- Manuals and documentation
- Exterior and user interface
- Teardown and design evaluation
- Firmware backups
- Direct calibration vs HP 3458A and xDevs standards
- Calibration accuracy verification
- Calibration test for another unit
- Python application for testing
- Conclusion and summary
Work in progress, this article may contain errors and mistakes
This article is about japanese-made Advantest R6581T review, repair, calibration, service and performance tests. We slowly getting to complete coverage on all 8½-digit DMM/DVMs that were in production, with this R6581T worklog article. Here’s a complete list 8½-digit long-scale meters that we know about:
This is 19” rack-mount size system DMM, just a little smaller than competition Keysight 3458A or Datron 1281. It is however much lighter, due to use of aluminum chassis and shield sheets, instead of heavy-duty steel frame in HP meter.
|Advantest/ADCMT||R6581 (T/D)||DMM||No longer available for sale|
|Datron||1281/1271||DMM||No longer available for sale, obsoleted by Fluke|
|HP||3458A (+Fluke HFL version)||DMM||Current model|
|PREMA||6048||DMM||No longer available for sale|
|Solartron||7081||DMM||Also known as Guildline 9578. No longer available for sale|
R6581 is especially interesting, because it’s design was direct response to mighty HP 3458A, and lot of ideas and features are directly featured against the HP meter. In some parts this meter is better, while in others worse. Take a look on specification. Specification from HP 3458A (002 option) is shown in blue color.
There are three main versions of the Advantest R6581:
- R6581 is metrology type instrument with full DMM functionality, for DCV, DCI, 2/4-wire resistance, TrueRMS ACV and ACI.
- R6581D is field type instrument with DC only DMM functionality, for only DCV, DCI, 2/4-wire resistance, without AC module.
- R6581T is field type instrument with DC only DMM functionality, for only DCV, DCI, 2/4-wire resistance, without AC module.
xDevs.com team got hands on R6581T version of the unit, which is equivalent to R6581, but without AC TrueRMS option board.
|DC Voltage||100mV||1 V||10 V||100V||1000 V|
|Accuracy (1y), Rdg ppm + Range ppm (HP3458A/002)||7+10 (5+3)||6+1 (4+0.3)||5+0.2 (4+0.05)||7+1 (6+0.3)||7+0.2 (6+0.1)|
|Temperature coefficient, ppm||1.2+1||1+0.1||0.5+0.01||1.2+0.1||1.2+0.01|
|RMS noise at NPLC100, Range ppm||0.1||0.05||0.05||0.05||0.05|
|Input resistance, Ω||>1010||>1010||>1010||106 ±1%||106 ±1%|
|AC Voltage (Model R6581 only)||10mV||100mV||1V||10 V||100V||1000 V|
|Best accuracy (1y), Rdg ppm + Range ppm (HP3458A)||300+2500 (200+110)||200+200 (70+20)||150+1500 (200+20)||500+200 (400+20)|
|Temperature coefficient, ppm||10% spec\°C||10% spec\°C|
|Bandwidth frequency (max accuracy)||1 Hz to 2MHz (100Hz to 10kHz)|
|Resistance 4W||10 Ω||100 Ω||1 KΩ||10 KΩ||100 KΩ||1 MΩ||10 MΩ||100 MΩ||1 GΩ|
|Source current (High power mode)||10 mA||1 mA||100 µA||10 µA||1 µA||100 nA||10 nA|
|Accuracy (1y), Rdg ppm + Range ppm (HP3458A)||14+20 (15+5)||12+5 (12+5)||10+5 (10+0.5)||8+0.5 (10+0.5)||14+1 (15+2)||50+5 (50+10)||500+10 (500+10)||5000+10 (5000+10)|
|Source current (Low power mode)||10 mA||1 mA||100 µA||10 µA||1 µA||100 nA||10 nA||10 nA|
|Accuracy (1y), Rdg ppm + Range ppm||14+20||14+20||10+5||12+5||14+5||50+5||500+50||5000+50||5000+10|
|Temperature coefficient, ppm||3+3||3+1||2+1||2+0.1||20+0.1||100+1||1000+1|
|RMS noise at NPLC100, Range ppm||0.1|
|DC Current||100 nA||1 µA||10 µA||100 µA||1 mA||10 mA||100 mA||1 A|
|Accuracy (1y), Rdg ppm + Range ppm (HP3458A)||25+400 (30+400)||20+40 (20+40)||20+20 (20+10)||20+20 (20+8)||20+20 (20+5)||20+20 (20+5)||30+20 (35+5)||110+20 (110+10)|
|Temperature coefficient, ppm (HP3458A)||10+200 (10+200)||10+20 (2+20)||2+4 (10+4)||5+3 (10+3)||5+2 (10+2)||10+2 (10+2)||20+2 (25+2)||20+3 (25+3)|
|Shunt resistance (burden voltage)||1010 kΩ(55mV)||105 kΩ(45mV)||10.1 kΩ(55mV)||1.01 kΩ(75mV)||102 Ω(100mV)||12 Ω(100mV)||3 Ω(250mV)||2 Ω (<1.5V)|
|RMS noise at NPLC100, Range ppm||2|
|AC Current (Model R6581 only)||100 µA||1 mA||10 mA||100 mA||1 A|
|Best accuracy (1y), Rdg ppm + Range ppm||200+200||800+200|
Redistribution and use of this article, any part of it or/and 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 (/fix/r6581t/) and the following disclaimer.
- Redistributions of files in binary form must reproduce the above copyright notice, this list of conditions, link to this page (/fix/r6581t/), 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 Advantest 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 Advantest/ADCMT instruments repairs or provide extra information, you can do so following these simple instructions
Manuals and documentation
The only available official manual data is in japanese language, but thanks to modern OCR and online translators, we can get most of important information about the meter.
There is no service manual or schematics available for this unit.
However great deal of useful information can be found in this thread at EEVBlog.com forums
Exterior and user interface
Teardown and design evaluation
Digital and interface PCBA
Master DC Voltage reference PCBA
Inguard power supply PCBA
Front panel PCBA
Troubleshooting and repair
h2 Firmware backups
TL866CS programmer with PLCC socket adapter was used to read back all firmware ROMs. Provided for use AS IS.
Firmware dumps received from internets:
Direct calibration vs HP 3458A and xDevs standards
External calibration for the Advantest R6581T was performed with help of xDevs’s primary standards:
- +10.00000 DC Voltage standard, 792X prototype build around 792A Power Pack and our own LTZ1000A-based buffered reference.
- 9999.9747 Ω ±0.33 ppm Fluke SL935, which was calibrated November 28, 2017.
Same standards were used to calibrate Fluke 5700A which was used as reference source for all ranges and functions verifications. Calibrations of both MFC and meter were performed on same day, so long-term drift can be ignored and only relative accuracy specifications are used.
Calibration accuracy verification
This section covering non-factory calibration procedure and testing of 24-hour stability performance on the meter. This is non-traceable calibration for education purpose only. By no means using even high-end calibrator can provide legally valid calibration certification.
To run some performance numbers on Fluke 5700A, let’s connect one to Fluke 5700A working standard multi-function calibrator and use cal.equipment’s calkit Python application to crunch some numbers.
Image 66: Advantest R6581T 8½-digit DMM with Fluke 5700A MFC during the calibration.
Reference source value = (Low limit + Hi limit) / 2;
DCV Function performance test
Here is automated performance verification test data before the calibration adjustment.
Before calibration adjustment, AS RECEIVED condition. Last calibration date or history of the unit is unknown.
|DCV Test||R6581T Data||Source uncertainty||Low Limit||Hi limit||Measured||24hr spec||Result|
|0.1 VDC (0.1 Range)|
Table 5: R6581T fresh after external calibration
As we can see, meter cannot meet it’s 24 hour specification for 3V and 30V ranges, but actually not that much off. 300 mVDC range was not tested due to bug in control code.
Resistance function test (4-wire with offset compensation enabled)
|OHM Test||R6581T Data||Source uncertainty||Low Limit||Hi limit||Measured||24hr spec||Result|
Table 8: R6581T performance test, as returned after calibration adjustment, 4-wire resistance
Test on highest range with 1 GΩ standard was performed using external standard, as 5700A cannot source 1 GΩ.
Calibration test for another unit
EEVBlog member Ramon contacted me regarding improving uncertainty and measurement ability for his DIY resistance wire tests and projects. He bought another R6581T locally, and we had short meetup to perform calibration accuracy check and get some data comparison going.
792X prototype 10V reference as DCV source, Keithley 2002 #2 as reference meter were used to do checks. Together with reverse verification to my primary HPAK3458A meters, this setup allow to verify performance of R6581 with relatively good uncertainty.
Brief teardown of his Advantest was also performed, with photographs presented below.
Transformers are all good and happy, overall unit is very clean and tidy.
Firmware version of this unit is very same to our own, so there was no action taken to back it up.
Unlike xDevs’s unit, in this one battery definately does not like factory-made installation, so it was replaced at some point. It still has juice and measured over 3V, so no problem here, for now.
This is quite interesting instrument with it’s own quirks and features.
Definately a grab if you see one at low price under few hundred USD.
Here is small table of featureset against Keysight 3458A and Keithley 2002:
|Feature||Advantest 6581T||HP 3458A||Keithley 2002|
|Performance||Good||Excellent||Good to excellent|
|Input terminal type||Non-standard banana jacks||5-way binding posts||Banana jacks|
|ADC DCV Linearity||Okay||Best on the market||Good|
|Ratio measurement modes||Yes, front/rear||Yes, force/sense inputs||No|
|Guard input||Yes, floating or tied to LO||No|
|Artefact calibration||Full, with 4W Zero, 10VDC and 10 kΩ only||AC only, need MFC fo DC/Resistance|
|Front/rear switching||Electronic, using relays. Front panel/GPIB||Mechanical switch only|
|Internal temp sensor||Yes|
|Display||Dual-line VFD||Single-line VFD 16×1||Dual-line VFD|
|I/O||GPIB only, SCPI||GPIB only, custom protocol||GPIB only, SCPI|
|Expansion options||Analog output, Scanner cards||None||Scanner cards, nanovolt Keithley 1801 preamp kit|
|Cooling||Active cooling with fan and airducts|
|Form-factor, weight||2U 19”, light||2U 19” heavy||half-19” rack, light|
Also some generic parts were ordered from DigiKey store, such as DC fan, capacitors. Here’s order list (there were some other parts for other projects, not shown here, to make up for >100$ bill, as then it’s eligible for free shipping to Taiwan :).
|Quantity||Part Number||Description||Extended Price USD|
|1||565-1846-ND||CAP ALUM 680UF 20% 50V RADIAL||2.46|
Table 22: First Digikey order for parts
Total repair cost in parts as for today: $1312.64 USD
|02/26/2018||Received unit, initial teardown||1 hour|
|02/28/2018||Repair evaluation, testing Outguard boards and PSU, fix VFD panel operation||3 hours|
|03/02/2018||Ordered caps and from DigiKey||1 hour|
|03/07/2018||Received parts, replace all electrolytic caps||2 hours|
|Taking photos, writing up article sections and posts||5 hours|
Table 23: Time worklog summary
Total man-hours spent on this project around 24 hours, give or take few. Was it worth doing? That’s up to you, so please leave your comments, so I can do more stuff like this.
Equipment used during restoration project:
- Soldering gear, ERSA I-CON station
- MiniPro TL866CS programmer to read/write ROMs/RAMs
- Keithley 2002 DMM (calibration 2007)
- HP 3245A
- Fluke 5700A multi-function calibrator (calibrated within 30 days)
- Fluke SL935 resistance standard (calibrated November 2017, to <1ppm accuracy)
- 792× 10V reference prototype (calibrated February 2018, to <2ppm accuracy)
- Nikon D3 for timelapse shooting
- Nikon D800 with 28-70/2.8D, Sigma 150/2.8 Macro, 35/1.8G DX lenses, tripod
- 1L of IPA and alcohol to clean chassis and parts
- Google – used lots of it!
I would like to express my big respect to “Mickle.T” EEVBlog member for his time and efforts on reverse engineering R6581 design in great detail. This article would not be possible without you. Спасибо за проделанную работу!
Discussion about this article and related stuff is welcome in comment section or at our own IRC chat server: irc.xdevs.com (standard port 6667, channel: #xDevs.com). Web-interface for access mirrored on this page.
Modified: Jan. 28, 2020, 11:51 p.m.