Keithley Model 2001 repair and service notes

Contents

Intro

The Model 2001 from Keithley (today part of Tektronix) is a high performance 7½-digit bench digital multimeter, which share same appearance to other Keithley 2xxx series measuring/sourcing tools and designed to fit into standard 19” rack. It have half-width size, allowing two units to be installed side by side. Such format perfectly fits main purpose of system meters, which is often a choice of use for production testing/validation and research. The build quality of the hardware is pretty good and construction is robust enough to withstand most of generic usage as well.

Setting up the 2001 is straight forward, it’s easy to use right from power on, and does not need reading operator manual to perform most of usual functions.

Key specs and features are:

  • 18 ppm base DCV Accuracy
  • High Throughput, up to 2000 Readings per Second
  • Solid State Protection to 1600V
  • 1µsec Spike Detection
  • AC Crest Factor Display
  • Multiple Measurement Displays
  • 10nV Sensitivity
  • On-Line INFO
  • 10-Channel Scanner Option
  • In-Circuit Current to 12 A
  • 4½, 5½, 6½, or 7½ Digits
  • 2µsec Trigger Latency
  • 2 MHz AC Voltage Measurement with Peak, Average, & True RMS AC Measurements
  • Bar Graph Displays
  • Frequency Measurements to 15MHz
  • RTD and Thermocouple Measurements
  • SCPI Commands, IEEE 488.2 Compatible
  • Trigger-Link
  • 128kB Reading Memory Option
  • Hi-Lo Limits with Digital Outputs

In this repair article we will go thru overview, check for broken component/blocks of few old used units we have, as well as cover repair methodology and service information notes. This hopefully will enable owners of Model 2001 meters to fix their damaged instruments and continue to enjoy their performance for few more decades. As new instruments emerge on market, such as DMM7510 and 34470A there are still very limited number of DMMs can archive specs which Model 2001 have, specially considering relatively affordable (often less than 1000$ USD) offerings for these meters on second market.

Even considering that Model 2001 was developed quite long time ago (around 1990-1992), it’s still manufactured and sold officially on Tektronix website for no less then $4880 USD (base unit w/o options). Given it’s specs and resolution, this is still one of not so many 7½ digit DMMs on market. It took more than 20 years till release of recent 7½-digit refresh unit, Model DMM7510. While new DMM7510 provide new fancy interface, in some aspects old-school Model 2001 is still superior and provide scan-card options.

Disclaimer

Redistribution and use of this article or any 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/kei2001/) 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/kei2001/), 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 any other 3rd party, including Keithley Instruments or Tektronix 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 action, arising out of or in connection with the use or performance of information published here.

If you willing to contribute or add your experience regarding any test instruments or provide any extra information, such as firmware dumps, internal photographs, you can do so following these simple instructions

Service information and downloads

Most important first, schematics for repairs:

Before we start with repairs, here are official manuals including schematics:

Unofficial schematics , reverse engineering from Rev.L PCB
This schematics provided as is, reverse engineered. Not complete, Use with caution.

Firmware dumps

Firmware version Combined binary ODD ROM EVEN ROM
Model 2001 A05 2001-A05 ROM binary LEND
Model 2001 A06 2001-A06 ROM binary LEND
Model 2001 A08 (Latest for A-verison, 4Mbit) 2001-A08 ROM binary LEND
Model 2001 B06 2001-B06 LEND 2001-803-B06 2001-804-B06
Model 2001 B07, Version 13.1, Feb 4 10:18:36 EST 1994 2001-B07 LEND 2001-803-B07 2001-804-B07
Model 2001 B08, Version 17.1, Sep 6 16:12:29 EDT 1994 2001-B08 ROM Binary 2001-B08 LEND 2001-803-B08 2001-804-B08
Model 2001 B09, Version 23.1, Jun 12 12:17:07 EDT 1995 2001-B09 LEND 2001-803-B09 2001-804-B09
Model 2001 B10, Version 29.1, Feb 5 09:56:39 EST 1996 2001-B10 ROM Binary 2001-B10 LEND 2001-803-B10 2001-804-B10
Model 2001 B11, Version 35.1, May 6 15:15:32 EDT 1998 2001-B11 LEND 2001-803-B11 2001-804-B11
Model 2001 B13, Version 45.1, Apr 29 13:10:54 EDT 1999 2001-B13 LEND 2001-803-B13 2001-804-B13
Model 2001 B15, Version 54.1, Oct 23 08:57:18 EDT 2000 2001-B15 ROM Binary 2001-B15 LEND 2001-803-B15 2001-804-B15
Model 2001 B16, Version 57.1, Sep 8 14:51:30 EDT 2003 2001-B16 LEND 2001-803-B16 2001-804-B16
Model 2001M B17, Version 7.1, Nov 5 08:32:34 EST 2009 2001M-B17 LEND 2001M-803-B17 2001M-804-B17

Older A-version (single 4Mbit ROM, no support Keithley 1801 nV preamp)
B-version firmwares are used on Rev.K digital PCB with two 4MBit ROMs (even and odd bytes).

Key differences of B-version code:

  • Supports Keithley Model 1801 nV preamp
  • Supports internal Keithley Model 2031 phase measurement extension module
  • Supports internal Keithley Model 2051 TC-OHM extension module
  • Supports SPRTD and Thermocouple Type N sensors from B11 version (same as Keithley 2002 A09+).

Do not flash little endian binaries (LEND) into ROM, they will not work. Motorola 68000 is using big-endian encoding, with LSB half first. Little-endian binaries are available only for firmware analysis purpose. To convert them simple python app was used:

# xDevs.com Firmware endian conversion app BE -> LE1
# https://xdevs.com/fix/kei2001

import os

with open('Input_BE.BIN','rb') as a:
    with open('Output_LE.bin','wb') as x:
        for cnt in range(0, 524288):
            y = a.read(1)
            z = a.read(1)
            x.write ("%s%s" % (z, y) ) 

Based on binary comparison, firmware since version B11 (similar to Keithley 2002’s A09) added support for Type N thermocouples and SPRTD which was absent in A06, at least based on text strings difference. Also there is new extra strings – HPR.HPRECISION, ISN, A4, B4, A7, B7, C7, RZERO, RZER. It seems that A06 have also only 9 settings profiles (SAV0 to SAV8), while A09 and later have 10 (SAV0 to SAV9).

HPR, RZERo, SPRTD string is related to GPIB readings format and RTD temperature sensors, as explained in User’s manual:

SPRTD (RZERO, A4, B4, A7, B7, C7) settings are explained on page 233-234 as well.

One more person mentioned in A09’s credits: Justin Noble

I do not know what is difference between other versions, as there are no public revision logs available.
Combined binary images were merged using simple python tool:

# xDevs.com Firmware combine tool 
# https://xdevs.com/fix/kei2001

import os

with open('2002_804_a06_even_0150fa76.bin','rb') as a:
    with open('2002_803_a06_odd_01c9ab74.bin','rb') as b:
        with open('K2002_A06_FULL.bin','wb') as x:
            for cnt in range(0, 1048576):
                x.write ("%s%s" % (a.read(1),b.read(1) ) ) 

So likely firmware version B11 changeset to older B09 is next:

  • Support for Type N thermocouples
  • Support for SPRTD which was absent in A06
  • New format setting : HPR (HPRECISION)
  • Unknown command ISN?
  • Coefficient settings for SPRTD: A4, B4, A7, B7, C7
  • RTD custom user setting RZERO, RZER, to set resistance at 0°C

There is easy way to read and backup calibration ROM as well, using external programmer, such as TL866. Procedure and connections shown on photograph below.

It’s wise to backup calibration ROM before attempting own recalibration or sending unit to service, just in case.

Some of my calibration ROM dumps:

Calibration ROM Dump
DigPCB: A07967, Year 2014, Unit 0544257, Tektronix TW cal Binary
DigPCB: A19956, Year 1993, Unit 0549310 Binary
DigPCB: A14641, Year 1998, Unit 0546015 Binary
DigPCB: A10142, Year 2000, Unit 0545483 Binary
DigPCB: A39541, Year 1994, Unit 0579220 Binary
DigPCB: A09513, Year 1992, Unit 0545274 Binary
DigPCB: A07543, Year 1999, Unit 0544015 Binary

Many other Keithley Instruments firmware dumps are available on xDevs.com documentation site

If you have Keithley Model 2001 firmware (or any other Keithley instrument firmware), please upload here.

Calibration decode for EEPROM contents

Calibration constants from K2001 can be read using GPIB command. They are stored in plain binary format in the ROM. Table C-2 of 2001-905-01G calibration manual lists the response to the :CAL:PROT:DATA? query. The response is an ASCII string of 99 numbers separated by commas, and is not affected by the FORMAT command. Constants listed in Table C-2 are shown in the order they are sent.

Constant name Unit K2001 #7 Unit K2001 #1 Detailed description ROM Offset Data format
g1 1.000038E+00 9.997173E-01 RMS gain for 2V, 200V, and 750V AC ranges 0×018 double float IEEE 754
of1 1.533712E-04 -1.583679E-04 RMS offset for 2V, 200V, and 750V AC ranges 0×020 double float IEEE 754
g10 1.000523E+01 1.000293E+01 RMS gain for 200mV and 20V AC ranges 0×028 double float IEEE 754
of10 1.704422E-04 -1.076852E-04 RMS offset for 200mV and 20V AC ranges 0×030 double float IEEE 754
gfwr1 9.999993E-01 9.999961E-01 Average gain for 2V, 200V, and 750V AC ranges 0×038 double float IEEE 754
offwr1 2.698059E-05 -3.915291E-05 Average offset for 2V, 200V, and 750V AC ranges 0×040 double float IEEE 754
gfwr10 1.000433E+01 1.000551E+01 Average gain for 200mV and 20V AC ranges 0×048 double float IEEE 754
offwr10 1.689877E-04 8.726960E-05 Average offset for 200mV and 20V AC ranges 0×050 double float IEEE 754
a100 1.000210E+02 9.998019E+01 100:1 divider attenuation factor 0×058 double float IEEE 754
a500 4.963352E+02 4.962888E+02 500:1 divider attenuation factor 0×060 double float IEEE 754
ofpkpos10 6.097749E-03 1.418733E-02 Positive peak offset for 200mV AC range 0×068 double float IEEE 754
ofpkneg10 1.117921E-02 1.722747E-02 Negative peak offset for 200mV AC range 0×070 double float IEEE 754
ofpk1 4.573312E-03 4.053523E-03 Positive and negative peak offset for 2V, 200V, and 750V AC ranges 0×078 double float IEEE 754
ofpkpos20 6.097749E-03 7.600356E-03 Positive peak offset for 20V AC range 0×080 double float IEEE 754
ofpkneg20 1.117921E-02 7.093666E-03 Negative peak offset for 20V AC range 0×088 double float IEEE 754
div100self 1.125000E+02 1.340000E+02 Self-calibration code for frequency compensation DAC, 100:1 divider 0×090 double float IEEE 754
div500self 9.000000E+01 1.260000E+02 Self-calibration code for frequency compensation DAC, 500:1 divider 0×098 double float IEEE 754
noise10 1.864359E-04 2.238526E-04 Noise factor for 200mV and 20V AC ranges 0×0a0 double float IEEE 754
cfc1 1.000331E+00 1.000316E+00 Crest factor correction factor for 2V, 200V, and 750V AC ranges 0×0a8 double float IEEE 754
cfc10 1.000021E+00 1.000284E+00 Crest factor correction factor for 200mV and 20V AC ranges 0×0b0 double float IEEE 754
acdclow 1.000346E+00 1.000340E+00 AC-coupled correction factor for 200mV and 2V AC ranges 0×0b8 double float IEEE 754
acdchigh 1.000199E+00 1.000192E+00 AC-coupled correction factor for 20V, 200V, and 750V AC ranges 0×0c0 double float IEEE 754
inputtc 1.081469E-01 1.044814E-01 Input time constant 0×0c8 double float IEEE 754
acdccur 1.000348E+00 1.000487E+00 AC-coupled correction factor for AC current 0×0d0 double float IEEE 754
compval1 1.215000E+02 1.190000E+02 RMS comparator DAC code 0×0d8 double float IEEE 754
div100 1.135000E+02 1.340000E+02 Frequency-compensation DAC code for 20V AC range 0×0e0 double float IEEE 754
div200 1.130000E+02 1.330000E+02 Frequency-compensation DAC code for 200V AC range 0×0e8 double float IEEE 754
div500 9.300000E+01 1.220000E+02 Frequency-compensation DAC code for 750V AC range 0×0f0 double float IEEE 754
div100off 1.000000E+00 0.000000E+00 Frequency-compensation DAC offset for 20V AC range 0×0f8 double float IEEE 754
div200off 5.000000E-01 -1.000000E+00 Frequency compensation DAC offset for 200V AC range 0×100 double float IEEE 754
div500off 3.000000E+00 -4.000000E+00 Frequency compensation DAC offset for 750V AC range 0×108 double float IEEE 754
Check sum 0xA5 0xB4 16-bit summator of cal values 0×18 – 0×108 , rotated byte by 4 0×110 2 bytes, 16 bits
dcv[200mV]gain 1.395384E+00 1.385034E+00 200mV DC gain 0×112 double float IEEE 754
dcv[200mV]offset -1.541521E-05 -5.612224E-05 200mV DC offset 0×11a double float IEEE 754
dcv[2V]gain 1.743986E+00 1.730974E+00 2V DC gain 0×122 double float IEEE 754
dcv[2V]offset -1.732121E-06 -5.877960E-06 2V DC offset 0×12a double float IEEE 754
dcv[20V]gain -6.976404E-01 -6.924144E-01 20V DC gain 0×132 double float IEEE 754
dcv[20V]offset -1.150898E-05 -3.787601E-05 20V DC offset 0×13a double float IEEE 754
dcv[200V]gain 1.743655E+00 1.730672E+00 200V DC gain 0×142 double float IEEE 754
dcv[200V]offset -1.008386E-06 2.410488E-06 200V DC offset 0×14a double float IEEE 754
dcv[1000V]gain 6.975011E-01 6.922715E-01 100V DC gain 0×152 double float IEEE 754
dcv[1000V]offset 5.484389E-04 -1.772726E-03 1000V DC offset 0×15a double float IEEE 754
dca[200uA]gain 1.396956E+00 1.386757E+00 200ΩA DC gain 0×162 double float IEEE 754
dca[200uA]offset -1.814614E-05 1.600430E-04 200ΩA DC offet 0×16a double float IEEE 754
dca[2mA]gain 1.397030E+00 1.386588E+00 2mA DC gain 0×172 double float IEEE 754
dca[2mA]offset -2.236752E-05 1.604392E-04 2mA DC offset 0×17a double float IEEE 754
dca[20mA]gain 1.394508E+00 1.384073E+00 20mA DC gain 0×182 double float IEEE 754
dca[20mA]offset -2.232714E-05 1.601482E-04 20mA DC offset 0×18a double float IEEE 754
dca[200mA]gain 1.379412E+00 1.370280E+00 200mA DC gain 0×192 double float IEEE 754
dca[200mA]offset -2.208545E-05 1.585523E-04 200mA DC offset 0×19a double float IEEE 754
dca[2A]gain 1.394483E+00 1.384269E+00 2A DC gain 0×1a2 double float IEEE 754
dca[2A]offset -2.232674E-05 1.601709E-04 2A DC offset 0×1aa double float IEEE 754
ohm220gain 1.500375E+00 1.516462E+00 2-wire 20Ω gain 0×1b2 double float IEEE 754
ohm220offset -9.411514E-03 -1.058161E-02 2-wire 20Ω offset 0×1ba double float IEEE 754
ohm2200gain 1.430062E+00 1.430543E+00 2-wire 200Ω gain 0×1c2 double float IEEE 754
ohm2200offset -9.552922E-04 -1.109982E-03 2-wire 200Ω offset 0×1ca double float IEEE 754
ohm2[2k]gain 1.787328E+00 1.787851E+00 2-wire 2kΩ gain 0×1d2 double float IEEE 754
ohm2[2k]offset -9.572465E-05 -1.112534E-04 2-wire 2kΩ offset 0×1da double float IEEE 754
ohm2[20k]gain 1.965853E+00 1.966221E+00 2-wire 20kΩ gain 0×1e2 double float IEEE 754
ohm2[20k]offset -1.134743E-05 -1.719503E-05 2-wire 20kΩ offset 0×1ea double float IEEE 754
ohm2[200k]gain 2.479002E+00 2.475204E+00 2-wire 200kΩ gain 0×1f2 double float IEEE 754
ohm2[200k]offset -3.401631E-06 -9.457002E-06 2-wire 200kΩ offset 0×1fa double float IEEE 754
ohm2[2M]gain 2.273309E+00 2.272104E+00 2-wire 2MΩ gain 0×202 double float IEEE 754
ohm2[2M]offset -2.351793E-06 -7.820685E-06 2-wire 2MΩ offset 0×20a double float IEEE 754
ohm2[20M]gain 2.503775E+00 2.493262E+00 2-wire 20MΩ gain 0×212 double float IEEE 754
ohm2[20M]offset -2.496137E-06 -8.477018E-06 2-wire 20MΩ offset 0×21a double float IEEE 754
ohm2[200M]gain 3.925414E+00 3.909291E+00 2-wire 200MΩ gain 0×222 double float IEEE 754
ohm2[200M]offset 1.234136E-02 -4.004449E-02 2-wire 200MΩ offset 0×22a double float IEEE 754
ohm2[1G]gain 1.570253E+00 1.563722E+00 2-wire 1GΩ gain 0×232 double float IEEE 754
ohm2[1G]offset 1.234442E-03 -4.003584E-03 2-wire 1GΩ offset 0×23a double float IEEE 754
ohm420gain 1.500375E+00 1.516462E+00 4-wire 20Ω gain 0×242 double float IEEE 754
ohm420offset 3.446987E-05 4.566308E-05 4-wire 20Ω offset 0×24a double float IEEE 754
ohm4200gain 1.430062E+00 1.430543E+00 4-wire 200Ω gain 0×252 double float IEEE 754
ohm4200offset 3.285450E-05 4.307594E-05 4-wire 200Ω offset 0×25a double float IEEE 754
ohm4[2k]gain 1.787328E+00 1.787851E+00 4-wire 2kΩ gain 0×262 double float IEEE 754
ohm4[2k]offset 3.026024E-06 4.242987E-06 4-wire 2kΩ offset 0×26a double float IEEE 754
ohm4[20k]gain 1.965853E+00 1.966221E+00 4-wire 20kΩ gain 0×272 double float IEEE 754
ohm4[20k]offset 3.328274E-06 4.666301E-06 4-wire 20kΩ offset 0×27a double float IEEE 754
ohm4[200k]gain 2.479002E+00 2.475204E+00 4-wire 200kΩ gain 0×282 double float IEEE 754
ohm4[200k]offset 4.197058E-06 5.874237E-06 4-wire 200kΩ offset 0×28a double float IEEE 754
n7vref 6.976333E-01 6.923925E-01 7V reference value 0×2d2 double float IEEE 754
rollover 9.999967E-01 9.999954E-01 ±2V rollover 0×2da double float IEEE 754
mux4d711 15813 15897 Multiplexer 4-1/2 digit counts (AC peak) 0×2e2 uint32_t
mux4d711p5 8951 9068 Multiplexer 4-1/2 digit counts (AC peak) 0×2e6 uint32_t
mux4d215 12388 12479 Multiplexer 4-1/2 digit counts (AC peak) 0×2ea uint32_t
mux4d011 29542 29562 Multiplexer 4-1/2 digit counts (AC peak) 0×2ee uint32_t
mux4d015 29542 29562 Multiplexer 4-1/2 digit counts (AC peak) 0×2f2 uint32_t
mux4d0150 29540 29561 Multiplexer 4-1/2 digit counts (AC peak) 0×2f6 uint32_t
mux4d011p5 29542 29561 Multiplexer 4-1/2 digit counts (AC peak) 0×2fa uint32_t
mux4dF150 1282 1296 Multiplexer 4-1/2 digit counts (AC peak) 0×2fe uint32_t
mux4dF15 6456 6461 Multiplexer 4-1/2 digit counts (AC peak) 0×302 float IEEE 754
i20 9.300236E-03 9.133327E-03 20Ω range current source value 0×306 float IEEE 754
i200 9.757503E-04 9.681874E-04 200Ω range current source value 0×30a float IEEE 754
i2k 9.757503E-04 9.681874E-04 2kΩ range current source value 0×30e float IEEE 754
i20k 8.871397E-05 8.803561E-05 20kΩ range current source value 0×312 float IEEE 754
i200k 7.035033E-06 6.993260E-06 200kΩ range current source value 0×316 float IEEE 754
i2m 7.671572E-07 7.618376E-07 2MΩ range current source value 0×31a float IEEE 754
i20m 6.965424E-08 6.942609E-08 20MΩ range current source value 0×31e float IEEE 754
i200m 4.442807E-09 4.427848E-09 200MΩ range current source value 0×322 float IEEE 754
i1g 4.442807E-09 4.427848E-09 1GΩ range current source value 0×326 float IEEE 754

Also extra checksum visible at 0×32A offset – 2 bytes – 0×05 0×97. Calibration date stored in ASCII , following MM/DD/YY formatting on offset 0×32C – 8 bytes. 0×334 offset has 0×00 byte. Next calibration date – ASCII MM/DD/YY on offset 0×335 – 8 bytes
All binary data stored in Big endian

Exterior

Display and user interface

2001’s control and menu system is quick and easy to navigate through, thanks to great dot-matrix VFD screen with great contrast and brightness. Screen have two main lines for readout and auxiliary information, like current settings, statistics, and service markers on top which shows current operation mode and status. Same VFD, is used in higher end 8½-digit brother, Model 2002, switch units 7001 and 7002, and many of 24xx series and 26xx SMU instruments, as well as some other specialized Keithley meters.

Display with part-number DD-51 is custom, specially made by Newheaven Display or Noritake Itron for Keithley Instruments. It is passive glass with multiplexed segments, so all driver circuitry and display/keypad microcontroller is located on front panel. Communications with main processor is done via TTL-level UART interface.

Internal structure and block details

Detail closeup for components on digital board:

ROM chip

Overall Keithley 2001 repair procedure

Based on repair experience of no less than seven (yes, that’s right, 7!) DMMs I had my hands on and lot of feedback from other Model 2001 owners, general methodology for this DMM repair fits procedure below very well:

  • DO NOT turn meter on first, to avoid more damage, take it apart.
  • If your meter more than 8 years old (check date codes on ICs), it is imperative to replace ALL electrolytic capacitors on BOTH analog and digital PCBs. This is first step, do not omit it, or more damage to parts and fragile 4-layer PCBs will occur.
  • Clean any corroded traces, carefully check and restore bad vias and connections around leaked capacitors. Make sure no pads or traces are broken/open on PCB. Sometimes trace is eaten inside and damage is hard to see,
  • Check avalanche diodes and CR513, CR514 zeners in bootstrap power supply section
  • Replace 40×40mm 12V fan, if meter is old. Clean air filter to have nice airflow.
  • Now after all power damaged parts checked and fixed, try to power on meter and probe supply voltages (+5, +15, -15, +38, -38, etc). Keep an eye on parts, make sure nothing get scorching hot.
  • Run SELFTEST diagnostics. Some of errors and their solution in my case were presented below in article.
  • If you have many 4xx.x series errors check U520 and U525. These op-amps have floating power and can be damaged from bad capacitors in doubled voltage supply.
  • Check front/rear inputs switch. If it’s dirty or does not actuate well, remove switch from PCB and carefully clean it. It’s serviceable, but be careful with contacts, easy to damage.
  • Backup your ROM firmware and calibration EEPROM (24C16 on digital board).
  • You will need stable and known 2VDC, 20VDC, 19-20k&Omega and 1MΩ standards to do user calibration Model 2001

Below I’ll cover repair example and worklog on four old meters, covering most of common issues with these old units. There are multiple PCBA revisions over the years exist in these multimeters, but most of work will be done covering Rev.J and Rev.L.

Unit #1 repair section

This meter is the one I bought first. It had same issues with leaked capacitors in voltage multiplier, bad avalanche diodes and PCBA damages on analog board.

After replacement of these parts and all electrolytic capacitors, issues with DCV and OHM functions were resolved. This was also confirmed by passing all 3xx.x series tests in self-diagnostics. But ACV, ACI and DCI modes are faulty. ACV reporting presence of 480VAC on 750V range, overflow on lower ranges and incorrect readings on current measurement as well, except DCI 2A range.

Many 4xx.x tests we failing in self-diagnostics.

After few days of troubleshooting culprit was found in Linear Systems U441 dual JFET pair in floating FE circuitry, causing current mirror to become unleveled and failing all related tests and operation. U520 had incorrect drive due to one of JFETs in Q512 gone bad. Temporary replacement with NXP PMBFJ620 in SOT363 corrected the issue and meter was happily passing SELF-TEST routines. Both parts are marked in red on schematics:

Rework was not pretty, as original Linear Systems U441 is in hermetic TO-71 6-lead package, while PMBFJ620 is tiny SOT-363. I have new U441 on it’s way to make it right later.

Hangs/No comm link error after warm-up

Two of repaired meters had issue related to digital section. After initial warm-up, which took 15-30 minutes, meter’s digital circuitry hang, not responding to any keys. After power cycle meter was unable to boot, and front panel had just “No comm link” message displayed. This happening when main processor unable to send any commands to display board MCU.

I checked power supply on digital board, which is only +5VDC from separate transformer winding, rectifier bridge, it was all fine. All electrolytic capacitors were replaced before hand, so it’s not them.

Issue on one unit was discovered to be bad NVRAM chip, Dallas DS1245Y. It was one of the chips bought from eBay, and with removal of memory option (leaving socket empty) – 2001 had no more issues even after 24 hours of operation. As DS1245Y is sitting on same memory bus as main SRAM and ROM, if it have intermittent faults it could corrupt processor data bus causing hangs.

Unit #1 calibration section

Unit #2 repair

This 2001 was bought broken, as for parts. Initial inspection indicated there were missing critical parts and damages, as shown in chapter below. Also there was no A/D converter board, so we had used spare board bought separately.

Analog board (Rev.L)

Missing parts were located around voltage reference

U329 196-600A IC, PRECISION REFERENCE Linear LM399 Precision ovenized 7V voltage reference
U330 Linear LTC1043 Capacitor switching block
U327 Linear LT1050 1.75V reference output opamp
Q540 BT3904 NPN transistor

p=.

Q330 TG-254 TRANS, DUAL N-CHAN JFET, IFN146 (TO-71) Matched JFET
Q512 TG-235 TRANS, N-CHAN. DUAL JFET, U441 Matched JFET

Previous owner tried to fix by putting wrong diode on CR111 :(. And as lot of old Model 2001 units this one had issues with leaking caps and damaged PCB around bootstrap capacitor supply nodes (capacitors C114,C115,C116,C117 and avalanche diodes near them).

Rest of board look OK, but likely to have hidden damages in some parts we don’t know about just yet.

Overall board photos for both sides:

Digital board (Rev.K, same as used in Keithley 2002 units)

Missing R758 and damaged footprint pad for it

Overall board photos for both sides:

Repair

First step was to get back all parts. While LM399, LTC1043 , LT1050 are readily available either from direct

Linear shop or Digikey, parts like matched obsolete JFETs are way harder to obtain. One of not as straight forward, but rather easy route is to buy another 2001 unit which is beyond repair, as donor for salvaging needed parts. Luckily during that time there was one for sale around 100$ USD with many parts missing on analog boards, but it had these Q330 and Q512 JFETs on it. So I got this, already third, meter just to donor parts.

First thing to do servicing old Model 2001, is to replace fan to fresh new one, to avoid overheating and capacitor life reduction. In this case new Delta EFB0412MA (40×40 mm, DC 12V, 90mA) was installed. Also ordered and installed capacitors, missing parts, such as voltage reference:

And capacitor block, Q330, Q512.

DIP holes for LTC1043 are easy to clean using copper solder wick. Be careful not to overheat PCB, since this 4-layer board don’t have solid copper polygons on inner layers and easy to delaminate from high temperature exposure. ERSA iCON soldering station with temperature set to +275°C.

Install 2001-162 Revision H A/D converter board

Now after check for shorts or open connections, power on unit. It did power on okay, initial check on +5V, +15, -15V, boost supplies +38 and -38V shown no problematic deviation.

Since testing function of such sophisticated DMM is not an easy job, Keithley implemented handy self-diagnostics mode, when meter using it’s ADC and internal switches to test for various blocks and compare measured signal with reference value stored in ROM. Running test is easy, by navigating GENERAL/TEST/SELF-TEST menu.
Resulting list of errors is scary-some:

  • 105.2
  • 304.6, 304.7,
  • 401.1, 402.1, 404.1
  • 404.2, 404.3, 404.4, 404.5,
  • 405.2, 405.4, 405.6, 405.8
  • 406.6
  • 407.1, 407.2, 407.3
  • 409.6
  • 410.1, 411.1, 411.2, 412.1

Error 105.2

This was easy fix FIXED by repairing connection for knocked off resistor R758 on digital board and putting 39 ohm 1206 resistor back.

Errors 304.6, 304.7

These were fixed by cleaning PCB with isopropyl alcohol around ohms current source on analog board and range switches very well. On high resistance ranges test current sourced by meter is mere hundred nanoamp, and contamination on PCB surface or dirty component leads can easily upset such sensitive signal path. General rule for handling high-precision PCBA is not to touch anything with bare fingers (as oil and dust from fingers contaminate surfaces and cause unwanted leakages).

40x series errors

Having so many similar errors usually indicates that either power rail or some bias level common for all affected parts/blocks is not correct or noisy. So these situations it’s often viable to use systematic approach, rather than trying to fix each item one by one.

In this case next operations were performed:

  • Install new Chemicon KMG 1000µF x 50V (C116,C117,C108) and Chemicon KZE low-ESR 470µF x 63V (C114,C115) capacitors. KZE caps are bigger so little lead forming and careful installation was performed. C106 was replaced earlier with KMG cap.

  • Install correct avalance BYD diode on CR111 location.

  • Desolder U507, U508, U509 and clean surface/pads in this area with clean acetone. After cleaning – solder U507,U509 back, and replace U508 to new 74LS00 as it’s just jellybean part and easier to put new in, than testing if original is ok.
  • Replace R526 (100R 1206), R280 (10K 1206) and R520 (330R 1206) resistors with new ones, as original resistors had heavy corrosion from electrolyte.

  • Replace U519 (donor taken from another unit analog board)

  • Replace U516 to ADA4898-1YRDZ opamp. Since this opamp have exposed pad there was capton mylar film placed under package to avoid short possibility to via under footprint.

One of other units with similar 4xx.x errors had bad U525 and U520. Diagnostics on that meter revealed function issues – DCI measurement on ranges other than 2A was throwing some incorrect readings even with short on inputs, ACV was showing fixed 480VAC on 750V range (with no input signal/short) and overflow on lower ranges and ACI currents.

Issues fixed after U525 and 4.3V zener replacement. Original LTC1050 opamp had leaky resistance 145Ω between V+ and V- rails making low-power floating supply circuitry unbalanced.

Other meter had bad LT1223 current mode operational amplifier, driving FET full-wave rectifier bridge to -12V, and screwing everything. Same dianosis errors and ACV,DCI and ACI function failures were observed. Removed U520 and powered meter without it, supply voltages on U525 become normal.

Now assemble unit.

After successful power on, we can run self-test again.

No erros detected, very well. If there are errors, meter showing “*” symbol after phrase All tests complete.
I had selftest running in loops for an hour, no issues were detected.

Unit 2 Calibration

Calibration manual states next about user-level comprehensive calibration procedure:

The comprehensive calibration procedure calibrates DCV, DCI (except for the 2A range), 2-wire ohm, and 4-wire ohm functions.
At the end of the DC calibration procedure, AC self-calibration is performed to complete the calibration process.

Also there is note that full low-level calibration would be required if unit is repaired:

If the unit has been repaired, you should perform the low-level calibration procedure explained in paragraph 2.10.

I did not have suitable sources for it, so for now skipped low level calibration. Most of it is for AC performance.

Comprehensive user-level calibration

Since we replaced lot of critical analog components, all functions require recalibration.
Let’s try run user-level comprehensive calibration first, to see if what can be calibrated at this level.
As performance check reference setup we will use next equipment:

Instrument Detail Last calibration / manufactured Calibration step S/N
Keithley 2002 DMM 2008 DCV, 2W-ohm, 4W-ohm, DCI, ACV, ACI 1167961
Keithley 2001 DMM 05 February 2014, +24.3°C DCV, 2W-ohm, 4W-ohm, DCI, ACV, ACI 0544257
Vishay VHP202Z 1MEG 0.01% Custom VPG 321217 VHA518-11 1ppm/°C 0.1W resistor 2W-ohm
Vishay VHP101 19K 0.005% Custom VPG 321217 VHP101 matched/tracked 0.3ppm/°C 2W-ohm, 4W-ohm
Vishay VHP202Z 10K 0.01% Y607110K0000T9L high-stability precision resistor 2W-ohm, 4W-ohm
Vishay VHP101 1.9K 0.005% Custom VPG 321217 VHP101 matched/tracked 0.3ppm/°C 2W-ohm, 4W-ohm
Vishay VHP101 95K 0.005% Custom VPG 321217 VHP101 matched/tracked 0.3ppm/°C 2W-ohm, 4W-ohm
Vishay VHP202Z 190R 0.005% Y6071190R000V0L high-stability precision resistor 2W-ohm, 4W-ohm
Vishay VHP202Z 19R 0.05% Y607119R0000A0L high-stability precision resistor 2W-ohm, 4W-ohm
Vishay CSNG 3c 2.5R resistor Custom VPG 321215 CSNG 0.5ppm/°C, 9W, 1.6A 2W-ohm, 4W-ohm
xDevs.com KX LTZ1000A REF Ultra-zener high-stability voltage reference 7V DCV
xDevs.com KX LTZ1000A REF As above, divided to 1.9V DCV
HP 33120A ARB 20ppm ARB ACV, Low-level AC cal US34006542
EDC MV106 DC Voltage standard Calibrated in February 2014 from K2001 DCV up to 10V

Custom VHP101 resistors are matched between within 0.01%, with 0.5 ppm/°C tracking tempco on best effort.
Custom resistors with VPG numbers were ordered with in-house PMO stabilization procedure. You don’t want to know a price for these :).

User-level calibration consist of these steps:

  • Low-thermal short calibration step.
  • +2V DC calibration step.
  • +20V DC calibration step.
  • 20kΩ calibration step.
  • 1MΩ alibration step.
  • Open circuit calibration step.
  • Calculate DC cal constants.

Low-level calibration

Low level calibration have lot more steps, mainly for AC function, as well as DCI,ACI steps.

  • 20V AC at 1kHz step.
  • 20V AC at 30kHz step.
  • 200V AC at 1kHz step.
  • 200V AC at 30kHz
  • 1.5V AC at 1kHz step.
  • 0.2V AC at 1kHz step.
  • 5mV AC at 100kHz step.
  • 0.5mV AC at 1kHz step.
  • +2V DC step.
  • -2V DC step.
  • 0V DC step.
  • +20mA AC at 1kHz step.
  • +0.2A DC step.
  • +2A DC step.
  • 2V AC at 1Hz step.

20V AC at 1kHz step setup

Ambient temperature: +24.0 °C
Here we use *20 bipolar amplifier to output 20 VAC sine wave at 1kHz. Source 1.000 VAC provided from HP 3245A, verified by HP 3458A.
Voltage level source is calibrated Keithley 2400 SMU.
Setup was allowed to stabilize thermally for 10 minutes before performing calibration step.

Details TBD..

20V AC at 30kHz step setup

Ambient temperature: +24.0 °C
Here we use *20 bipolar amplifier to output 20 VAC sine wave at 30kHz. Source 1.000 VAC provided from HP 3245A, verified by HP 3458A.
Voltage level source is calibrated Keithley 2400 SMU.
Setup was allowed to stabilize thermally for 10 minutes before performing calibration step.

Details TBD..

200V AC at 1kHz step setup

Ambient temperature: +24.0 °C
Here we use *20 bipolar amplifier to output 200 VAC sine wave at 1kHz. Source 10.000 VAC provided from 33120A, verified by calibrated Keithley 2001 DMM.
Voltage level source is calibrated Keithley 2400 SMU.
Setup was allowed to stabilize thermally for 10 minutes before performing calibration step.

Details TBD..

200V AC at 30kHz setup

Ambient temperature: +24.0 °C
Here we use *20 bipolar amplifier to output 200 VAC sine wave at 30kHz. Source 10.000 VAC provided from 33120A, verified by calibrated Keithley 2001 DMM.
Voltage level source is calibrated Keithley 2400 SMU.
Setup was allowed to stabilize thermally for 10 minutes before performing calibration step.

Details TBD..

1.5V AC at 1kHz step setup

Ambient temperature: +24.0 °C
Here we use output 1.5 VAC sinewave at 1kHz from HP 3245A, verified by HP 3458A.
Setup was allowed to stabilize thermally for 10 minutes before performing calibration step.

Details TBD..

0.2V AC at 1kHz step setup

Ambient temperature: +24.0 °C
Here we use output 0.2 VAC sinewave at 1kHz from HP 3245A, verified by HP 3458A.
Setup was allowed to stabilize thermally for 10 minutes before performing calibration step.

Details TBD..

5mV AC at 100kHz step setup

Ambient temperature: +24.0 °C
Here we use divided output 1.000 VAC sinewave at 100kHz to 5 mVAC voltage from HP 3245A, verified by HP 3458A.
Setup was allowed to stabilize thermally for 10 minutes before performing calibration step.

Details TBD..

0.5mV AC at 1kHz step setup

Ambient temperature: +24.0 °C
Here we use divided output 1.000 VAC sinewave at 1kHz to 0.5 mVAC voltage from HP 3245A, verified by HP 3458A.
Setup was allowed to stabilize thermally for 10 minutes before performing calibration step.

Details TBD..

+2V DC step setup

Ambient temperature: +24.0 °C
Here we use divided output 2.00000 VDC voltage from battery-powered KX reference (LTZ1000-based) connected to DMM in normal polarity.
Setup was allowed to stabilize thermally for 10 minutes before performing calibration step.

Details TBD..

-2V DC step setup

Ambient temperature: +24.0 °C
Here we use divided output 2.00000 VDC voltage from battery-powered KX reference (LTZ1000-based) connected to DMM in reverse polarity.
Setup was allowed to stabilize thermally for 10 minutes before performing calibration step.

Details TBD..

0V DC step setup

Ambient temperature: +24.0 °C
For this step input terminals were shorted with 4W-star short PCB.
Setup was allowed to stabilize thermally for 10 minutes before performing calibration step.

Details TBD..

+20mA AC at 1kHz step setup

For this step will use HP 3245A, verified by HP 3458A.

Details TBD..

+0.2A DC step setup

Ambient temperature: +24.0 °C
For this setup will use Keithley 2400 freshly calibrated to HP 3458A.
Current measured also with calibrated Keithley 2001. Setup was allowed to stabilize thermally for 30 minutes before performing calibration step.

Details TBD..

+2A DC step setup

Ambient temperature: +24.0 °C
For this setup will use 2000.000 mA current MOSFET source, with 2 x Vishay PG CSNG 2.5 ohm resistor (9W, 0.5ppm/°C, 1.6A rating).
Current measured with calibrated Keithley 2001. Current source circuit controlled from LTZ1000A reference VREF fixed voltage output.
Trim for current level was performed after reaching system thermal equilibrium.
Setup was allowed to stabilize thermally for 30 minutes before performing calibration step.

Details TBD..

2V AC at 1Hz step setup

HP 3245A, verified by HP 3458A. Output voltage also measured by calibrated Keithley 2001.

Unit #3 repair

Unit #3 calibration section

Unit #7 repair

This section is still work in progress, sorry.

Same issue with bad old capacitors:

Don’t wait for your capacitors become this old, replace them each 5-7 years even if they look OK from outside.

New capacitors (Chemicon) ready to install:

Let’s take a look on analog board parts in this unit:

Voltage reference is old aged LM399, OHM current source resistors are Vishay foil unlike other unit which had PTF56 resistors.

Digital board

Unit #7 calibration section

Infrared thermal images

Here are thermal images of two units, K0 (faulty) and K1 (good) for comparison. All images were done using Fluke Ti32 imager.


Image 1: Back analog PCB, both shots faulty unit


Image 1: R286 comparison, left is faulty unit, right is good unit


Image 1: Top w/o ADC module, left is faulty unit, right is good unit


Image 1: Top images, both faulty unit


Image 1: Top images, both good unit


Image 1: LM399 voltage reference circuitry, left – bad unit, right – good


Image 1: VREF on bad unit, rear side of ADC module


Image 1: U324 area, faulty at left, good at right side

Memory options

Since we likely fixed unit, worth maximize it’s features as well, by installing option MEM2 on digital board.
This is rather simple operation, already covered before in this article so will not put too much details in here.

Option Hardware Storage memory size Setup profiles
STD SRAM IC xxx only one, #0
MEM1 DS1230AB xxxx or xxxxx w/o timestamps up to 5, #0 to #4
MEM2 DS1245Y xxxx or xxxxx w/o timestamps up to 10, #0 to #9

Memory option require to have DS12xx programmed to valid contents, which at least have matching serial number to one in calibration EEPROM.
Otherwise storage function will not work correctly (reset contents on power off, don’t allow extra setup profiles to be saved).

Performance verification

Testing such high-performance and resolution meter require even higher performance standards and calibration sources, such as Fluke 5700A, which are usually available only metrology labs.Keithley recommends next gear to check meter specifications:

Function Reference standard
DCV Fluke 5700A

During writing of this article there is no access to such equipment, so this simplified performance check without calibrator source/standards is not as accurate as it should be. I used next test gear to test my meters:

Function
DCV 200mV xDevs.com KX LTZ1000 reference output divided
DCV 2V xDevs.com KX LTZ1000 reference output divided
DCV 20V xDevs.com KX LTZ1000 reference output direct and 20V ratio
DCV 200V Keithley 2400 calibrated in January 2014 by Tektronix
DCV 1000V 210V, Keithley 2400 calibrated in January 2014 by Tektronix

DCV Noise evaluation at shorted inputs

K2001 and many other DMMs were tested and compared in regards of zero voltage noise (short input) to get idea of overall measurement path noise. Methodology and test results are covered in detailed article here

Here’s some comparison data with Keithley 2001 & 2002’s, together with 3458A.

Setup:

  • No filters on Keithley meters.
    3458A – NPLC 100, AZER ON
    Keithley’s – NPLC 10, AZER SYNC, no filter
    Number after dash is just GPIB address.

All meters measure same KX LTZ1000 reference, direct 7V output, which is verified to have noise <0.2ppm. Stuff in the middle of the graph – just me, messing with other gear nearby, causing jumps due to knocking on cables and such.

Noise measurement result Voltage, µV Deviation ppm/range
VDC 3458A <1.2 peak-peak ~0.12 ppm/10V
VDC 2002-4 <2.5 peak-peak ~0.125 ppm/20V
VDC 2002-6 <2.5 peak-peak ~0.125 ppm/20V
VDC 2001-20 <15 peak-peak ~0.75 ppm/20V
VDC 2001-21 <22 peak-peak ~1.1 ppm/20V

Conclusion

Hope this helps to fix your Keithley Model 2001/2002. These instruments worth working on them, and once repaired and well cared able to deliver great performance and joy.

Author: Illya Tsemenko
Published: Sept. 17, 2014, 2:22 p.m.
Modified: Dec. 5, 2016, 12:08 p.m.

Comments