Fluke introduced the 8020A in the late 1970s as its first handheld DMM. These 3½-digit, 2000-count models can still be found on the used market, and can still be very useable, if they’re in working condition. Unfortunately though, LCD problems are common on this 30+ year-old device.
This example was obtained rather inexpensively in a group of three derelict handheld meters. (See also: Elenco LCR Meter, Keithley 135 DMM) The LCD is obviously suffering from some malady that causes it to look dark. None of the segments can be seen at all.
In an attempt to find out exactly which part of the LCD has turned dark, I carefully removed the polarized reflective backing. This reveals that the glass is still clear, and the backing is responsible for the darkening. But even placing the glass over a new piece of polarizing film and testing it shows that none of the segments work, so this display cannot be recovered.
Getting a new LCD that exactly matches this Fluke part may be next to impossible. But this DMM is very basic and only needs 3½ digits with decimal points, a leading minus sign, and a low-battery annunciator. There are several readily-available generic 3½-digit LCDs that can possibly work.
I chose the Lumex LCD-S3X1C50TR/B because it has exactly the same dimensions as the original LCD, and has exactly the features needed with no unneeded segments. The original display has a colon and a plus sign, which are not used, but have to be connected to the backplane clock signal to prevent being accidentally activated.
The original LCD used a single elastomeric connector on the top edge for the 30 signals coming from the main PCB. The chosen replacement has a row of 20 0.1-inch pitch pins on both top and bottom edges. Therefore the plastic LCD support bracket will have to be milled to create relief areas for these pins. Great precision is not really required, but there are support posts on the back side that need to be left attached. This makes for a tight fit for the end pins on the back row.
I chose to use single strands of solid 30-gauge insulated wire-wrap wire. For a mechanically secure connection to the LCD pins, I used a manual wire-wrap tool to make a secure wrap around each the LCD pins used. Small pieces of heat-shrink tubing provide strain relief and prevent accidental shorts.
All of the wires are routed around to the front and threaded through a plastic strip that has been drilled with a row of 30 tiny holes at the same pitch as the PCB pads. This keeps the wires in order so that they can be carefully sorted out and lined up before being soldered to the pads.
Soldering was a fairly quick process. The PCB pads were given a generous coat of flux. Holding the wire on the pad with a soldering aid and touching with a wet soldering iron tip created reliable joints.
I checked and double-checked the wires against the connection diagram before soldering (and caught a couple of mistakes.) So after soldering, a power-up test reveals that the LCD is working correctly and there are no misconnections. Since this is the first time I’ve seen this particular DMM work, a quick calibration and function check is in order. As is often the case with old Fluke DMMs, this unit is still quite accurate.
In order to check the low battery annunciator, the meter was powered from a variable DC power supply. Surprisingly the low battery condition is signaled below 7.65V, which seems high. Especially since the meter appears to work all the way down to 4.00V, below which the display fades out. The average current drawn is about 2.4mA.
Thanks to user ‘lowimpedance’ and others for starting a discussion at the EEVBlog forum that provided inspiration for this project.
DMMCheck calibration check results for this Fluke 8020A:
|AC||5V||5.55||nominal=5.55 for non-TRMS|
|1mA||1.096||nominal=1.111 for non-TRMS|
|Ohms||100Ω||100.2||0.3 with leads shorted|