Resistor reference for the home lab

[Warhawk]

Hello colleagues,
I am about to get a few precision resistors (vishay s104, 0,01%, 2ppm) from a local gentleman. They are removed from an old equipment and measured using calibrated HP3458A in 4wire configuration (30 minutes to stabilize them) in an air conditioned lab.
I want to use these as a resistor reference for my home lab. Is it worth to think about some oven ? There are a lot of topics regarding the voltage or current reference but not a much about the resistor counterpart.

Thanks 

[Alexei.Polkhanov]

These often come in oil-filled cans like here: http://www.ohm-labs.com/resistance-standards/200-series-oil.html. I assume that for simple convenience value of precision resistor should be a round number like 50ohm or 1.0000000 ohm sharp at, lets say, 25C. You can adjust it if you use resistive wire, but not with resistor. I think it would suck if you have to "punch in" something like 10.00679... ohm on calculator every time when you use it for measurements. When I used precision resistors in the past I always placed them into ice bath because this way I don't have to worry about accuracy of temp measurements as much. It is easier to get precisely 0C using ice bath than precisely 25C in in home conditions. 

[Conrad Hoffman]

It would be nice, but that's not how it usually works. For precision work the actual certified value is used (and punched into the calculator). For lower precision work you can just round to the even value, but then why have high precision resistors? The same is true for many standards. It's valuable to have the long term history so you can see drift, thus readjustment is often not done. "Calibration" may be nothing more than a recording of the current value that's close to the nominal desired value. A practical example is the Fluke 5450A resistance calibrator. The resistors are never adjusted; the exact value is simply stored so it can read out on the display when a particular resistance is selected.

For the OP, unless you have equipment that can resolve the tempco, an oven won't serve much purpose.

[Warhawk]

It would be nice, but that's not how it usually works. For precision work the actual certified value is used (and punched into the calculator). For lower precision work you can just round to the even value, but then why have high precision resistors? The same is true for many standards. It's valuable to have the long term history so you can see drift, thus readjustment is often not done. "Calibration" may be nothing more than a recording of the current value that's close to the nominal desired value. A practical example is the Fluke 5450A resistance calibrator. The resistors are never adjusted; the exact value is simply stored so it can read out on the display when a particular resistance is selected.

For the OP, unless you have equipment that can resolve the tempco, an oven won't serve much purpose.

thanks, that makes sense. I will just probably create only a small PCB with with four golden banana plugs and super simple acrylic case for each resistor.

[Andreas]

I want to use these as a resistor reference for my home lab. Is it worth to think about some oven ?

Hello,

you will only get stable results with constant temperature and constant humidity.
(of course it depends on your precision needs).

I made some T.C. measurements on S102 resistors which are similar:

https://www.eevblog.com/forum/projects/t-c-measurements-on-precision-resistors/msg462303/#msg462303

So you have also watch out to hysteresis effects.


there is also a thread about resistor drift after soldering (even on better Z201 resistors).

https://www.eevblog.com/forum/projects/vishay-bulk-foil-drift-after-soldering/msg445297/#msg445297

With best regards

Andreas

[Warhawk]

I want to use these as a resistor reference for my home lab. Is it worth to think about some oven ?

Hello,

you will only get stable results with constant temperature and constant humidity.
(of course it depends on your precision needs).

I made some T.C. measurements on S102 resistors which are similar:

https://www.eevblog.com/forum/projects/t-c-measurements-on-precision-resistors/msg462303/#msg462303

So you have also watch out to hysteresis effects.


there is also a thread about resistor drift after soldering (even on better Z201 resistors).

https://www.eevblog.com/forum/projects/vishay-bulk-foil-drift-after-soldering/msg445297/#msg445297

With best regards

Andreas

I knew that something excellent will come off my post ! Thanks !

[Dr. Frank]

Hello colleagues,
I am about to get a few precision resistors (vishay s104, 0,01%, 2ppm) from a local gentleman. They are removed from an old equipment and measured using calibrated HP3458A in 4wire configuration (30 minutes to stabilize them) in an air conditioned lab.
I want to use these as a resistor reference for my home lab. Is it worth to think about some oven ? There are a lot of topics regarding the voltage or current reference but not a much about the resistor counterpart.

Thanks 

It all depends on your requirements..

The 3458A may determine the value of your resistors to about 10ppm (0.001%) uncertainty, so the temperature of the resistor during that measurement would also have to be measured exactly.. due its T.C., which may be as high as +/- 4.5ppm/K.. see datasheet: +/-2ppm/K typ. +/- 2.5ppm/K max. spread!!

Intended as a reference of that uncertainty level , you would have to attach properly a precise thermometer to the resistor.

Then your colleague, owning that 3458A, would have to determine the R(T) curve of your S104, so that you could later correct for temperature deviations.

What's the alloy technology used? C-alloy or K-alloy? (i.e. S104C or S104K?)

How to solidly attach the resistor and the thermometer is discussed in Andreas' thread. (iso-thermal mass required)

Depending on the case type of your S104?, i.e. plastic (as shown in the spec: http://www.vishaypg.com/docs/63001/63001.pdf) or metal can, you also have to take care about heat flow, over the case or over the leads.

These metal foil resistors may drift over time if they are NOT hermetically sealed AND oil filled (VHP case). This annual drift may be as high as 20ppm/yr., shelf life, disregarding that these are aged resistor. See longterm ageing curve in spec, also!
Therefore, regular recalibration , i.e. annual monitoring of its value  is required.

Also take care for proper 4W jacks, especially if its value is equal or below 10k.

My construction of such a resistance reference, involving all of the mentioned features, can be seen here:
https://www.eevblog.com/forum/projects/t-c-measurements-on-precision-resistors/30/

I achieve < 1ppm/K, correction to <0.2ppm by measuring temperature, and < 0.5ppm/ 4yr. ageing with VHP202Z.

Your S104? may achieve about 10 fold worse stability figures, which still would be quite good.

Frank

[quarks]

as mentioned before, by Conrad, in calibrators, the reference resistor values are only stored and displayed.
There is no need for a 10.00000 value.
For example, it will be no problem if the value is 0.1% "off", as long as the value is stable over time and has as low as possible changes with environmental influence.