- Manual references
- Initial photos and conditions
- Hardware differences due to Option HFL
- Power checks and initial diagnostics
- A1 DC/resistance board assembly repairs
- A2 board tests and repair
- Meter verification and calibration
- Comparisons to other xDevs.com lab gear
- Thermal infrared images
Hard to believe but it’s been six years since xDevs’s first rusty 3458A repair project . That “Golden grade” meter still going strong in lab, helping with daily measurements and calirations. Most use is done at main 10V DC range, capturing biggest benefit of 3458A DMM with excellent ADC performance and very low thermal coefficient errors.
This article will show a worklog for yet another 3458A, this time from Agilent manufacturing days. It’s not standard meter but the elusive Option HFL that was sold and maintained by Fluke back in early 2000s when they just gobbled up poor Wavetek instrument division but didn’t finished yet the transforming 1281 into ugly-looking Fluke 8508A. And to follow our best tradition, unit was received broken and non-functional. There is no fun in buying working equipment. No room for learning how to fix stuff! :)
As usual, all photos are clickable for high-resolution version.
Recently many threads pop up, with question “Is it worth to buy bad 3458A for $1000?. Well, simple answer is – No, it’s a big risk investment. If you asking yourself about this, then you are not ready for it. Buying broken or used 3458A for “good price” and fixing it might not be an easy way of getting good 8½ DMM, when multiple hidden costs considered. If unit stated as “untested/unknown/don’t know how to test” and not to have shown self-test result, it could have hardware failures. Paid cost for such a unit WITH proper repair can easily exceed cost of buying a fully tested unit. If unit’s photo shows it as powered up but you can see “ERR” on display, it also means there is fault detected. Keysight have standard 3458A repair + recalibration cost at $2660 USD. Calibration alone of fully functional unit is likely to be over $1000 USD. Also old units manufactured earlier than 2005, would have Dallas NVRAM batteries already near end of life, with risk of losing calibration data any day. So even if you buy meter for $1000, you can be easily be held back by another $2000-3000 to fix it and get properly calibrated.
There are also special versions of 3458A with options, such as:
|3458A-001||Extended memory option|
|3458A-002||Improved DC reference stability 4 ppm/year|
|3458A-H01||Special 1000 Vrms ac maximum input voltage|
|3458A-H52||US Air Force NSN|
|3458A-HFL||Fluke version 3458A/HFL with 2ppm/year stability and better resistance|
|3458A-OGC||Precision calibration, intended for metrology use only|
Table 2: 3458A options and versions
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 (/fix/hp3458a/) 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/hp3458a/), 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 HP/Agilent/Keysight 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 HP/Agilent/Keysight instruments repairs or provide extra information, you can do so following these simple instructions
There are service notes/engineering changes were published during years of 3458A’s lifecycle. If your unit is old, worth to check if any of them required to do.
Initial photos and conditions
Like a wizard with all-see crystal sphere, here are list (likely not complete) of issues present with the box.
- Dim VFD TBD
- Not working resistance function Fixed
- Not working correctly voltage function Fixed
- Busted mounting for front binding post terminals
- Large offsets and AC failures on all AC functions. Fixed
Power checks and initial diagnostics
Instrument was throwing “REFERENCE LOOP FAILURE: 201” codes and sometimes “Error 114: balanced rundown convergence” errors upon first power ups. Resistance function and AC voltage functions were not functional and DC voltage was errornous.
A1 DC/resistance board assembly repairs
A2 AC assembly troubleshooting and repairs
Most of the troubleshooting process was recorded during livestream. Beware, it is not edited or skipped footage, so it has nearly 7 hours of pondering around 3458A guts.
After spending 15 or so hours all AC voltage issues were narrowed down to one broken jelly-bean 3904 NPN transistor :)
Meter verification and calibration
SN18 ADC stability test
Comparisons to other xDevs.com lab gear
Thermal infrared images
Thermal analysis and understanding of hot spots and thermal gradients present in metrology-level instrument like HP/Agilent/Keysight 3458A is vital to achieve stable and repeatable ppm-level performance. Some of thermal images are captured using Fluke TiX560 320×480 imager.
Any second-hand broken 3458A buying decision should weight in probable cost of A3 A/D PCBA replacement. There are reasons why many of 3458As that end up on eBay have faulty boards or part mules. But in this Option HFL unit we happen to avoid the trouble with A3 swap and it was just relatively simple cheap component level repair.
|Dead Agilent 3458A||$1800||$100||eBay|
|Dallas/MAXIM NVRAMs 1 x DS1220, 2 x DS1230||$80||N/A||Maxim|
|3 x ADG441 mux ICs||$15||N/A||Digikey|
|1 x OP07 operational amplifier||$5||N/A||Digikey|
|4 x CD4094 shift-register IC||$2||N/A||Digikey|
Table 9: Project cost
Total restoration cost for this meter: $2003 USD. Not cheap by any measure, yet still in range of secondary-market value for 3458A.
|05/05/2021||Received unit, initial teardown and checks, repair for ohms functionality||6 hour|
|06/05/2021||Repair for DCV functionality, finished A1 repairs||5 hour|
|12/05/2021||Troubleshooting A2 issues||3 hour|
|13/05/2021||Test other boards with good donor A2, repair bad A2||15 hour|
|19/05/2021||A2 repaired, started calibrations and performance tests||12 hour|
|20/05/2021||Testing SN18 start||8 hour|
|26/05/2021||Started datalog or 10VDC base range to verify mid-term A3 stability||150 hour|
|Documentation, photos, calibrations, article write-up||12 hours|
Table 10: Timelog summary
I would like to express my gratitude and appreciation to supporters of this project: Todd Micaleff, Mr. T, Nikonoid and of course all our readers. Stay tuned, many more interesting ideas are going online at xDevs.com in coming months. 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. 27, 2023, 8:40 a.m.