Page 2 of 2

Re: Fluke 8050A Display Upgrade

Posted: Sun Oct 24, 2021 2:06 pm
by mediatechnology
This OLED is a little bigger, 1.3" diagonal and would fit the smaller bezel opening: ... B08FR1PW5G

The Fluke 8050As small bezel is 2" diagonal and the big opening 2.75".

In the "Oops I did it again department:"

I just scored a "for parts" 8050A on eBay which looks to be in pristine shape but won't power up.
It's a battery-powered 8050A and the seller wrote that sometimes on power up the display would briefly flash than go off.
That's a classic bad NiCd symptom.
It may just happen to have a good LCD.
I needed a clean case with unbroken handle and this one also had the vinyl carrying case.
For $20 it wasn't much of a gamble.

I mentioned earlier that on the display board the F2 switches had long plungers. (Relative and Power)
I theorized that was for voltage standoff but the range switches on the boards below it are normal length.
If I do a complete drop-in board the switches on the original may not have to be re-used.

Re: Fluke 8050A Display Upgrade

Posted: Wed Oct 27, 2021 6:12 am
by mediatechnology
I have a MAX7219 DIP in my shopping cart at Mouser along with some common cathode 0.36" LEDs.

Reading through some of the Amazon reviews I found that the OLED module wasn't particularly reliable.
My instincts tell me that any advanced display is going to eventually present the same problem as the original LCD.

Most of the serial LED 7 segment modules are 4 digit. I need 4-1/2 to display -1.999.
On one of the mods the "-" symbol was segment G with the right-hand portion of the segment shortened since "+/-1" digits are not made.
The mod requires 5 display digits plus an additional one for the 4 discrete LEDs.
The MAX7219 has a mode where digit 0 can be individually-addressed and digits 1 and above use the 7 segment character set.

I'm still tempted to go old-school and demux 3 pins to drive 5 digits using a 4051.
I'll have to use outboard discrete drivers.
In this day avoiding sole-sourced parts seems like a good idea.

Whatever I use to drive the display I still think I'm going to take all 20 pins off the display board to the driver.
The connector becomes the right angle mounting bracket and the large number of pins provide rigidity.

Re: Fluke 8050A Display Upgrade

Posted: Thu Oct 28, 2021 3:30 pm
by mediatechnology
Well Fluke 8050A number 5 is up and running: All it needed was new batteries.

Not bad for $20, shipping, and about $20 in batteries.
Almost perfect LCD.
All functions work.
Calibration looks to be good when compared to the other meters.


I should probably just stop now... :lol:

Re: Fluke 8050A Display Upgrade

Posted: Fri Oct 29, 2021 7:15 am
by mediatechnology
With the newest Fluke 8050A I got a vinyl carrying case with strap, a full manual, and the Army manual.

The Army manual has completely different schematics.
I already had an 8050A manual and needed to scan a better copy than the one I've posted here snipped from a pdf.

This Fluke 8050A has the "-01" battery-powered option.
While initially troubleshooting the new meter with a DC supply I measured the average battery current to be in the range of 120-150 mA.
In the 8050 the NiCd stack is the 5V supply.
There is no regulator: The battery connects - through the power switch - directly to the MK3870's Vcc and Ground pins.
When powered the NiCds act as a shunt Zener.
The switching regulator hangs off the NiCd stack.
Without good batteries the 8050A just won't work.

I already had a new set of NiCds on hand and am ordering a another set of 4 from for a third meter that will need them.
I use a 2500 mA/hr Dantona C-2500WT which is a standard "C" NiCd with solder tabs.

The newer 2500 mA batteries may be able to supply the extra LED and Arduino Nano current but the question remains is can the charger circuit supply enough extra current to maintain charge when it's operated on AC.
That power comes from a transformer-based supply.

My MAX7219 and common cathode LEDs should be in today.

Re: Fluke 8050A Display Upgrade

Posted: Sun Oct 31, 2021 7:16 am
by mediatechnology
I've been staring at the schematic of the 8050A again and wanted to correct an earlier error.

The battery-powered version has the NiCd stack sitting between the -5V and -10V rails.
The MK3870 MPU sits between the -5V rail and 0V.
The 3870 is powered by the switcher - it's Vee pin is connected directly to the NiCd + terminal.

I think I've figured out why the CPU has its Vee (Vss, Ground) connection to -5V.

The 4016 CMOS switches used in the A/D and front-end are connected to the -5V and +6V supplies.
To avoid level translation from a +5V-powered MPU they chose to make logic 0 -5V.
What confused me is that the CPU can then only swing from -5V to 0V: The 4016's need -5V to +6V.

There are pull-ups from the 4016 control lines to +6V suggesting that the MK3870 has >Vcc tolerant open collector or open drain outputs.
I checked the MK3870 F8 datasheet and it can be masked-programmed for open drain outputs.

The ST0-ST4 lines also appear to be open-drain: There is a 70K resistor pack on the display board that pulls the outputs "up" to Ground.
That explains the RC slope I see on ST0-ST4 on the inactive rising edge.

The HV WXYZ BP and DP data lines appear to be totem-pole outputs without pull-ups.

Budget Metrology

Posted: Mon Nov 01, 2021 5:33 pm
by mediatechnology
I wanted to know that the Fluke 8050A "Tower of Measurement" was accurate.

Since I don't own any calibrated voltage sources that becomes a problem...

I don't need to know that they're traceable but are they even in the ballpark?

I remembered a National Semiconductor Voltage Reference sample kit I had and never used.
I decided to crack it open and see if I had anything remotely accurate.



The LH0070H "-0" high accuracy caught my eye. It's 10V 0.03% typical 0.1% maximum at 25°C.
10.000V +/- 0.003V typical
Close enough to know whether the meters are close or not. :lol:

How do the Fluke 8050As measure?



I'll take that!
Now it could be that they're all wrong but its a start.
I decided to check it against all 5 meters.
Three meters were 10.000 V +/-1 count.
One was 10.002 V.

I also tried the second LH0070H and it was within 1 count of the first.
You can combine several LH0070 outputs through low-value resistors and get a statistical reference.
I didn't try that since they were within 1 count of each other.

The final meter was 10.03 V which I calibrated.
The calibration procedure calls for 190V on the 200V scale with the 20V scale dependent on the 200V range adjustment.
Since I had 10V I first set the cal correct on the 200V scale and then switched to the 20V scale and adjusted for exactly 10.000 V.
That didn't affect the 200V reading.

I also knew that one meter didn't read AC volts accurately.
It was few percent off.
The rest of the meters were within 0.01 dB of each other so I calibrated the RMS detector adjustments to the "consensus" reading.

So I'm using a 40 year-old voltage reference to calibrate 40 year-old meters.
I could send one to Tucker Electronics to have it calibrated but what fun would that be?

Re: Fluke 8050A Display Upgrade

Posted: Mon Nov 01, 2021 6:08 pm
by flyboy71
Close enough for government work! Speaking of...... way back in the early 90s a company I worked for used those meters for testing on a government project. Whatever we were measuring had to be within a gnats behind tight spec so they literally stacked and paralleled those until one of them read the data within the spec so we could say it tested ok and record the serial number of the DMM on the data sheet for signoff.

Re: Fluke 8050A Display Upgrade

Posted: Thu Nov 04, 2021 2:42 pm
by mediatechnology
I scored yet another 8050A on eBay and received it yesterday.
LCD display was perfect and the calibration almost spot on.
The only issue it had - easily solved - was an intermittent 2A current fuse.

Just found this site:


They have a 10 V 0.002% reference for about $86 US.