xDevs.com | Datron 1281 DMM repair and calibration worklog

Intro
Disclaimer
Manual references
Initial inspection and basic disassembly
Repair workflow, unit #1
Repair workflow, unit #2
Firmware
Custom FRAM adapters and EEPROM programmer adapter
Other possible issues
- Front panel damages
Tweaks and tricks
- DIY cable for rear CHA and CHB ports
Calibration
Performance verification
- DCV Noise evaluation at shorted inputs
Restoration summary

Intro

We all know Keithley, HP/Agilent/Keysight, Tektronix, LeCroy and Fluke. But back in days, less known T&M manufacturer Datron was also known for their high-performance DMM and calibrators lineup. Datron was later acquired by Wavetek in 1987 for $15.8M. Wavetek was still manufacturing original Datron DMMs and calibrators, and eventually sold it’s T&M division to Fluke in 2000 for $40M. As of today there is no service provided for obsolete lineup, but lot of knowledge and technology from Datron was used in Fluke’s high-performance products, such as 8508A 8½ DMM and obsoleted now 7001 DC standard.

Having all that said, 1271 and 1281 are still welcome instrument in many labs, with 8½-digit resolution and metrology grade performance. Common rival for these system multimeters is well known HP 3458A and Keithley 2002. These Datron DMMs are still going on secondary market, for prices range around USD $2000.

This article will cover two different unit, with second generation Wavetek-branded logo, produced somewhen in 1992. These were purchased as “used”. There are not many DMMs of this performance level on the market even today, so it will be interesting to fix and see what these are capable by today’s standards. Below is table of all 8½-digit DMMs for your reference

Manufacturer	Model	Type	Status	DCV accuracy	Resistance accuracy	DCI accuracy
Advantest/ADCMT	R6581D	DMM	No longer available for sale	±5ppm + ±0.2ppm	±8ppm + ±0.2ppm	±20ppm + ±20ppm
Datron	1281/1271	DMM	No longer available for sale	±3.1ppm	±6ppm	±25ppm
HP	3458A	DMM	Current model
Fluke	8508A	DMM	Current model
Keithley	2002	DMM	Current model
PREMA	6048	DMM	No longer available for sale	±5ppm + ± 1ppm	±7ppm + ±5ppm	±30ppm + ±10ppm
Solartron	7081	DMM	No longer available for sale
Transmille	8080	DMM	Current model		Transmille	8081	DMM/Electrometer	Current model
Transmille	8881	DMM	Current model		ZIP Nauchpribor KM300	KM300	DCV/Res/Calibrator

Table 1: 8½-digit DMMs

HP and Keithley we already looked at on pages of this site, so it’s time for Wavetek-Datron.

So what’s special about this DMM? Let’s take a look on few key specifications for Datron 1281. Detailed specifications are presented later in manuals section of this article.

Function	Ranges	Accuracy
DC Voltage	0 to ±1000V	Best 1 Year Specification: ±3.1 ppm
AC Voltage	100 mV to 1000V @ 40 Hz to 1 MHz	Best 1 Year Specification: ±60 ppm
DC Current	0 to ±2A	Best 1 Year Specification: ±25 ppm
AC Current	100 mA to 2A @ 10 Hz to 5 kHz	Best 1 Year Specification: ±200 ppm
Resistance	10Ω to 1 GΩ	Best 1 Year Specification: ±6 ppm

Table 2: Specifications brief for Model 1281

Unlike 3458A or 2002, Datron 1281 also has unique Ohms measurement feature, such as active ‘Ohms Guard’ terminal to guard out leakage current paths when measuring very high value resistors. And its ‘LoI’ mode can prevent excessive self-heating in PRTs. The 1281’s 2-input ratio function allows you to use it as a high performance automated bridge.

Also 1281 have enhanced possibilities for ratio measurements, as it’s rear panel have two channels for input signals to perform direct ratio measurements without external scan-cards or armature involved. This is a reason why some of calibration laboratories prefer using Datron 1281 or modern Fluke 8508A to HP 3458A.

Datron also have self-cal technology, with is targeted to compete with Autocal feature of HP 3458A.

Let’s see what we have to do to get units back to operational state. Our plan on this project is to perform component level repair, and perform brief study of instrument’s internal construction, as repair progress. There are no official man manufacturer service for this old DMMs anymore, so hopefully we will not need custom unobtainable components, as that can easily put repair into BER. Datron 1281 is still accepted in many calibration labs, and it’s calibration procedures are similar to other long-scale meters.

But if we get units in question repaired, we will be doing calibration in-house, using calibrated HP 3458A-002 and Fluke 732B DC voltage standard. As usual, all photos are clickable for high-resolution version.

There are different versions of 1281 with options keys, such as:

Option ID	Description
Option 10	True RMS AC Converter
Option 20	2-Wire & 4-Wire Resistance Converter
Option 30	Current Converter (must have option 20)
Option 50	10A Shunt
Option 70	Isolated Analog Output
Option 80	115V, 60Hz Line Operation
Option 90	Rack Mounting Kit

Table 3: 3458A options and versions

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:

Redistribution of article must retain the above copyright notice, this list of conditions, link to this page (/fix/d1281/) and the following disclaimer.
Redistribution of files in binary form must reproduce the above copyright notice, this list of conditions, link to this page (/fix/d1281/), 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 Fluke 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 add your experience regarding test instruments repairs or provide extra information, you can do so following these simple instructions

Manual references

Datron Model 1281/1271 Datasheet and specification

Datron 1281 Users manual, Issue 6, Feb 1999

Alternative User’s manual

Datron 1281 Calibration and service manual, Volume 1 and 2, Issue 1, July 1989

Datron 1281 Calibration and service manual with schematics, Volume 2, Issue 2, January 1989

Datron 1281 Calibration and service manual with schematics, Volume 2

There are also useful manuals for similar Datron DMM, Model 1271. Credits for storing these manuals goes to KO4BB. Here’s mirror of them:

Datron 1271 Schematics

Datron 1271 Service manual, Calibration and servicing

Datron 1271 User’s manual

DJVU-format files can be opened with freeware WinDjView from Djvu.org.

Initial inspection and basic disassembly

Sample photos:

Image 1: DMM front face

Image 2: DMM rear face

Display:

Its a custom DC405E2 from ITRON/Japan. Display size is 1×40 char, 5×7 DOT matrix, 5mm height. Itron does not have any stock for the display and do not have public documentation for it. But thankfully, we got and able to share full datasheet for DC405E2.

Image 3: Testing VFD display output and test connections

Image 4: DMM rear face

Teardown

Datron 1281 (and lesser brother 1271) using interesting arrangement for primary power supply with two big toroidal transformers. First transformer converts mains to low voltage supply for earth-referenced digital domain, and powers low-voltage PWM DC-DC around second transformer, with switching frequency in sync with mains.

Patent data

Secondary outputs of this low voltage transformer are rectified and regulated with usual LDOs to provide clean analog supply voltages. What is the benefit in such double conversion? This design approach targeted to reduce capacitive coupling and keep power consumption by linear power supplies constant, so temperature inside meter’s chassis does not change in different conditions.

We already know, that temperature changes are our main enemy in high precision voltage and current measurements, so Datron had to make extra effort to reduce instability of temperature contributors, instead of adding fans or heat sinks around critical components. Similar idea technology was also used in Fluke’s 7000 DC reference standard and recent Keithley’s 7½-digit DMM7510.

Datron 1281 and 1271 are the using unusual arrangmenet for main voltage reference. Model 1281 using two identical Linear LTZ1000 based modules, labeled Datron 400763-1, with custom hermetically sealed Vishay metal foil resistor network, while Model 1271 using LTZ1000 module and LM399 module to save cost. In out units refereces were enclosed in black epoxy, and would be troublesome to remove from PCB for teardown.

Repair workflow, unit #1

Unit was received as advertised. Meter was failing with error message 9005 – 00150C03 immediately after power on.

Service manual decode message for 9005 states : Serial Interface Fault.

While testing optocouplers with second Datron 1281, it was noticed that error codes were different on second meter if optocouplers were left out of circuit. Possibly one of the daughterboards was failing? A quick check of all cables to the main DC board showed one that was not connected and stuffed under the main board. Not good, as this shows someone was diagnosing an issue that required the board to be possibly removed from the frame.

After reconnection cable, this 9005 error code went away and first DCV readings were acquired. It looked good with the input shorted, however the OHMs readings were open. Another search for connector issues showed the input connector was hastily plugged in and had missed a pin on the DC board, pushing it aside. A quick reconnection and all the ranges appeared to respond as expected with the short on the input. A screw was also missing that held the connector in place.

A quick self-test gave what probably were the original errors, DCV ERROR 2182 DC 100mV range zero magnitude and NVRAM error. The errors are broken down based on the type of test. The Fast test failure code is explained in section 2.7.3 for True Zero Checks. It tests the mean value with the inputs shorted on the 100mV range.

Conditions are for test 2182 are : -250µV < Mean 100mV Zero < +250µV.

The NVRAM error is most likely due to a bad old battery, similar to what we already know from repairing old instruments like HP 3458A, scope CSA7404 and R&S SMT03. Good rule of thumb is to replace all batteries if instrument is older than 10 years.

Without further diagnosing of the error, the inguard and outguard supplies needed to be tested as the parts have datecodes from 1988 and the lithium backup battery is dated from 1987. Its voltage was measured at 3.7V which is pretty good for a nearly 30yr old battery.

The schematics for both the inguard and outguard supplies were scanned. Test points for the DC board were easily found as Datron did not use silkscreening on their boards and used unmasked copper for the TP numbers on the DC board. The Digital board, however, has no markings and does not match the component layout drawing in the service manual.

Neither boards have any REV markings on the top, but might be located on the bottom side.

Digital board

This PCBA have main meter’s brain, good old 16 bit Motorola HD68000P8 processor, manufactured by Hitachi. It’s indeed very popular chip, present in many instruments of those years, such as HP 3458A, Keithley 200x, etc. It is NMOS based 5V ASIC, specified to run at 8.0 MHz clock and able to address 16 MBytes of memory space.

There are multiple different versions of this PCBA for Datron 1281. Another kind of board is shown below:

Image X: Digital board, courtesy of www.amplifier.cd .

Lot of 7400 series logic can be found around processor, having glue function for various I/O interfaces and control signals. All DIP-packaged ICs are replaceable, with using nice collet-type sockets. Re-seating all chips would be a good idea, considering age of the instrument and unknown history.

Other ceramic ASIC is located in center of the board, interfacing GPIB interface controller TMS9914ANL.

But first things first, power supply checks. The outguard supplies located on this PCBA were measured once suitable test points were found:

Test point	Specification	Measured voltage	Ripple	Condition
	+5V	+4.99 VDC	0.32 V_PK-PK	OK
	+45V	+47.1 VDC	0.43 V_PK-PK	OK

The Digital board has four EPROMs that are stacked 2ea on a small piggyback boards. The part numbers show them to be Hitachi Semiconductor HN27512G-25, which are 512Kbit (64Kbit × 8-bit) UV or OTP 5V EPROMs. The firmware version appears to be 3.02 but will need to be verified through menus. The board plugs into a single socket and has an additional wire connecting to it marked A17 most likely for an additional address line to other memory ROMs.

The digital board is removed by taking out screws from the rear cover. The power supply and ribbon cables also need to be disconnected. The board and rear cover come out as one assembly.

(Picture showing which screws to remove)

All seven capacitors were changed with only one gotcha. Just like the DC board, one of the caps had a lead bent offset to line up with the PCB holes.

Index	Quantity	Part Number on Digikey	Manufacturer Part Number	Description
1	1	493-16250-1-ND	ULD1H100MDD1TD	CAP ALUM 10UF 20% 50V THRU HOLE
2	2	493-6802-ND	UHW1E222MHD6	CAP ALUM 2200UF 20% 25V RADIAL
3	4	493-1622-ND	UHE1H102MHD6	CAP ALUM 1000UF 20% 50V RADIAL
4	1	493-1675-ND	UHE2A221MHD6	CAP ALUM 220UF 20% 100V RADIAL
5	1	493-1647-ND	UHE1J471MHD6	CAP ALUM 470UF 20% 63V RADIAL
6	2	1189-1869-ND	25ZLH470MEFC10×12.5	CAP ALUM 470UF 20% 25V RADIAL
7	1	565-3939-ND	EKYB500ELL221MJ16S	CAP ALUM 220UF 20% 50V RADIAL
8	2	565-3964-ND	EKYB630ELL221MJ20S	CAP ALUM 220UF 20% 63V RADIAL
9	4	493-14725-1-ND	ULD1H101MPD1TD	CAP ALUM 100UF 20% 50V THRU HOLE
10	10	399-15737-ND	ESL105M100AC3AA	CAP ALUM 1UF 20% 100V THRU HOLE
11	1	P19617CT-ND	EEU-EB1J101BJ	CAP ALUM 100UF 20% 63V RADIAL

The lithium battery was replaced with a TL-5101P. Replacement of battery was performed with unit powered on and a battery powered soldering iron was used to avoid loosing calibration constants and SRAM data.

Here’s BOM list of parts used for repairs:

Digikey Part Number	Made by	Part Number	Specification	Customer Description	Quantity
439-1003-ND	Tadiran Batteries	TL-5101/P	3.6V ½AA	950mAh	1

DCV board assembly

After disconnecting lots of cables and wires, main DC board was released and removed. Might be worthy to take a note of cable locations before removing PCBA:

Released board:

This colorful PCBA have all critical components for DC and main voltage reference modules, covered by black plastic covers. Unusual amount of transformers and inductors differentiate Datron’s approach in multimeter design. We will look into those later on, after repair.

Let’s check power rails from primary supply. Capacitors which should be replaced are marked on schematics in blue.

A quick test of the inguard supplies on the DC board showed the following values:

Test point	Specification	Measured voltage	Ripple	Condition
TP901	+15V	+15.0 VDC	0.6 V_PK-PK	OK
TP902	-15V	-14.8 VDC	0.04 V_PK-PK	OK
TP903	+35V	+35.4 VDC	0.09 V_PK-PK	OK
TP904	-35V	-33.7 VDC	0.87 V_PK-PK	OK
TP905	+5V	+5.0 VDC	0.08 V_PK-PK	OK

All but two capacitors were replaced on the DC board (12 of 14). A couple of gotchas were found during the change out. The three 1000 µF caps had one lead bent offset to the bottom. It appears the mounting holes were too close together and did not match capacitor pin spacing. Another gotcha is that it is important to note which lead is offset as there are some components that will prevent it from going in properly. Two of the capacitors actually would touch each other and this is solved with the offset leads.

Another gotcha is that most of the four 100 µF capacitors leads are not soldered straight to the board. At least one positive lead is soldered to the pcb in through a pad but it is not usually shaped for the polarity (square-vs-round). The other leads are soldered straight to components with insulating sleeves. Two of these capacitors were relatively easy to remove by clipping them at the mating lead. The old lead was left on the part and the new lead was soldered next to it. There was concern that additional heat could have caused issues with the aging parts. The other two capacitors were surrounded by other parts preventing them from being changed easily. It was decided to continue troubleshooting at this moment and replacing them would have been done later.

The 470 µF cap was replaced with a higher rated voltage. The new part is as tall as the three 1000 µF caps. The lid has polymide tape directly above the three parts so the same precaution was added above the 100 µF cap.

The remaining capacitors had no additional gotchas.

An IET DE-5000 was used to measure the 1000 µF capacitors for value and ESR. The old parts were at least 20% below rated capacitance and the ESR was approx 1Ω @ 120Hz. In comparison, the new parts were all within 10% and the ESR was < 0.1 Ω.

Compressed air was used to clean the dust and debris from the DC PCB. The board was reinstalled in meter and the FAST SELF-TEST was performed several times over the period of several hours. All tests completed now without errors.

Here’s BOM list of parts used for repairs:

Digikey Part Number	Vendor	Part Number	Specification	Customer Description	Quantity
493-1647-ND	Nichicon	UHE1J471MHD6	470 µF/63V	10000H/105 °C 7.5mm/sp	2
493-1622-ND	Nichicon	UHE1H102MHD6	1000 µF/50V	10000H/105 °C 7.5mm/sp	4
493-6802-ND	Nichicon	UHW1E222MHD6	2200 µF/25V	10000H/105 °C 7.5mm/sp	2
732-9242-1-ND	Wurth Electronics	860040778011	270 µF/63V	10000H/105 °C 5mm/sp	1
493-6817-ND	Nichicon	UHW1H221MPD	220 µF/50V	10000H/105 °C 5mm/sp	1
732-9325-1-ND	Wurth Electronics	860240578010	470 µF/35V	10000H/105 °C 5mm/sp	1
732-9281-1-ND	Wurth Electronics	860240474005	100 µF/25V	8000H/105 °C 3.5mm/sp	4
493-14361-ND	Nichicon	UBT1J330MPD	33 µF/63V	2000H/125 °C 3.5mm/sp	2
732-8966-1-ND	Wurth Electronics	860020772005	1 µF/63V	2000H/105 °C 2mm/sp	4

Options

Now time for option boards, located on bottom side of instrument chassis

Option 10 – True RMS AC Converter

No issues were noted with the AC assy. It was inspected, photographed, and dust/debris were removed before it was reinstalled.

Option 20 – 2-Wire & 4-Wire Resistance Converter

After removal of Option 20 PCBA remnants of spider with it’s home was detected. That likely caused some excessive currents flow thru webs and poor bug, causing measurement issues.

No issues were noted with the Ohms assy. It was inspected, photographed, and dust/debris were removed before it was reinstalled.

Option 30 – Current Converter

Current shunt resistor network consist of separate set of resistors:

RefDes	Value	Specification	Manufacturer
R111	800 Ω	0.1%
R112	90 Ω	0.1%
R113	9 Ω	0.1%
R114	1.0 Ω	0.1%	Mann Components
R115	0.1 Ω	0.1%	Mann Components

Pair of teal colored precision 4-wire shunts are manufactured by Mann Components Ltd (Ayton Road, Wymondham). Mann Components was an English resistor manufacturer with sales of about $2M. It was bought by Vishay in early 1983.

During reassembly of the bottom half of the meter, it was discovered that the Current assembly had some difficulty with the three connectors at the front of the pcb. Looking at the wires an interesting issue was found. It appears that the two outside connectors were placed in the opposite location. The wires are a little short and there is some noticeable stress. Looking for other photos on the internet, a ’94 era 1281 had the connectors located in the right position.

Other repairs

The lithium battery was replaced with a TL-5101P. The power was left on and a battery powered soldering iron was used.

The power input filter was also replaced as a precaution. The old part is a Waycom WF 120-3/05 and it was replaced with a Schaffner FN9222-3-06. The ground cable is exposed so additional heat shrink was added to the P and N terminals.

The power supplies were remeasured. All supplies are close to as-found except the -35V supply. It was now measuring -35.6VDC with 0.13Vp-p noise. The +15V supply measured 15.1 with 0.22Vp-p.

Quick check after assembly, with short on input terminals:

Input DCV +10V test:

Repair workflow, unit #2

This second unit actually happens to be the the first Datron 1281 acquired. It was chosen to repair last repair due to the number of errors.
The NVRAM in this meter is the same as the first 1281 but the battery is a few years newer.

Repair procedure is very similar to previous box. This unit comes with Options 10, 20, 30 and 70, so it have full set of boards and modules to play with.

This unit does not have mount hardware for 19” rack (Option 80) like previous, but comes with foot to angle it on the bench for easier operation. It is still older version of Model 1281, still carry Datron Instruments logo, not Wavetek-Datron like on newer meters.

Model 1281 is rare DMM providing separate active resistance guard terminal, unlike HP 3458A or Keithley 2002. Fluke 8508A have combined Guard terminal, switchable between passive or active mode.

Rear panel is same as first unit, with all the usual connectors and without any fans or vent holes. Keylock to enable calibration mode on this unit is damaged by previous owner, likely by attempts of using wrong key. This is rather unusual today to see such hard keylock to protect calibration, but back then it was easy go-to solution.

Of course it’s just single pole switch and any dedicated calibration pirate cannot be stopped by simple lock, as photo above able to indicate well. ROM/RAM chips have same style paper-written labels with version/serial number.

The fast self-test returned the following errors including test limits from the service manual:

Error code	Limit min	Value description	Limit max
DCV 2182	-250uV	Mean 100mV Zero	+250uV
Ohms 2752	+0.01FR – 4%	Mean Magnitude	+0.01FR + 4%
DCI 2572	-50ppm of FS	Mean Magnitude	+50ppm of FS

The 2752 and 2572 test both use the Ohms 10mA current source. The current source was tested using Keithley Model 2002 as control meter. The 2002 showed a low noise source but the value was nowhere near specifications. Most likely the calibration constants are too far off.

It is hoped that the capacitor replacement will cure the DCV 2182 Error as it did on the first meter. The power supplies will need to be checked first, as usual with old equipment repairs.