Kelvin-Varley voltage divider Dekapot DP1211, teardown and what to do with it ?

[DC1MC]

Hello, I've got a K-V Dekapot voltage divider because I thought that it's a resistor decade , seem to be operational for the little testing I could do, just some very faint blueish  oxidation on the switch contactas, that I think it will be easy removed by my VARENA contact spray.
So it seem to be operational, and I've snapped some pictures for your enjoyment.

the datasheet and USER/Service Manual that I was able to found are there:

https://www.ietlabs.com/pdf/Datasheets/TegamDP1211-1311.pdf

https://www.ietlabs.com/pdf/Manuals/DP1211_1311_im.pdf

Looking at the schematic of the device, I struggle to find a use for it, could they still be used for something nowadays or it's some kind of fossile from bygone times ?

If even the User Manual has this quote at the very beginning:

Before you reject this unit
This is not a calibrated resistor. It is a voltage divider.
It has an equivalent circuit which includes a variable resistor in series with the output lead. This does not impair its accuracy as a voltage divider but makes it difficult  to  find  its  linearity  by  resistance  measurements.

So it's not that obvious for the intended users as well, if they have to print these warnings  .

I wanted to dremel the resistors out and use them for something, but then I've found this serious warning a bit down on the page:

Important Note on Surplus Units
Surplus units cannot be relied upon since the preci-
sion resistors used have a shelf life and drift over time.
As such, surplus units cannot be put into service.

So much for what I've heard here on the forum, that the resistors drift stabilizes with time, it seems that those guys managed to produce some self-destroying resistors  !!!

So now what I should do with it, I'm pretty sure nobody will want it with those expired resistors and it's not worth the cost of shipping, so it's ANYTHING that could be done to it or should I throw it directly in the garbage bin ?!?

Please try to suggest me some useful usage for it, I'll wait a bit of before scrapping it, out of respect for the forum member that helped me with the shipping (thanks again Ulf).

 Cheers,
 DC1MC

 

  Right now I'm going for lunch and maybe take some of these_{anti-trolling} pills, my condition it's getting worse at the year's end.

[tggzzz]

It is an accurate voltage divider, provided the output load draws zero current.

Personally I wouldn't touch the contacts.

Do not scrap it; they are rare and have value: see fleabay. Alternatively, send it to me.

Typical uses: to measure the linearity of voltmeters, or to make voltmeters that are accurate to <<0.1% with a meter is only accurate to 2%. Or to just play with the pornographically tactile rotary switches Key search term: "kelvin-varley divider".

(And please remove your "size" tags!)

[JS]

I always wanted a kvd. It's useful to generate very precise voltages starting from a fixed, stable voltage source.

They are great for many things, mostly metrology related as the precision is usually a few orders of magnitudegreater than needed for every day use. Checking linearity of ADCs, zeroing comparators and bridges, generating reference voltage to make transfers and calibrations are some of the uses for them.

Whatever you do don't turn it into a decade resistor, find someone who wants one and he will make it happen, making a decade resistor or getting one for you.

JS

JS

[tggzzz]

I always wanted a kvd. It's useful to generate very precise voltages starting from a fixed, stable voltage source.

They are great for many things, mostly metrology related as the precision is usually a few orders of magnitudegreater than needed for every day use.

Yup. I have a 7 decade KVD like this https://www.ebay.com/itm/Julie-Research-Lab-Vintage-Rare-JRL-VDR-106-7-Voltage-Divider-NICE-Mod-A-KVD/302999427247 but I didn't pay anything near that for it

Checking linearity of ADCs, zeroing comparators and bridges, generating reference voltage to make transfers and calibrations are some of the uses for them.

See image of an old 6 digit "DVM" with a 2% (i.e. <2 digit) meter

Whatever you do don't turn it into a decade resistor, find someone who wants one and he will make it happen, making a decade resistor or getting one for you.

Just so!

[Conrad Hoffman]

Perfectly good unit and probably still crazy accurate. Don't touch the contacts or clean anything unless it isn't actually working well. Put the can back on and find a use for it!

[capt bullshot]

Are you crazy, scavenging the resistors from a good KVD?

If you don't have a use for it, send it to me.

Edit: I'd try to repair your MG3633A sig gen for it.

Edit 2: You're just trolling, aren't you?

[DC1MC]

I always wanted a kvd. It's useful to generate very precise voltages starting from a fixed, stable voltage source.

They are great for many things, mostly metrology related as the precision is usually a few orders of magnitude greater than needed for every day use. Checking linearity of ADCs, zeroing comparators and bridges, generating reference voltage to make transfers and calibrations are some of the uses for them.

Whatever you do don't turn it into a decade resistor, find someone who wants one and he will make it happen, making a decade resistor or getting one for you.

JS

Well, it has an '84 production date, so the resistors are most likely drifted away and useless as they say, I've feed it up with 10.00006V, set it it to 9.9 and fiddled with the potentiometer to make 9.90000, managed to go to 9,00004 (+/- 1), so far so good, then set it up to 5.0V and guess what: 4,99996V, well so much for linearity, I guess the resistors are truly expired and being a cheap device there is no way to calibrate it and must be discarded, I guess only Dave manages to get those impeccable 8 digits resistive decades that are in parameters after 30yrs and also looks good on the outside  .
At least it seems that it has some short term stability, i was depressed that 4,40000 showed like 4,40016, set it to 4,40000 with the potentiometer and now it stays like this (+/-1digit) for something like 20 minutes on my 3456A, so probably it could be used for transfer and calibration of 5,5 digit instruments or something.

Who will give a reasonable useful resistor decade for it, nobody I tell you, and shipping it to Argentina will probably cost like one with 3 digits, that is in the validity period, bought from the US.

 Well, this is it, at least it didn't cost too much (for Germany) and it's reasonably short-term stable, so yeah, nice to have, even with the expired resistors  .

  Cheers,
  DC1Mc

[tggzzz]

Well, it has an '84 production date, so the resistors are most likely drifted away and useless as they say,

Er, no. KVDs are relatively subtle, and bits of wire don't drift much.

I've feed it up with 10.00006V, set it it to 9.9 and fiddled with the potentiometer to make 9.90000, managed to go to 9,00004 (+/- 1), so far so good, then set it up to 5.0V and guess what: 4,99996V, well so much for linearity, ...

Measured it with what? What's the measuring device's input resistance and specification?
What's your voltage source's stability specification?
What's the KVD's specification?

If you are using an HP3456A, then at best the accuracy measuring 9V would be 9*23e-6 +- 20uV (i.e. +-227uV) and 5V would be 5*23e-6 +- 20uV (i.e. +-135uV). Your 5V measurement is within 40uV, so I really don't see any problem with the KVD.

I'll bet that the KVD is OK, provided you haven't put too much current through it, nor fouled up the switch contacts.

[DC1MC]

Are you crazy, scavenging the resistors from a good KVD?

If you don't have a use for it, send it to me.

Edit: I'd try to repair your MG3633A sig gen for it.

Edit 2: You're just trolling, aren't you?

Hi Capt, why would I be crazy to dismantel it, it's a two digits divider, some while ago was a guy killing a Fluke 760 calibrator, "because he can't repair it..." and snipping the resistors, he even did an Youtube video where he was happy that he salvage a neon lamp  , I think the post and video are still up.
And why would I troll it, I've attached a picture of the actual reading, sorry for the potato quality, but it really become dark here and I didn't have the blitz on, I swear to God that I didn't touch it since I've set it to 4.40000 (with potentiometer adjustment, as I've said earlier), I've said that it seem stable.
Anyway, if you want to have it AND if you pick me up from Mannheim and drive me back your train station (I'll contribute to gas and the Anritsu it's 32Kg !!!), I have the last WE from January and the next two after free, so we can arrange something and no dremeling will be applied to the divider  .

 Cheers,
 DC1MC

[DC1MC]

Well, it has an '84 production date, so the resistors are most likely drifted away and useless as they say,

Er, no. KVDs are relatively subtle, and bits of wire don't drift much.

I've feed it up with 10.00006V, set it it to 9.9 and fiddled with the potentiometer to make 9.90000, managed to go to 9,00004 (+/- 1), so far so good, then set it up to 5.0V and guess what: 4,99996V, well so much for linearity, ...

Measured it with what? What's the measuring device's input resistance and specification?
What's your voltage source's stability specification?
What's the KVD's specification?

I'll bet that the KVD is OK, provided you haven't put too much current through it, nor fouled up the switch contacts.

Well, I didn't say that it's expired, the bloody manufacturer has this warning on the first actual page of their manual, but it actually make sense, with no way calibrate it, there will be some drift in different directions of the resistors and then it will finally pass out of specification, of course for hobby use will be enough or to test/calibrate 4..5 digits voltmeters in the best case. It's not cheese to become toxic when spoiled  .

The voltage source it's an AD584JH based device, that is not the most long term stable thing in the world, even if I keep it running for something like a year, but it's not so unstable that is drifts in a minute.

The measuring instrument it's the weak point, just a humble 3456A, but this is what I have at the hand right now and it's pretty stable.

I do take offense at the "fouling" of the contacts, this is a bloody expensive chemical for contacts, non-corrosive, no-leftovers, and I've did (unmentioned) tests before and after, it was much worse when it was uncleaned, anyway, the lowest reproducible value of the potentiometer was initially 2,4ohm and after cleaning 1.6ohm.

I've feed it with 10V, it should be OK according to the data sheet and the resistors didn't look overloaded at all when I've open it up, clean it and IMMEDIATELY close it.

Even now, it stays in between 4,39999 and 4,00001V and this makes me happy, it does seem to have some stability, at least as seen by my humble instrument, from a 3458A POV it may be another story, so if the Cap doesn't want it I will keep it around for a while, thanks to all that contributed to the thread.

Stupid/Naive idea: could it be used together with an OP in the feedback loop or somehow otherwise than the OP as a repeater, in a way that will preserve the stability ?

 Cheers,
 DC1MC

[tggzzz]

Well, it has an '84 production date, so the resistors are most likely drifted away and useless as they say,

Er, no. KVDs are relatively subtle, and bits of wire don't drift much.

I've feed it up with 10.00006V, set it it to 9.9 and fiddled with the potentiometer to make 9.90000, managed to go to 9,00004 (+/- 1), so far so good, then set it up to 5.0V and guess what: 4,99996V, well so much for linearity, ...

Measured it with what? What's the measuring device's input resistance and specification?
What's your voltage source's stability specification?
What's the KVD's specification?

I'll bet that the KVD is OK, provided you haven't put too much current through it, nor fouled up the switch contacts.

Well, I didn't say that it's expired, the bloody manufacturer has this warning on the first actual page of their manual,

Are you blaming the (highly respected) manufacturer because the DekaPot isn't a DekaStat?

but it actually make sense, with no way calibrate it, there will be some drift in different directions of the resistors and then it will finally pass out of specification, of course for hobby use will be enough or to test/calibrate 4..5 digits voltmeters in the best case. It's not cheese to become toxic when spoiled  .

The voltage source it's an AD584JH based device, that is not the most long term stable thing in the world, even if I keep it running for something like a year, but it's not so unstable that is drifts in a minute.

Popcorn noise can be a problem. I have a voltage reference that sometimes jumps 15uV (1.5ppm).

The measuring instrument it's the weak point, just a humble 3456A, but this is what I have at the hand right now and it's pretty stable.

I modified my message to include the expected readings from a 3456A; you may not have seen that. For simplicity, here they are again. "If you are using an HP3456A, then at best the accuracy measuring 9V would be 9*23e-6 +- 20uV (i.e. +-227uV) and 5V would be 5*23e-6 +- 20uV (i.e. +-135uV). Your 5V measurement is within 40uV, so I really don't see any problem with the KVD."

I didn't bother to include the expected precision of the KVD itself, but it is a 2-decade KVD plus 2 decade potentiometer having a linearity of +-20ppm and resolution of +-30ppm. Those values can be added to the above numbers.

I do take offense at the "fouling" of the contacts, this is a bloody expensive chemical for contacts, non-corrosive, no-leftovers, and I've did (unmentioned) tests before and after, it was much worse when it was uncleaned, anyway, the lowest reproducible value of the potentiometer was initially 2,4ohm and after cleaning 1.6ohm.

Shrug. I've no idea what the chemicals were, nor their effects on the precision contacts, e.g. on any Seebeck voltages. In my limited experience, with such precision contacts it is sufficient to rotate them multiple times to remove oxide. A more definitive answer would be appreciated.

Not all restorations add value, e.g. https://en.wikipedia.org/wiki/Ecce_Homo_(Mart%C3%ADnez_and_Gim%C3%A9nez,_Borja)

Beyond that, I suspect the resistance values you mention aren't especially important for a KVD.

Stupid/Naive idea: could it be used together with an OP in the feedback loop or somehow otherwise than the OP as a repeater, in a way that will preserve the stability ?

Don't put too much current through it, have an infinite resistance (i.e. >>10Mohm!) load on its output, don't forget the frequency dependence of the inductive resistors, slug any feedback loop so that long leads don't cause the opamp to oscillate, shield to keep noise/interference from getting into your circuit.

[vindoline]

You could build this nifty precision voltage source: https://youtu.be/9KtJr7Jv16I

I agree with the others that it's highly unlikely that the precision resistors have drifted appreciably.

[texaspyro]

Maybe IET wants to scare people into buying really expensive new units instead of cheap, perfectly good, surplus units?     I paid $20-$50 for my DekaPots.

[tggzzz]

Maybe IET wants to scare people into buying really expensive new units instead of cheap, perfectly good, surplus units?     I paid $20-$50 for my DekaPots.

For real metrology applications, they won't need to scare people.

Metrology is either intrinsically scary (i.e. at the boundaries of the possible), or is about peace of mind (legally valid measurements).

For people playing in their spare room, there's no need to be scared - merely grateful that high quality kit can be easily found by the amateur.

[DC1MC]

Maybe IET wants to scare people into buying really expensive new units instead of cheap, perfectly good, surplus units?     I paid $20-$50 for my DekaPots.

Well, if it wasn't clear on the first post, I did a little bit of end of year trolling, more like joking< .

Of course I LOVE this thing, it's rock stable and I've spend a good part of an hour yesterday playing with funny voltage values, like 1,23456 and 2,32425 and of course 6,66666  and seeing them staying put.
( but being a man of my word, if Capt Bulshot follows the offer to repair my Anritsu it's going to him).

The video above, with the hipster AD guy making a lame buffer out of a FIVE DECADE !!! coaxial divider and some OP was a confirmation that it's feasible to make a nice poor-man calibrator out of such a thing.
For relatively rich poor man of course, because such a divider it's over 1,8K, IF you're able to find it. In any case, if I ever make such a precision voltage source I'll either use better buffers or don't make a lamest attempt to cover the voltmeter display with my hand when it miserably fails the load test   .

Now that I've got the taste of it (two decades it's nice and handy, but more is better  ) I was prowling the fleabay to see what it's available, in between SF and Fantasy prices  , I did see one offer from a Canadian seller that seem reasonable:

https://www.ebay.de/itm/ELECTRO-SCIENTIFIC-ESI-DP311-DEKAPOT-VOLTAGE-DIVIDER-MODULE-1000-OHMS-710V-RMS/332607664425
https://www.ebay.de/itm/FLUKE-60A-C-5003-DECADE-POTENTIOMETER-0-to-1K-RESISTANCE-MAX-VOLTAGE-70-DEKAPOT/23295083196

I wasn't able to find any 4 decade coaxial ones, much less 5 decade, most likely that was some one of a kind isolated model and it's unobtanium now (but please prove me wrong, and I'm talking about the proper coaxial ones, not their lamer modern variant with these tumbler switches).

 So the Argentinian forum member (@JS) who wanted to have one, can get it if he wants, and I was wondering if we, the DE users, should we organize a group buy and release the 3 Canadian ones, 23EUR shipping it's much better than 67EUR shipping , if interested, please let me know.

 Cheers,
 DC1MC
   

[tggzzz]

Maybe IET wants to scare people into buying really expensive new units instead of cheap, perfectly good, surplus units?     I paid $20-$50 for my DekaPots.

Well, if it wasn't clear on the first post, I did a little bit of end of year trolling, more like joking< .

Noted.

Don't expected me to spend my time trying to help you in the future.

[DC1MC]

Noted.

Don't expected me to spend my time trying to help you in the future.

Now, now, let's not start the year in this mood, should we ?!?

First, for you or anybody else that felt "trolled", I sincerely apologize, the trolling was just emulating the typical clueless user that gets a piece of cool gear and hacks it away to recover the most useless part of it, it did left an impression to me the guy hacking to death the poor Fluke calibrator and proudly showing that he recovered this cool neon lamp.

I indeed didn't know too much about K-H dividers, mostly because they were outside of my price reach, and I appreciate EACH and EVERY meaningful contribution made in this thread and others, the metrological masters here have my undivided respect, if I can make this little pun.
Seriously speaking, I've learned so much in the time I've spent here and I want to give thanks to everybody that took some of this precious time to help and educate an (almost)newbie like me.
That doesn't mean that, now and then, there will be no tongue'n cheek (or whatever you spell this) post, but as always, it will be clearly marked  .

 Cheers,
 DC1MC
 

[tggzzz]

I don't like people wasting my little remaining life.

I like helping people; trolling prevents me helping people that might benefit from my help.

[DC1MC]

As I've said, ALL my questions regarding ALL of my topics on this forum were of genuine interest in the thematic or help/requests for help, and not to steer controversy or useless discussions, I'll leave this to the "It's the moon landing real" people.
Now and then I'm using a joking formatting that should be easy to detect, but I guess this is why nowadays you have sarcasm tags 

 In any case, thanks for all the fish, they were tasty 

 Cheers,
 DC1MC

[tggzzz]

As I've said, ALL my questions regarding ALL of my topics on this forum were of genuine interest in the thematic or help/requests for help, and not to steer controversy or useless discussions, I'll leave this to the "It's the moon landing real" people.
Now and then I'm using a joking formatting that should be easy to detect, but I guess this is why nowadays you have sarcasm tags 

You admitted you were trolling which, by definition, is not genuine. One person suspected that, but everybody else treated your statements as being genuine.

Emoticons are used precisely because words alone often don't convey sufficient meaning, especially across language and cultural divides.

[capt bullshot]

Well, if it wasn't clear on the first post, I did a little bit of end of year trolling, more like joking< .

Of course I LOVE this thing, it's rock stable and I've spend a good part of an hour yesterday playing with funny voltage values, like 1,23456 and 2,32425 and of course 6,66666  and seeing them staying put.
( but being a man of my word, if Capt Bulshot follows the offer to repair my Anritsu it's going to him).

Yes, I'm looking for a KVD (any style, not necessarily a DekaPot one). But I believe you've also a good use for this one, so I won't insist. Admitted, the trolling worked for me for some amount of time, but then I didn't believe anymore you really wanted to get rid of it. Nice year's ending fun anyway. Nevertheless, if your Anritsu should find somehow and somewhen a way to me, I'd still try to repair it or help you with the repair.

BTW
Happy New Year

[DC1MC]

....

Yes, I'm looking for a KVD (any style, not necessarily a DekaPot one). But I believe you've also a good use for this one, so I won't insist. Admitted, the trolling worked for me for some amount of time, but then I didn't believe anymore you really wanted to get rid of it. Nice year's ending fun anyway. Nevertheless, if your Anritsu should find somehow and somewhen a way to me, I'd still try to repair it or help you with the repair.

BTW
Happy New Year

Now that's more like I want to hear  , I'll try to organize some travel stuff in the time frame I've mentioned, speaking of KVDs, besides the Canadian ones that I've mentioned, on our side of the pond I've only found this Danish used equipment seller that sells these two device, one KVD and one resistive decade, without asking for a large slice of a kindney:

KVD: https://www.ebay.de/itm/Fluke-60A-C-5001-Decade-Potentiometer-10kohm/112255619421
Decade: https://www.ebay.de/itm/ESI-DS1464-Coaxial-Decade-Resistor/112255615307

I would have happily got them both, kept the decade and surprised you  , but this thing mentioned about both made me hesitant:

"Resistors are OK, but the contacts needs some attention - but still in good condition!"

And as R. Kipling said: "When the jackal says it's gray, how black it must be...", so coming from a profi seller it could be that the previous owners used them to divide 380V on highly-inductive loads or maybe just some patina, one never knows, and I've seen that the opinion of the people in know is to never touch the contacts with anything, so I'm really wondering what to do.

 What do you say, these or the Canadians ?

  Cheers,
  DC1MC

[capt bullshot]

I'd prefer an EU source, faster delivery and less hassle with customs and whatever. The Canadian ones don't look better in terms of visual shape anyway.

At least the Danish seller tells you about the contacts, that may have two causes IMO - not being used at all, or heavily used. I'm not in a hurry with the KVD, it's on my long term wish list, and at some time something interesting may pop up (e.g. like that ET Systems AC source and the ZES power meter http://wunderkis.de/AC-Source-EAC/pic/meas1.jpg ) at a low price tag. So for the unknown condition, the price tag isn't low enough for these to me. Alas, these things are quite rare in Europe.