Fluke SL935 resistance standard prototype

Contents

Intro

Following up with metrology-level test equipment collection, I got unique chance to acquire something rare, Fluke SL935 prototype module. This unit has serial number 001, and there are high chances it is only one in the world. There was no information or mention anywhere in Fluke documentation or Internet, so this is our exclusive chance to play with this special box.

Disclaimer

Redistribution and use of this article or any images or files referenced in it, in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

  • Redistributions of article must retain the above copyright notice, this list of conditions, link to this page (https://xdevs.com/article/fsl935/) and the following disclaimer.
  • Redistributions of files in binary form must reproduce the above copyright notice, this list of conditions, link to this page (https://xdevs.com/article/fsl935/), and the following disclaimer in the documentation and/or other materials provided with the distribution, for example Readme file.

All information posted here is hosted just for education purposes and provided AS IS. In no event shall the author, xDevs.com site, or Fluke or any other 3rd party be liable for any special, direct, indirect, or consequential damages or any damages whatsoever resulting from loss of use, data or profits, whether in an action of contract, negligence or other tortuous action, arising out of or in connection with the use or performance of information published here.

If you willing to contribute or have interesting documentation to share regarding pressure measurements or metrology and electronics in general, you can do so by following these simple instructions.

Fluke SL935 idea

SL935 is 1 Ω and 10 KΩ Resistance Standard prototyped in Fluke 732B chassis by some Fluke calibration facility. Reason to think so is use of no less than 14 precision custom thin-film networks from Fluke 5700A calibrator. Most of them have same manufacturing date and week (beginning of 1993), so these came from calibrator service facility or Fluke’s spare parts inventory. It would be difficult for any hobby-level enthusiast to get this amount of precision custom components, due to cost of even broken and destroyed 57XX series calibrators. Not mentioning requirement to have expensive Fluke 732B available to use as donor for parts.

We can confirm it’s based on Fluke 732B chassis, as same oven and front panel are used. Prototype with S/N 001 has nameplate label sticked on top of 732B panel to cover original DC standard silkscreen.

Manuals and comparison to production standards

There are no manuals for Fluke SL935 prototype available, so here are other products relevant information which can be helpful.

Comparison for known specifications of production resistance standards:

Specification Fluke SL935 Fluke 742A-1 Fluke 742A-10K ESI/IET SR104 IET SRL-1
Output resistance 1 Ω and 10 KΩ 1 Ω 10 KΩ 10 KΩ 1 Ω
Accuracy (1y) ?, ? ±2 ppm ±2 ppm ±1 ppm ±2 ppm
Temperature coefficient, ppm <1 ppm?, <1 ppm? ±3 ppm ±1.5 ppm <±0.1 ppm/°C ±3 ppm max (18-28°C)
Long term 1 year stability, ppm ? , ? ±8 ppm ±4 ppm ±0.5 ppm ±2 ppm
Max load current 100mA, 500µA 200 mA 600 µA 10 mA 418 mA
Construction type Hermetic Fluke film networks Precision wire-wound Wirewound resistor in oil tank Precision wire-wound
Active thermal compensation Yes, +35 °C oven assembly None Thermistor in oil tank None
Temperature sensor Yes, 100 KΩ thermistor No No Yes, 10 KΩ thermistor No
Power requirements AC mains or +12VDC 300mA Not required, passive device
Backup/offline power supply Internal 7.0Ah +12V VRLA Not required, passive device
Dimensions, weight 86 × 105 × 127 mm, 910g 254 × 206 × 311 mm, 4.8kg 86 × 105 × 127 mm, 730g
MSRP N/A, Prototype $1700 USD $1700 USD $7147 USD $1974 USD

Table 1. Hi-end resistance standards, brief specification list

Design and construction


Image 1-2: Fluke SL935 oven module assembly, as received.

Restoration


Image 3-4: Oven module construction and wiring.


Image 5: Resistance can removed


Image 6: 10 KΩ resistance hermetic network array from Fluke P/M 798330.

10 KΩ derived from 10 x Fluke 5700A hermetic resistor network Z1 Fluke 798330 from A10 Ω MAIN PCBA and 2.1111 Ω trim (9 × 19Ω).


Image 7: 1 Ω resistance hermetic network array from Fluke P/N 815889.

1 Ω derived from 4 x Fluke 5700A hermetic resistor network Z5 Fluke 815889 from A9 Ω CAL PCBA.


Image 8: Board assembly with silver-coated wires and perf-board.


Image 9-10: Ceramic heater substrate resistors on each side of the oven block.


Image 11: Modified oven 732B control PCBA with TL062C opamp


Image 12-14: Another view on resistor networks


Image 15-16: Wiring to binding posts and copper washers


Image 17: Prior to reassembly

To bring resistance standard into useful condition, Fluke 732B-7001 power pack chassis was reused.


Image 18: Fluke 732B-7001 power pack, bottom side.


Image 19: Fluke SL935 with the remaining faceplates


Image 20: Fluke SL935, remaining front and rear parts.


Image 21: Overall view.


Image 22-23: Fluke SL935 without the top cover shield.

Initial 1 Ω test before teardown
Initial 10 KΩ test before teardown

Fluke 732B-7001 battery/charger design

Modifications and assembly SL935

Thermistor comparator circuit around U507 on charger board was modified to support lower +35°C temperature setpoint of SL935 oven thermistor. This was done by adding 15 KΩ resistor to existing R547 and adding 79.6 KΩ resistor to R549 to move window for proper IN CAL LED operation.

Temperature coefficient performance

Temperature coefficient results with oven

10 KΩ test after assembly with oven ON


Image 24: Initial test with HP 3458A and 10 KΩ output.


Image 25: Initial test with HP 3458A and 1 Ω output.

Initial graphs from HP 3458A(3). Settings are NPLC 200, DELAY 3, 4-wire 10KΩ fixed range.

Calibration procedure

Since Fluke SL935 is not a commercial product, we have to rethink exact calibration procedure for this unit. Given the level of this standard stability, expected to be at least no worse than Fluke 5720A calibrator specification, we can have procedure similar to Fluke 742A and Fluke 732B calibration. Below is listed procedure steps in sequence.

Performance specifications verified with both 1.0000 Ω and 10000 Ω outputs presented in Table 2.

Parameter Condition Minimum Typical Maximum Unit
1.0000 Ω resistance output switch position
Stability ITEST =< 100 mA 28 45 ppm/year
Resistance value ITEST =< 100 mA 1.0000 ±85ppm Ω
Resistance tempco 3 ppm/°C
10000 Ω resistance output switch position
Stability ITEST =< 500 µA 3 6 ppm/year
Resistance value ITEST =< 500 µA 10000 ±9ppm Ω
Resistance tempco 1 ppm/°C
Internal oven temperature ±18-28 °C ambient +34.0 +35.0 +36.0 °C
Thermistor value DB9 connector, pin 4-8 64.35 65.00 65.65
Battery life After full charge 168 180 hours
Charge time 40 hours

Table 2: Fluke SL935 specification defined prior to calibration.

Resistance values calibration

STEP 1 – Initial pre-condition.

If standard was moved to the new bench station/location, it should be powered by mains and fully charged prior to calibration. Connect clean copper spade lug cables for both force and sense terminals of the Fluke SL935. Leave standard for minimum 2 hours to reach thermal equilibrium with environment. Lab temperature shall be kept in range +20…+25 °C.

STEP 2 – Connect resistance bridge and 10 KΩ reference standard for 10 KΩ comparison.

To establish high accuracy resistance transfer it’s recommended to use precision resistance bridge, such as Measurements International 6010 or Guildline Instruments 6622A in combination with high-stability 10 KΩ resistance reference, such as IET Labs SR104. Maintain reference standard as per standard laboratory guidelines.


Image 26: Connections for RX Fluke SL935 DUT, DCC bridge system and reference RREF IET SR104.

Use clean metrology-grade low capacitance cables with insulation resistance >1 TΩ to obtain best results. Use 10k switch position on Fluke SL935 front panel, to bring 10 KΩ at the front terminals.

Ensure that SL935 is disconnected from AC power to avoid possibility of ground loops and unwanted EMI/RFI noise pickup.

STEP 3 – Thermal soaking delay.

Let the all equipment and cables rest 15 minutes to reach thermal equilibrium to reduce risk of unwanted thermal EMF errors. Perform internal SL935 thermistor measurement, to establish internal oven temperature during the calibration. Nominal thermistor resistance during normal operation is 65.00 KΩ ±1%.


Image 27: Fluke SL935 thermistor pinout on rear DB9 port.

Thermistor can be probed from rear DB9 connector, using pin 4 and pin 8, same pinout as Fluke 732B DC Voltage Standard.

STEP 4 – Perform measurement.

Perform resistance measurement using known value of reference RREF resistance standard and 1:1 bridge comparison mode. Follow standard metrology guidelines for bridge operation for 1:1 transfer.

STEP 5 – Replace 10 KΩ reference standard and connect 1 Ω reference standard for 1 Ω comparison.

To establish high accuracy resistance transfer it’s recommended to use precision resistance bridge, such as Measurements International 6010 or Guildline Instruments 6622A in combination with high-stability 1 Ω resistance reference, such as Leeds and Northrup (L&N) 4210 Thomas-type. Maintain reference standard as per standard laboratory guidelines.


Image 28: Connections for RX Fluke SL935 DUT, DCC bridge system and reference RREF L&N 4210.

Use clean metrology-grade low capacitance cables with insulation resistance >10 GΩ to obtain best results. Use 1 switch position on Fluke SL935 front panel, to bring 1 Ω at the front terminals.

Ensure that SL935 is disconnected from AC power to avoid possibility of ground loops and unwanted EMI/RFI noise pickup.

STEP 6 – Thermal soaking delay and oven temperature measurement.

Let the all equipment and cables rest 15 minutes to reach thermal equilibrium to reduce risk of unwanted thermal EMF errors. Perform internal SL935 thermistor measurement, to establish internal oven temperature during the calibration. Nominal thermistor resistance during normal operation is 65.00 KΩ ±1%. Image 28 can be referenced for thermistor probe pinout connections.

STEP 7 – Perform measurement.

Perform resistance measurement using known value of reference RREF resistance standard and 1:1 bridge comparison mode. Follow standard metrology guidelines for bridge operation for 1:1 transfer.

STEP 8 – Assign resistance value and reconnect AC power.

Once measurement is complete, disconnect test equipment and connect Fluke SL935 DUT to AC mains to maintain oven operation and battery charge.

Assign determined resistance value to each of Fluke SL935 resistance outputs, using metrology-standard guidelines and lab methodology. Include uncertainty with assigned value for further use.

Temperature coefficient calibration

Typical metrology range for high-performance resistance transfer standards is +18 °C to +28 °C, which gives 10 degree window for temperature stability evaluation. Suggested workflow to perform this calibration is outlines in steps below.

If standard was moved to the new bench station/location, it should be powered by mains and fully charged. Connect clean copper spade lug cables for both force and sense terminals of the Fluke SL935. Leave standard for minimum 2 hours to reach thermal equilibrium with environment. Lab temperature shall be kept in range +20…+25 °C, following standard metrology guidelines in place.

For tempco test airbath chamber that can fit Fluke SL935 shall be used. Recommended air bath is Measurement International Model 9300A. Temperature range shall be adjustable in range from +18 °C to +28 °C to allow full metrology range calibration.

Only DUT need to be subjected to temperature variation, to establish temperature/resistance correlation. Ensure that SL935 is disconnected from AC power to avoid possibility of ground loops and unwanted EMI/RFI noise pickup.

Perform automated measurement, following same connections and procedure as in STEP 2-4 during resistance value calibration, but this time with air bath temperature variation in 1 °C steps or finer. As the result there should be at least 10 points available with measured resistance:

Formula defined from the calibration, using 0.30 mADC test current:

RTEMP = R[TREF] × (1 + α × (TEMP – TREF) + β × (TEMP – TREF)2 ),

Where for 10 KΩ values are:

α = 0.00890E-6 Ω/°C,
β = 0.00042E-6 Ω/°C2,
R[TREF] = 9999.9747 Ω and TREF = +23.00 °C.

Calculated resistance table would be:

Resistance value Airbath Temperature Measurement Uncertainty Total resistance deviation from +23.00 °C
9999.97436 Ω +18 °C ±0.05 °C ±0.33 ppm -0.034 ppm
9999.97441 Ω +19 °C ±0.05 °C ±0.33 ppm -0.029 ppm
9999.97447 Ω +20 °C ±0.05 °C ±0.33 ppm -0.023 ppm
9999.97453 Ω +21 °C ±0.05 °C ±0.33 ppm -0.016 ppm
9999.97461 Ω +22 °C ±0.05 °C ±0.33 ppm -0.008 ppm
9999.97470 Ω +23 °C ±0.05 °C ±0.33 ppm Reference
9999.97479 Ω +24 °C ±0.05 °C ±0.33 ppm +0.009 ppm
9999.97489 Ω +25 °C ±0.05 °C ±0.33 ppm +0.019 ppm
9999.97500 Ω +26 °C ±0.05 °C ±0.33 ppm +0.030 ppm
9999.97512 Ω +27 °C ±0.05 °C ±0.33 ppm +0.042 ppm
9999.97525 Ω +28 °C ±0.05 °C ±0.33 ppm +0.055 ppm

Table 3: 28 November 2017 data of 10 KΩ measurement for temperature coefficient evaluation

Once test is complete, switch SL935 output resistance to 1 Ω, reconfigure measurement equipment and perform measurement for 1 Ω value as in STEP 5-7. As the result there should be at least 10 points available with measured resistance:

Resistance value Airbath Temperature Measurement Uncertainty Total resistance deviation from +23.00 °C
1.00005896 Ω +18 °C ±0.05 °C ±0.17 ppm -0.062 ppm
1.00005897 Ω +19 °C ±0.05 °C ±0.17 ppm -0.048 ppm
1.00005898 Ω +20 °C ±0.05 °C ±0.17 ppm -0.036 ppm
1.00005900 Ω +21 °C ±0.05 °C ±0.17 ppm -0.023 ppm
1.00005901 Ω +22 °C ±0.05 °C ±0.17 ppm -0.011 ppm
1.00005902 Ω +23 °C ±0.05 °C ±0.17 ppm Reference
1.00005903 Ω +24 °C ±0.05 °C ±0.17 ppm +0.011 ppm
1.00005904 Ω +25 °C ±0.05 °C ±0.17 ppm +0.021 ppm
1.00005905 Ω +26 °C ±0.05 °C ±0.17 ppm +0.031 ppm
1.00005906 Ω +27 °C ±0.05 °C ±0.17 ppm +0.041 ppm
1.00005907 Ω +28 °C ±0.05 °C ±0.17 ppm +0.050 ppm

Table 4: Example of 1 Ω measurement range for temperature coefficient evaluation

Formula defined from the calibration is same as in case with 10 KΩ, this time using 100 mADC test current:

RTEMP = R[TREF] × (1 + α × (TEMP – TREF) + β × (TEMP – TREF)2 ),

Where for 1 Ω values are:

α = 0.01117E-6 Ω/°C,
β = 0.00023E-6 Ω/°C2,
R[TREF] = 1.00005902 Ω and TREF = +23.00 °C.

Once measurement is complete, disconnect test equipment and connect Fluke SL935 DUT to AC mains to maintain oven operation and battery charge.

Calibration reports

First calibration was performed by Process Insturments in USA, 28 November 2017.

It didn’t go 100% smooth, as unit arrived from CAL with LOW BAT light on and no IN CAL light. Perhaps calibration tec did not hook it back to AC power after the measurements.

It was shipped to calibration lab fully charged, with battery life minimum 120 hours. That was enough time for SL935 to receive calibration in “hot” state, without any temperature interruptions, even if it was never plugged into mains. Shipping back from PI was 2-day service, which took just 52 hours from pickup to delivery.

Summary & Conclusion

Here I will keep own record for calibration history since the standard acquisition in 2 May 2017.

Date 10 KΩ 1 Ω 10K,ppm Δ 1Ω,ppm Δ Instrument 10 KΩ Δ/day 0, ppm 1 Ω Δ/day 0, ppm Ambient, °C Thermistor
30 April 2017 w/o oven 9999.988 A 1.0000575 A 0 0 3458A-1 0 +25.6 DNT
1 May 2017 w/o oven 9999.995 A +0.70 (reference) 3458A-1 0 (reference) +23.6 DNT
2 May 2017 w/o oven 9999.985 C 1.0000605 C -1.00 0 (reference) 3458A-1 0 0 (reference) +26.2 DNT
7 May 2017 assembled 9999.9938 +0.88 3458A-1 +0.12 +22.8 65.02 kΩ
8 May 2017 assembled 9999.9954 +0.16 3458A-1 +0.04 +23.3 65.xx kΩ
9 May 2017 assembled 1.0000594 -0.7 (noise) 3458A-1 -1.1 +23.6 65.xx kΩ
10 May 2017 assembled 1.0000595 +0.1 (noise) 3458A-1 -1.0 +22.8 65.xx kΩ
11 May 2017 assembled 1.0000601 +0.6 (noise) 3458A-1 -0.4 +22.0 65.xx kΩ

Table 5. Initial measurement data results

Date 10 KΩ 1 Ω 10K,ppm Δ 1Ω,ppm Δ Instrument 10 KΩ Δ/day 0, ppm 1 Ω Δ/day 0, ppm Ambient, °C Thermistor
13 Oct 2017 9999.9594 3458A-2 +23.6
13 Oct 2017 100mA ISRC 1.0000665 3458A-2/9823 +24.6 64.72 kΩ

Table 6. Pre-calibration measurement data results

Date 10 KΩ 1 Ω 10K,ppm Δ 1Ω,ppm Δ Instrument 10 KΩ Δ/day 0, ppm 1 Ω Δ/day 0, ppm Ambient, °C Thermistor
28 Nov 2017 0.3mA 9999.9747 ±0.33 ppm Reference MI 6010B + L&N 4210 Reference +23.00 ±0.05
28 Nov 2017 100mA 1.00005902 ±0.17 ppm Reference MI 6010B + L&N 4210 Reference +23.00 ±0.05

Table 7. Calibration measurement data results

As bonus, we can perform some math acrobatics, to attempt our own meters accuracy, using previously captured data.
Here are my own measurements which I did at home, prior to sending SL935 for calibration + ManateeMafia’s meter:

HP 3458A, meter 2 over 10 KΩ, direct 4-wire, OCOMP ON, DELAY 3, NPLC100 = 9999.971 Ω, +26 °C ambient, 18 May 2017. Error from PI value : -0.37 ppm
HP 3458A, meter 1 over 10 KΩ, direct 4-wire, OCOMP ON, DELAY 3, NPLC100 = 9999.985 Ω, +25 °C ambient, 20 May 2017. Error from PI value : +1.03 ppm
HP 3458A, meter 2 over 10 KΩ, direct 4-wire, OCOMP ON, DELAY 3, NPLC100 = 9999.965 Ω, +24.5 °C ambient, 25 Sept – 28 Sept 2017. Error from PI value : -0.97 ppm

Meter 1 calibrated by using Vishay VHP103 resistor at 10 KΩ as transfer standard, which are in turn measured against 90-hour fresh cal’d ESI SR104 (by PI too) in August 2016. Meter is 24/7 powered. Meter 2 calibrated directly vs same ESI SR104 back in January 2017, shipped to me in Feb and in 24/7 powered use since then.

HP 3458A, ManateeMafia’s meter over 10 KΩ, direct 4 wire, OCOMP ON, DELAY 3, NPLC100 = 9999.966 Ω, +26 °C ambient, 3 Nov – 7 Nov 2017. Error from PI value : -0.87 ppm

Based on this brief data, with adding uncertainties and error factors, I’d have pretty good confidence stating that such transfer path can provide <4 ppm absolute resistance accuracy at 10 KΩ range.
Noise + short-temp variations are usually under ± 1.5 ppm.

Now similar data for 1 Ω.

HP 3458A, meter 1 over 1 Ω, direct 4-wire, OCOMP ON, DELAY 3, NPLC100 = 1.00006111 Ω, +22 °C ambient, 5 May – 25 May 2017. Error from PI value : +2.1 ppm , noise about 30 ppm.
HP 3458A, meter 2 + TE 9823 calibrator sourcing +100mA and -100mA over 1 Ω, direct 4-wire, NPLC 100 = 1.00006657 Ω, +23.5 °C ambient, 13 October 2017. Error from PI value : +7.5 ppm, noise around 2 ppm
HP 3458A, meter 2 + TE 9823 calibrator sourcing +20mA and -20mA over 1 Ω, direct 4-wire, NPLC100 = 1.00006574 Ω, +21 °C ambient, 21 Sept – 22 Sept 2017. Error from PI value : +6.7 ppm , noise around 2 ppm

HP 3458A, ManateeMafia’s meter over 1 Ω, direct 4 wire, OCOMP ON, DELAY 0, NPLC100 = 1.00005831 Ω, +24 °C ambient, 31 Sept – 3 Nov 2017. Error from PI value : -0.7 ppm

This unique resistance standard will be used in xDevs.com projects and lab comparisons and future HP/Agilent/Keysight 3458A DMM and Fluke 5700A calibrator performance maintenance.


Image 29-30: Fluke SL935 front and rear side after restoration

Now it’s ready to be used in lab.


Image 31-32: Fluke SL935 front and rear side after restoration

Cost breakdown for this project presented below:

Item Cost Shipping Supplier
Fluke SL935 prototype $630 $30 bbs.38hot.net forum
Fluke 732B-7001 charger $360 $50 eBay
VRLA 12V 7.0Ah $20 N/A Local electronics store
Shipment for calibration $300 Express shipping
Calibration from Process Instruments $900 Both 1 Ω and 10 KΩ cal + tempco measurement 18°C-28°C
Total $1090 USD + calibration $1200 USD = $2290

Table 8. Costs summary

Given the performance and level of this resistance standard, total cost sounds reasonable. Even single Fluke 742A-10K resistance standard on secondary market usually costs more than this, let alone brand new from Fluke Calibration. Only time and periodic measurements would tell how stable is this SL935 resistance box, and even then stability measurement would be questionable due to measurement instrument own long-term stability and drift. To perform resistance measurements at ppm-level accuracy, any single DMM, even great HP 3458A is not enough. After getting some initial data during next few months, I plan to send SL935 for calibration lab to have low uncertainty reference. After doing this procedure for few years, we will eventually come up with annual stability/drift figurues.

And of course SL935 looks happy with it’s production brother, Fluke 732B:


Image 33: Fluke 732B and SL935 standards.

Until then, stay tuned and let us know your feedback!

Author: Illya Tsemenko
Published: April 28, 2017, 3:11 p.m.
Modified: Dec. 2, 2017, 1:04 a.m.

Comments