FS (UK/World) 7 decades (10K to 10G) resistor set.

[Alex Nikitin]

For some years now I use a small box with BNC sockets and a number of reasonably accurate resistors and find it very useful in many applications. Recently I've made a somewhat better looking copy, so here it is for sale. All resistors will be measured as accurate as I can do it (from <20ppm for 10K to <0.5% for 10G). All BNC sockets are PTFE isolated and the resistor chain leakage to the case is over 1000 Tohm.

Resistors used are as follows:

10K Vishay Bulk Metal Foil 0.1%, 2ppm/C

100K  Vishay Bulk Metal Foil 0.1%, 2ppm/C

1M0 RC55Y 0.1% 15ppm/C

10M Caddock USF340 0.1% 5ppm/C 

100M Dale RNX 3/8 1% 100ppm/C 100M Ohmite Slim-MOX 1% 25ppm/C

1G  Dale RNX 3/8 1% 100ppm/C 1G Ohmite Slim-MOX 1% 50ppm/C

10G (2x5G) Ohmite 102E 1% 100ppm/C

£150 (UK Pounds) + postage (£5 in the UK, £15 abroad)

Please PM me for further details.

Cheers

Alex

P.S. the shorting plug is not included now included in the kit, a 50 Ohm termination plug can also be used in its place (adding 50 Ohm to the chain).

P.P.S. The maximum voltage applied to the resistors is 10V for 10K, 30V for 100K, 100V for 1M0, 300V for 10M, 1000V for larger values, the maximum voltage between the case and the resistors chain is 1000V (limited by the BNC connectors) . 

[Alex Nikitin]

The box as pictured is sold to a forum member (that was quick). I will keep the thread open as I will build more units, so please state your interest here or by PM.

Cheers

Alex

[krish2487]

Hello Alex,
I m interested if you are willing to take the trouble to make one more. :-)
It looks cool and a solid bit of kit btw!

[Hydron]

PM sent expressing interest, but another quick question - do you have the kit to accurately measure the higher value resistors before shipping? Would be especially useful for the >10M ones, as this is tricky to do with a standard bench multimeter.

[Alex Nikitin]

PM sent expressing interest, but another quick question - do you have the kit to accurately measure the higher value resistors before shipping? Would be especially useful for the >10M ones, as this is tricky to do with a standard bench multimeter.

https://www.eevblog.com/forum/metrology/precise-measurement-of-high-value-resistors/msg5572707/#msg5572707

Cheers

Alex

P.S. - I am making 3 more units, will PM those who ordered, when ready.

[Alex Nikitin]

The first unit is packed and ready to go.

Measured resistances are:

9.9996K
99.998K
1.00012M
10.0012M
100.6M
1.005G
10.035G

Cheers

Alex

[Alex Nikitin]

Two more units are sold, I have one more at £100 + postage price, after that I can make them on order at £150 + postage to cover the new resistors cost. The resistors used would be as on the photo attached, with 25ppm/C Ohmite Slim-MOX for 100M and 1G values.

Cheers

Alex

[Alex Nikitin]

The last unit for £100 is provisionally sold to a forum member, from now on the price will be £150 (still essentially a cost price for the box, resistors and connectors) + postage, please PM me for ordering.

Cheers

Alex

[ause]

Hi Alex,
Apologies for the newbie question but why do you use the BNC jacks vs using Banana jacks, is that for a specific purpose, RF or calibrating?
Cheers.

[Alex Nikitin]

Hi Alex,
Apologies for the newbie question but why do you use the BNC jacks vs using Banana jacks, is that for a specific purpose, RF or calibrating?
Cheers.

Good quality BNC connectors (PTFE isolation, Beryllium Bronze central contact, machined body, not the cheap crap you can buy on eBay etc and not the cheaper versions from decent makes but with Nylon isolation) are probably the only affordable option suitable for this application with the high value resistors up to 10G. Most if not all 4mm binding posts are far too leaky to use at Giga-ohm range and do not provide sufficient screening. The nice bonus is that BNC connectors double as high quality mounting supports for the resistors. The solid Al alloy case levels the temperature gradients, reducing possible thermal offsets to essentially insignificant levels for most measurements. At the bottom end the added contact resistance of a quality BNC connector is sufficiently small (milliohms) to provide a good accuracy at 10K (10mohm = 1ppm from 10K).

Also this box is primarily a reference set, it has many other uses, including a high voltage divider, a set of external precision resistors for a picoammeter, both as a feedback and in series with the input. Just as an example: using 10G+100M at the input and 10M+1M in the feedback loop you convert a picoammeter into an accurate HV buffer for a voltmeter with 1000V input range, 1V output, and 11G input resistance. And for this kind of arrangement BNCs are the best option.

Cheers

Alex

[ause]

Nice one thanks Alex..Now that I know all that I should've got in earlier, I'll have to save up a bit more now..

Thanks again for your reply..

[unseenninja]

The first one of these boxes is now in my hands. Thanks to the thorough packaging, it survived the trip from the UK to Sweden unscathed.

As far as I can tell (with an 8.5 digit Datron 1271 meter), the resistances are unchanged by the journey. Shipping in winter might be a different story. My measurements match up well with the original measurements by Alex.


Value   My Measurement  Original Measurement
10K     9.999,557,8KΩ   9.9996KΩ
100K    99.998,389KΩ    99.998KΩ
1M      1.000,127,42MΩ  1.00012MΩ
10M     10.001,39MΩ     10.0012MΩ
100M    100.575,9MΩ     100.6MΩ
1G      1.003,812GΩ     1.005GΩ

The agreement between my measurements and the original measurements diverges as the resistance rises. This is caused by a problem which was present in my 1271 when it was last adjusted and calibrated. It was great to see Alex's accurate calibration expose what I already suspected.
I can't, at the moment, give a measurement for the 10GΩ resistor - my Datron doesn't go that far and my Keithley 617 needs adjustment and recalibration.

If you need a well constructed and well thought out resistance decade that goes from 10K to 10GΩ and comes with an accurate calibration, then I can recommend Alex's resistor set without hesitation.

[Alex Nikitin]

The new look for the "7D" - hopefully an improvement  , as I keep receiving requests to build more. The label has an acrylic lamination so should survive even an IPA wipe.

Cheers

Alex

[ause]

 

[EC8010]

The pedant in me has to point out that kilo should be lower case "k". Upper case "K" is kelvin. Other than that, very nice.

[Alex Nikitin]

The pedant in me has to point out that kilo should be lower case "k". Upper case "K" is kelvin. Other than that, very nice.

I guess it shows my age! When I’ve learned the craft, a capital K was used happily for kOhm. Let’s just hope it won’t mislead anyone  .

Cheers

Alex

P.S. - two more units are sold to forum members.

[krish2487]

Received  a very well packed and robust kit from Alex.

Thank you very much Alex!

I m no voltnut but a preliminary measurement with a 34460a shows that the measurements are off slightly on some higher ranges (10M and above) and more or less as specified
on the back of the case for the lower ranges ( some noise in the least significant digit).
But,
1. I have to subtract 50 ohms from the measurement.
2. cables were long and added significant error and are a mess
3. the box was cold to touch due to transit, I ll check again tomorrow after the box has settled in this environment.
4. I need to rig up a 4W measurement and short cables to make a more accurate measurement.
5. the 34460a only goes upto 100M

[Alex Nikitin]

Thank you!

1. I have to subtract 50 ohms from the measurement.

No, the plug you've received is a true short (less than 0.01 Ohm), not a 50 Ohm termination.

Cheers

Alex

[krish2487]

Great!! :-)
I ll keep that in mind..

the BNC connectors look and feel top notch and the sticker with the measured values is priceless! :-D
Thanks again Alex!

[krish2487]

Okay.. so I got some of the stuff over the weekend and gave another spin.
The measurement setup is as follows ->
2 multimeters - keysight 34460a and fluke 8845, both set to 100 plc and 6.5 digits. readings were taken after allowing the meter to settle for a minute.
I used a off the shelf BNC to dual banana plug and connected 4 cables of 50cm length in 4W measurement mode. The cables had stackable banana plugs on either ends.

and these is what I measured.

Marked                            Measured
9.9997K                           9.99653K
99.994K                           99.7283K
1.00038M                        1.00008M

and for some reason any measurement above that simpy reads around 5M.. no matter which resistance I try to measure..
I m almost certain the cable length is still a huge factor here and the BNC to banana adapter as well..

I tested it on both the multimeters and granted, I do not know anything about the calibration on either of them but they seem to agree with each other to a few (tens) of counts.

@Alex Nikitin any suggestions on what I can try to measure some of the higher ranges ??
I m reasonably happy with what I m seeing on the meters now, I m less worried about the absolute readings and more about the meters agreeing with each other, which they do. :-)
I just want to validate it on the 10M and 100M ranges and I will be relieved. :-D

[Alex Nikitin]

Something is not right, perhaps you are connecting the earth side in more than one point. Could you post a photo of your setup? Try to use a simple 2W ohms first, it will add about 0.5 Ohm or less to the value so the error will go progressively less with the increase in resistor value. 

Cheers

Alex

P.S. - for the right way of connecting to 4W Ohm, look at the post #5 above.

[krish2487]

Thanks for the input Alex.
I redid the connections and tried again. with 2W and 4W.. and I tested it on the 10M resistor and I m still seeing about 3.5M 

[Alex Nikitin]

What the meter shows if you unplug the BNC?

Cheers

Alex

[krish2487]

ohhhhkay...
It  shows the same reading as when it was connected.. 3.5-3.6M ohm...

[Alex Nikitin]

I'm not sure what else to try. What happens if you ground the case? If you have a BNC socket to 4mm banana plugs, you could try it with a screened BNC-BNC cable. It is not that much could go wrong with the box .  Open it and have a look, perhaps something has wriggled its way in during the transportation... .

Cheers

Alex