today.

My Eppley cell was manufactured during May 1975.

Reading Dr. Frank's post on the volt, it would seem to me that you need the 9ppm correction, whereas I just need a 1-2ppm correction(wrong since the Muirhead is probably a British volt - so I'd need a 5ppm drop). Since the volt dropped about 10ppm in 1972 and went back up 9ppm in 1990. Your correction should be by multiplying the voltage by 1.000009264 as per the NIST paper.

There is a museum in Connecticut USA that has many Eppley manufacturing documents. I paid USD $5 for a copy of the manufacturing document. My cell was one of a batch of 50 pieces. Sadly there is no calibration information. The cell has been on my equipment rack since July 2015 and is aging/decaying at a very consistent rate during that time. It has some faint writing on paper stickers that help with calculating long term aging/decay.  Since I received the cell, it has decayed about 2.4 uV per month.  I paid something like $20 plus shipping.

Does something like this exist for Muirhead? I guess I can try emailing them...

Your purchase for $50 was a bargain. The original cost in 1967 was most likely more than a hobbyist could afford.

I have a Keithley 515 resistance bridge built in the 1960s that was $2200 in the 1968 catalog. Google says that is $15,000 in today's money. I paid something like $300 for it .

Some of the very old equipment that was state-of-the-art when manufactured is worth owning. The gear can still perform well. One of my two KVDs was built in the 1960s and it works very well.

Well, the key pull of older equipment (and I think behind the steampunk fad) is that they were more connected to the physics than devices today. Electronics in those days were crude and one couldn't linearize with a lookup table... So it had to use physics. The temp regulator that is calibrated to the heat loss through the Styrofoam using two thermistors would not have been a contemporary design choice. An AVR or a PIC would more than likely have been there.

Once again, excellent job of bringing the cell to working condition. Cheers!

Thank you.

[Assafl]

I am trying to figure out the best way to measure this thing. There are three main uncertainties I can think of:

1. Voltage during manufacture (or anytime since then to yesterday). I sent an email to Muirhead - lets see if they answer.
2. Nil (see below)
3. DMM calibration uncertainty (10.3uV). I really don't know if that can be mitigated somehow...

Edit: Calibrated the temp to negligible impact on the measurement (certainly less than the DMM uncertainty). So 2 no longer relevant.
(2. Temperature. Class B PT1000 has an accuracy of 0.3+0.005t so at 37C about  0.5C - (huge!). Since body thermometers are very (very!) accurate at the 37C mark - perhaps I'll try calibrating the probe myself. )

Anyway - here is the (latest) spreadsheet:

[VintageNut]

For your #3., that is the ultimate question for everybody who has a hobby volt-nut metrology lab.

Is your 10.3 uV the number that was reported by the calibration lab? What instrument did they use to calibrate your DMM? What is the uncertainty of that instrument?

As an example, My K2000 is the best calibrated DMM in my lab for 1V DC and 10V DC. I had this DMM serviced and calibrated at Keithley USA HQ. The report for both the 1V and 10V range is 1 ppm off of perfect after they adjusted and re-tested the K2000. Then I read the last page where it shows that a Fluke 5700A was used to calibrate my K2000. I downloaded the datasheet for the 5700A and the uncertainty of the 5700A is something like 8 ppm (from memory).

So, my wonderful 1 ppm K2000 is maybe 9ppm off of 1V and 10V. No longer so wonderful.

[Assafl]

For your #3., that is the ultimate question for everybody who has a hobby volt-nut metrology lab.

Is your 10.3 uV the number that was reported by the calibration lab? What instrument did they use to calibrate your DMM? What is the uncertainty of that instrument?

As an example, My K2000 is the best calibrated DMM in my lab for 1V DC and 10V DC. I had this DMM serviced and calibrated at Keithley USA HQ. The report for both the 1V and 10V range is 1 ppm off of perfect after they adjusted and re-tested the K2000. Then I read the last page where it shows that a Fluke 5700A was used to calibrate my K2000. I downloaded the datasheet for the 5700A and the uncertainty of the 5700A is something like 8 ppm (from memory).

So, my wonderful 1 ppm K2000 is maybe 9ppm off of 1V and 10V. No longer so wonderful.

They use a Datron 4800. I don't think that you can use the datasheet. The numbers for the uncertainty come from the UKCal labs that calibrates the Fluke or Datron Calibrator. At most you can look at the drift from that point...

See below - the 1V proved to be a specifically hideous range (percentage wise) for the Datron...


These guys have been doing calibrations for quite a bit of the defense industry here after they spun the cal lab from an existing high tech company. So I sort of trust them - especially with all the medical device and aerospace electronics industries here.

The good news is that the HP does not drift. At all.....

[Assafl]

So all this discussion in the measuring the cell thread about tempco served as a reminder that I still had tempco to compensate.

Used an Ice Cream calorimeter as a sort of a thermal oven "cell chest" - lined it with two pad heaters and a freeze pack. Used a CAL controls 3200 PID.

Adjusted the external and internal loops - and hopefully was able to get to within a few uV per the entire range of 15-32C. Will see how that stabilizes in the office over the next few days/weeks.