Author Topic: T.C. measurements on precision resistors  (Read 396316 times)

0 Members and 2 Guests are viewing this topic.

Offline guenthert

  • Frequent Contributor
  • **
  • Posts: 706
  • Country: de
Re: T.C. measurements on precision resistors
« Reply #1025 on: August 24, 2018, 10:58:24 am »
guenthert,

Is it possible to determine the type of metal used in the binding posts? I have wondered about those resistors and what was used to make them. Are the posts silver plated or possibly ferrous?
I know next to nothing about materials.  All I can say is that it has the colour of steel (more or less -- the white balance in the picture above is off) and isn't (strongly) ferromagnetic.
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4543
  • Country: ua
    • xDevs.com
Re: T.C. measurements on precision resistors
« Reply #1026 on: August 24, 2018, 10:58:31 am »
Annual stability 20ppm for 0.01 class tolerance, tempco determined by alpha, beta values, which provided for each standard experimentally :) anything else you want to know? :D
YouTube | Metrology IRC Chat room | Let's share T&M documentation? Upload! No upload limits for firmwares, photos, files.
 
The following users thanked this post: guenthert

Offline MegaVolt

  • Frequent Contributor
  • **
  • Posts: 909
  • Country: by
Re: T.C. measurements on precision resistors
« Reply #1027 on: August 24, 2018, 01:36:18 pm »
Thanks a bunch.  Wished I'd be able to read Russian though.
Ask what is interesting I will translate :) I think it will be easier for me to type in Russian and translate into Google.
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4543
  • Country: ua
    • xDevs.com
Re: T.C. measurements on precision resistors
« Reply #1028 on: September 04, 2018, 12:19:24 pm »
Phew. My initial sorting of 75 recently acquired resistors completed.
Enjoy sorted table with results at attached. I sorted them by tempco.

It took a month of runtime my instruments, to collect the data. :)

« Last Edit: September 04, 2018, 12:23:00 pm by TiN »
YouTube | Metrology IRC Chat room | Let's share T&M documentation? Upload! No upload limits for firmwares, photos, files.
 
The following users thanked this post: Echo88, Dr. Frank, Andreas, e61_phil, doktor pyta, chuckb, razvan784, Magnificent Bastard

Offline babysitter

  • Frequent Contributor
  • **
  • Posts: 893
  • Country: de
  • pushing silicon at work
Re: T.C. measurements on precision resistors
« Reply #1029 on: October 09, 2018, 03:39:31 pm »
Time to enjoy a cool drink. Broken P331 give stylish cocktail containers when you replace the resistor windings in the hermetic part with alcohol-water-mix (as integrated cooling pad) and solder two tin layers with air gap (for isolation) into the bottom.

I'm not a feature, I'm a bug! ARC DG3HDA
 
The following users thanked this post: Dr. Frank, Andreas, Henrik_V

Offline AndreasTopic starter

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: T.C. measurements on precision resistors
« Reply #1030 on: October 09, 2018, 06:01:30 pm »
Hello,

I am missing the thermometer for the "measurement thread"

with best regards

Andreas
 

Offline thermistor-guy

  • Frequent Contributor
  • **
  • Posts: 365
  • Country: au
Re: T.C. measurements on precision resistors
« Reply #1031 on: November 21, 2018, 02:52:54 am »
I have just read the entire thread.

Thanks to Andreas for all the hard work, and to everyone who contributed their comments. There is a wealth of information here.
 
The following users thanked this post: Andreas

Offline dietert1

  • Super Contributor
  • ***
  • Posts: 2018
  • Country: br
    • CADT Homepage
Re: T.C. measurements on precision resistors
« Reply #1032 on: June 25, 2019, 07:27:20 am »
Yesterday i measured some dividers made from Nomca resistor arrays. The arrays give division ratios of
- 13:1 (without external resistor)
- 13.5:1 (with 25K external resistor)
- 14:1 (with 10K external resistor)

I made three 14:1 dividers from 8x 5K Nomca arrays and measured division ratios connecting the dividers to a 10V reference and measuring voltages. Division ratios were 4 ppm,  32 ppm and 60 ppm off nominal value.

In order to determine TC of division ratio i did bridge measurements with one bridge leg in the oven. My 3456A measures difference voltage with 0,1 uV resolution, that is about 0,14 ppm of the 0,714 V divided voltage. So with a temperature change of 25 °C i can determine TC to about 0,01 ppm/°C.

TCs measured were (A) -0,83, (B) 2,40 and (C) -0,57 ppm/°C.

As far as i understand 1 ppm/°C on the division ratio with an industrial part without preselection is a nice result and perfect for a LTZ1000 reference. When i removed the external resistor 10K, (A) improved to -0,49 ppm/°C. The external resistor is a 5 ppm/°C MF, up to 30 ppm/°C different from the 25 ppm/°C of the Nomca array. Its changes get attenuated 21:1.

Next step will be to understand better the result (B). I did not yet observe a certain mixture of the eight resistors.

Regards, Dieter
 
The following users thanked this post: 3roomlab

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 6445
  • Country: hr
Re: T.C. measurements on precision resistors
« Reply #1033 on: June 25, 2019, 08:19:48 am »
Dieter,
take a look here if you didn't already,

https://www.eevblog.com/forum/metrology/statistical-arrays/msg2026729/#msg2026729

Regards, Sinisa
 

Offline dietert1

  • Super Contributor
  • ***
  • Posts: 2018
  • Country: br
    • CADT Homepage
Re: T.C. measurements on precision resistors
« Reply #1034 on: July 08, 2019, 03:05:06 pm »
Recently i was drying 8G16D 10K econistors in a silica drypack and observed rapid and huge resistance changes of about 2 ppm/ % rH within the first 10 or 20 hours. Resistance of the dry resistors was lower by 60 ppm or more starting from 32 % rH. I think this humidity dependence can explain many of the strange deviations in the TC plots above (what was called hysteresis). When heating a resistor to determine its TC it starts drying at the same time, reducing the apparent TC and making the temperature/resistance plot appear more like an ellipsoid than a line when returning back to lower temperature.

A plausible explanation of the humidity effect would be a humidity dependent bobbin size  that puts stress on the resistance wire when it expands. Now i have another question: While a bobbin heats, it will expand, too (independent of humidity). Isn't there an effect of similar size? I am wondering whether a small apparent net TC like 1 ppm or less may be the residual of two much bigger effects canceling each other, one from the wire itself and the other one from the bobbin size. This (near) cancellation could be the reason why TiN measured mixed positive and negative TCs from resistors that were made from wire of similar TC.

Regards, Dieter

PS: Vishay mention this cancellation in their description of Z-Foil resistors. Apparently the foil material has positive TC and the ceramic substrate shrinks with rising temperature, compressing the resistive element. They don't tell numbers though.
http://www.vishaypg.com/docs/63501/how_to.pdf p. 12
« Last Edit: July 08, 2019, 03:18:21 pm by dietert1 »
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14072
  • Country: de
Re: T.C. measurements on precision resistors
« Reply #1035 on: July 08, 2019, 04:06:49 pm »
With the foil resistors the stress effect is relatively simple: the foil is much thinner than the substrate and they are well bond together. Thermal expansion of the substrate is still positive, but smaller than the foil, so on heating it compresses the foil.

With wire would resistors things are more tricky: there are often not well bond together, so that the bobbin can only pull, but hardly push. A polymer bobin can also change with time and react to humidity. The reaction to humidity can be rather slow - so it does not instant humidity but with a delay of maybe days or weeks. So at best there is enough room that the bobin will not cause much stress - this is how the standard resistors a build with the wire free to move. For encapsulated wire wounds it's not just the bobbin, but also the glue / filler that matters.
 

Offline AndreasTopic starter

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: T.C. measurements on precision resistors
« Reply #1036 on: July 08, 2019, 07:46:25 pm »
Apparently the foil material has positive TC and the ceramic substrate shrinks with rising temperature, compressing the resistive element. They don't tell numbers though.

Typical CTE of ceramic substrates is around +7..8 ppm/K
https://global.kyocera.com/prdct/semicon/material/

the metal foil has usually higher CTE (e.g. 13-16 ppm/K) as Kleinstein already mentioned.
So as difference we have a compression of the metal foil with rising temperature.

From the hysteresis side: yes a part of the hysteresis is humidity related.
When temperature cycling usually the hysteresis is larger on the first cycle than on the following cycles.

with best regards

Andreas
 

Offline dietert1

  • Super Contributor
  • ***
  • Posts: 2018
  • Country: br
    • CADT Homepage
Re: T.C. measurements on precision resistors
« Reply #1037 on: July 08, 2019, 08:39:55 pm »
Thanks for the numbers. As far as i understand CTE is the linear expansion, so the compression of about -7.5 ppm/K ( = 13..16 - 7..8 could result in a resistance change of up to -15 ppm/K ("wire" gets shorter and thicker). Which means the foil needs to be made from material with a TC of about +15 ppm/K to get a part with overall TC near zero.
If these numbers are correct they explain a little why even the best foil resistors exhibit TCs in a +/- 2 ppm range.

Can the experts tell whether those Z-foil voltage divider parts are made with one substrate or with two substrates?

Regards, Dieter
 

Offline z01z

  • Regular Contributor
  • *
  • Posts: 151
Re: T.C. measurements on precision resistors
« Reply #1038 on: July 09, 2019, 06:48:05 am »
Can the experts tell whether those Z-foil voltage divider parts are made with one substrate or with two substrates?
I'm no expert but AFAIK these are separate resistors.
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: T.C. measurements on precision resistors
« Reply #1039 on: September 14, 2019, 04:05:04 pm »
Measured Ohmite HS520A10k00 resistor, but with very disappointing results. It's spec'ed to have temperature coefficient of ±3 ppm/°C, -10°C to
+80°C, but I found more like -5.6ppm/K with linear regression, while fit with squared function gives a=-0.025374   b=-3.748446 c=106.824362

-branadic-
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline Martin.M

  • Frequent Contributor
  • **
  • Posts: 953
  • Country: de
  • in Tek we trust
    • vintage Tek collection
Re: T.C. measurements on precision resistors
« Reply #1040 on: September 14, 2019, 05:33:43 pm »
Hello,

I am missing the thermometer for the "measurement thread"

with best regards

Andreas

picture: John Fluke 2190A with option 6 "limits" and K Type Sensor (1m)


835338-0

the Limits Option is to set to <> 23 degrees and the relais terminal to connect to a fat red light " Temperature problem "
A little printer 2030A can be connected to the thermometer to print a tag with date/time stamp and actually temperature. (requires opt. 2 = digital output")

- Martin -
« Last Edit: September 14, 2019, 05:51:03 pm by Martin.M »
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: T.C. measurements on precision resistors
« Reply #1041 on: September 14, 2019, 06:06:34 pm »
I can call an Arroyo Instruments 5305 my own, that I upgraded my incubator with and use a PT100 1/10 class RTD4W. Still I'm not quite happy with the results yet, as I currently run it with P-parameter only. Autotune didn't work for me, I guess the 25l volume of the incubator is to big to find parameters in adequate time. After over 36h it was still searching for PI values, so I aborted it. Calculating parameters with P=1, I=0 and D=0 plus a step response didn't work either.

http://www.aaabbb.de/ControlTheory/ControlLoopTuningStepResponse.php

-branadic-
« Last Edit: September 15, 2019, 10:10:21 am by branadic »
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 
The following users thanked this post: shodan@micron

Offline Edwin G. Pettis

  • Frequent Contributor
  • **
  • Posts: 451
  • Country: us
  • The plural of anecdote is not data.
Re: T.C. measurements on precision resistors
« Reply #1042 on: September 15, 2019, 04:45:10 am »
@Branadic, I presume this is a new resistor from Ohmite, this is basically the old Ultronix resistor line.  I was told that they had 'fixed' all of the problems they had been having with the Ultronix parts (mainly due to the fact that Vishay had given them almost nothing in documentatio).  The 3 PPM/°C TCR spec is a bit tighter than the original.
 

Offline dietert1

  • Super Contributor
  • ***
  • Posts: 2018
  • Country: br
    • CADT Homepage
Re: T.C. measurements on precision resistors
« Reply #1043 on: September 15, 2019, 05:40:55 am »
Arroyo 5305 Autotune worked on our 5305 after i updated firmware to version 1.44 (download from their website). Another info i got from Paul Corr at Arroyo Instruments: Even if you think you have setup proportional mode, there is a small I term. The manual isn't 100% precise on this.

Regards, Dieter
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: T.C. measurements on precision resistors
« Reply #1044 on: September 15, 2019, 07:55:37 am »
The firmware on my unit is 1.44 too, did the update when it arrived. However, it seemed like autotune would take forever, that's why I aborted it.

-branadic-
« Last Edit: September 15, 2019, 11:03:43 am by branadic »
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: T.C. measurements on precision resistors
« Reply #1045 on: September 15, 2019, 07:20:52 pm »
Here is a measurement of a hermetically sealed 10k ± 0.01% AE resistor spec'ed with ±1ppm/°C, linear t.c. is 1.4275ppm/K, while squared fit delivers a=-0.014724 b=2.537869 c=-54.482701

-branadic-

EDIT: corrected labels at diagrams
« Last Edit: September 16, 2019, 04:26:29 pm by branadic »
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 
The following users thanked this post: TiN

Offline dietert1

  • Super Contributor
  • ***
  • Posts: 2018
  • Country: br
    • CADT Homepage
Re: T.C. measurements on precision resistors
« Reply #1046 on: September 15, 2019, 07:45:28 pm »
When looking at the curve i was asking myself, whether Vishay is delivering +/- some ppm at 23 °C. Is this true or result of some adjustments you made before measurement?
Also your measurement is another example how they (ab)use "limit" and "typical" specs to their advantage.
Do these parts come with a date code?

Regards, Dieter
 

Offline AndreasTopic starter

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: T.C. measurements on precision resistors
« Reply #1047 on: September 15, 2019, 07:58:34 pm »
Hello,

the only trick I see is +/-1ppm = 2 ppm according to box method.
So a maximum deviation of 120 ppm over the reduced 0..60 deg C range.

At least the hysteresis seems to be very low.
So you can calculate out the error.

with best regards

Andreas
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Re: T.C. measurements on precision resistors
« Reply #1048 on: September 15, 2019, 08:48:53 pm »
Yes they come with a date code, mine is 1737. The resistors I measure are simply connected with a Keysight 11059A to not introduce any heat for now by soldering cables to it.
This candidate will be used as a reference resistor. But I need to measure its 1K partner, date code 1649 and prepare some aluminium block with a thermistor.

-branadic-
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline FriedLogic

  • Regular Contributor
  • *
  • Posts: 115
  • Country: gb
Re: T.C. measurements on precision resistors
« Reply #1049 on: September 16, 2019, 07:26:07 am »
Here is a measurement of a hermetically sealed 10k ± 0.01% AE resistor spec'ed with ±1ppm/°C, linear t.c. is 1.4275ppm/K,

That seems very high. This is a test of the 1K and 10K Alpha HCZ resistors that I have.
(The centre line is 45°C, and the red trace is just a reference resistor)
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf