EBAY: LED Display Module FLUKE 8840A/42A (DIY Kit) replace broken, damaged VFD

[AHOL]

This is an updated post.
The information corresponds to the modified module.
Only + 5V power supply is used.
The problem that was discussed up to 8 post is REDACTED. Fully working.


Hello All! That will be my first topic here.
Immediately I apologize for my English, I use Google translate.
I present to your attention LED DISPLAY MODULE FLUKE 8840A / 42A (DIY kit).





This story began a year ago, like many of you, I bought a FLUKE 8840A on eBay with a broken display.
The search for solutions to replace it did not lead to special results.
It was decided to develop our own LED module with a custom LED display.

ASSIGNMENT
This LED display module can be used to replace your broken, damaged, cracked, unusable, etc. VFD (itron CP2067A ) of FLUKE 8840A and 8842A multimeters. (FLUKE PART NO. 728873(FLUKE STOCK NO. 680843)).
This LED display module is intended for hobbyist simulation of various measurement devices, as well as for educational purposes, and provided AS IS as a DIY kit.
In addition, this display module can be used in electronics hobby projects, where a 5.5 seven-segment display with indication of additional characters is required.

DESCRIPTION
The LED display module is designed as close as possible to the original and does not require any modification to the multimeter. The module is mounted in place of the dismantled non-working VFD display.
The control scheme for one segment is shown in Fig. 1.



Fig.1

This module uses a custom-made, industrially produced 7-segment display.

BASIC TECHNICAL SPECIFICATION



BOM



FRONT PANEL DESIGN AND ALL LIGHT ON SEGMENT FEATURE



Pin assignment



Before shipping, each module is assembled (excluding the pin header) and tested. I guarantee a long service life provided that the consumer observes the conditions of transportation, storage and installation.
This kit has been tested with FLUKE 8840A and FLUKE 8842A multimeters.

INSTALLATION




















Photo of the installation in full resolution follow the link
https://drive.google.com/drive/folders/17eKHPnh537Mgt9A4NKM6ya5zAWMvPNiQ


I am the developer and sole producer of this module.
All other solutions on the market use the module purchased from me.
I only sell on eBay under a nickname:
grudn_2015


My shop on eBay:
https://www.ebay.com/str/aholstore

The price at which I sell the module (without shipping and taxes): - $34.4

The LED display module is sold here:
https://www.ebay.com/itm/155853129116


Thank you for your interest!
Good luck with your research and measurements!

[bdunham7]

That looks really good, but where does it derive the LED power from?

[AHOL]

Hello bdunham7! Saw yours display too, not a bad job either.

Driver VFD delivers current up to 40mA

[bdunham7]

OK, you run the LEDs directly from the VFD driver?  The problem I ran into is that the VFD is driven off of the +30V front end amplifier supply and when I tried to power the LEDs from it, the +30V power supply became overloaded and dropped out below regulation.  The problem is the power supply, not the VFD driver chips.  This will cause the meter to not function properly at certain inputs.   So I think you have a problem there.  I solved that problem by modifying the main board, but my main obstacle at the moment is getting the LED display itself made up.  If I had a complete LED  display, I'd have a complete solution and I can use 30mA (or more) LED current drive in my setup. 

[bdunham7]

Also, where do you get your ground (V_ss) since there is not a ground (Digital LO) connection to the display driver board?

[AHOL]

Yes, you are right, there is a voltage drop, but +30V is used to power the VFD driver and the U306.
But here is what the instruction requires in order to pass the calibration and work correctly:

To test load sources, set the 8842A to 20V and measure the voltages across TP301, TP302, and TP303. TP301 should be 6.3V (typical) above TP302 and TP303 should be 6.2V (typical) below TP302.

When the module is running, these parameters are normal.
And the multimeter passes a self-test.

Vss is the VFD power winding.

[AHOL]

Continuation of the post about power U306 (LT1012)

Aalmost the same and for AD547 who has it installed.

Data from spec:

[bdunham7]

Yes, you are right, there is a voltage drop, but +30V is used to power the VFD driver and the U306.
But here is what the instruction requires in order to pass the calibration and work correctly:

To test load sources, set the 8842A to 20V and measure the voltages across TP301, TP302, and TP303. TP301 should be 6.3V (typical) above TP302 and TP303 should be 6.2V (typical) below TP302.

When the module is running, these parameters are normal.
And the multimeter passes a self-test.

I'm sorry to say that I think you've made a grave error (big mistake), but you may get away with it for a while depending on exactly how much current your LEDs draw.

The +/- 30 volts are required because the circuit (picture attached) floats the supply voltage for U306 up and down with the output, so the + supply would vary up to as much as 26.3 volts or so.  Testing this without an input won't reveal a shortcoming in the +30V supply.  I found during testing that as little as 120mA of current would draw the system down well below 26.3 volts and 160mA would draw it down below 20 volts, where it can't possibly work correctly.  I just reconfirmed this and a 160mA load draws the +30V supply down to just below 20 volts and the meter is severely inaccurate on a 20V input, reading less than 16 volts.  However, with the +30V supply loaded like this it will still pass a self test (!) and the TP301/TP302/TP303 voltages all read as you state--provided there's no input.  So that's no guarantee that all is good.

I realize that everyone seems to want a 'drop in' replacement for the VFD, but I concluded a while back that it just isn't feasible.  Even if your LEDs just happen to not draw quite enough current to cause a gross reading error, you've still overloaded that section of the power supply, which may cause a burnout in the future.  Also, I just couldn't see deliberately causing a change from specification--and lots of noise-- in the regulated power supply to one of the most critical components in the meter.  I've attached 3 scope shots--just driving the VFD, driving just 3 LED segments (out of 8 ) and with a fixed 160mA load.

I've ordered a few of your units--if the LED array is good, it's probably worth it for that.  I see they are selling quickly!

[AHOL]

Dear forum users! First of all, I want to thank everyone who bought my display module.

I confirm that the problem with the drop of + 30V is present in my module.

There is an exit.

I previously made a prototype with  duable-key control and power supply from the VFD filament Voltage, but refused it, since the filament winding does not provide enough LED current,
and I did not want to supply two wires from a +5V multimeter, BUT THIS CANNOT be AVOIDED.










The problem did not appear to me since I used CCR NSI50010YT1G. When using it, the drop is about 1.5V, but the voltage is still unstable.

The module can still be used in other applications according to the stated specification, BUT NOT IN THE FLUKE MULTIMETER.

Within 2-3 weeks I will make new boards with LED display, and I will send new boards in the quantity corresponding
to your current order to all buyers of the module for FREE.

I apologize for the inconvenience caused, this was not done intentionally.


Now a corrected module with additional + 5V power supply from a multimeter is now on sale.

[bdunham7]

Yes, the problem with a drop-in solution is that there just isn't enough current available anywhere.  Let me share how my setup works in case it helps you.  Mine requires some work to the main board, but is 100% electrically solid and provides plenty of LED current--my issue has been getting the display physically right.  If you have a good, bright display that fits, the other problems can be solved.

So the problem is translating the all-positive VFD control signals to a positive segment control and a ground-side digit control.  What I did is reuse the existing control circuitry and replaced the digit-grid driver, U215, with a ULN2803A darlington array.  This takes the same inputs as the original NE594s but instead of connecting the outputs to V_cc, it connects them to V_ss.  I then took U217 and U218 and changed their V_cc from the +30V regulated supply to the unregulated ~10.5V input to the 5V regulator--at the junction of CR601, CR602, C601 and VR601.  I just nipped the terminal with a flush cutter and soldered on a wire--truly pro!  This circuit can supply plenty of current and since I now have 10+ volts to work with, there's enough to handle the drops across the ULN2803A, the NE594, a CCD LED driver and two blue LEDs in series where needed.  So my board is just LEDs and CCD drivers.  I had to use 12 drivers, because although I could run two in series for the polarity sign, the '1' digit has both segments on the 'A' circuit and since I had common-cathode displays, I had to run them in parallel.

Since your board has the ground-side translator drivers on it, I think it could be adapted to work properly with just two bodges required.  First, the ground side would have to be disconnected from the VFD filament and run to the DIGITAL LO terminal.  Second, the segment-drive NE594s U217 and U218 would have to be bodged over to the 10V unregulated (or pre-regulated) supply.  I had thought of making a small board to go in place of U215/U217/U218 to make it look neater, but the bodge works well as-is.

If you wanted to stay as close as possible to drop-in, you could put 19 or 20 drivers on the board and just run two extra wires to ground and the power supply.  I wouldn't use the +5V regulated supply though, as tempting as it is to make it an easy plug-in installation.  I'm pretty sure most people wanting to replace their display on these can do a simple solder connection.

[AHOL]

Thanks for the answer.
Yes, I also thought about replacing the driver chip that is pin compatible but switches the outputs to ground instead of V+, but it is fiddles with soldering the drivers, which I wouldn't want to do.

And using + 5V (two wires) is probably the smallest change in the circuit;)

Drop on the + 5V bus when connected module only 20mV.

Usually you can receive signals from the processor and make an LED module with additional functions. I even bought a suitable display, but the price will be comparable to the price of a used multimeter.

Maybe I'll do it at my leisure;)

By the way, did you receive a letter via eBay?

[bdunham7]

I did get the message. I'll wait to see what you come up with in the meantime.  I'm also interested to see how bright the display is--i had to work pretty hard to get good viewing angles and acceptable levels of brightness after the front bezel is installed.  Is there any way your display LED could be made in a high-intensity red version?

I understand that a kit will sell better if less or no soldering is required.  I was looking for a good repair method, not a kit design.  I have no problems hacking away at the main board and even the bodge wires don't bother me much--they are ugly but solid.

Also, just an idea--if you are going to go from the +5V or the unregulated, you could bump the current up on the CCDs to 30 or even 40mA if you are using all on-board drivers.

[AHOL]

Can you email me on eBay?

This may not be the email address that eBay showed me.

The viewing angles and brightness are excellent, you'll see.
By the way, in my design the font was larger, but a factory that has been producing LEDs for many decades advised me to ensure that this type of font is displayed correctly on such a workspace. Both engineers and ergonomists do not recommend red as the primary color, only for signaling messages.

The LEDs used in the module have a current of 12-15mA and a supply voltage of 2.9V. This was my condition to reduce the problems of lack of power on some windings of the FLUKE transformer.

Most importantly, the low current of the LED ensures a long lifespan.

And you can do anything, including a red or red ERROR, there would be money;)

[YetAnotherTechie]

I previously made a prototype with  duable-key control and power supply from the VFD box winding, but refused it, since the filament winding does not provide enough LED current,

Just curious, but isn't the filament winding able to a watt or two?

[AHOL]

My module has LEDs with an operating current of 15mA and a supply voltage of 2.9V.
And even this is not provided by the filament winding.
The maximum that i managed to get is 4-5 mA.

After all, it is necessary to provide Ipulse=Icc*q^0,714

q - duty cycle 8

[up8051]

Can you share information where did you order custom LED?
I'm looking for manufacturer of custom 14-segment display.

[AHOL]

Any LED manufacturer will be able to fulfill your order.
The only advice I can give is to look for manufacturers in Hong Kong or China, they have a flexible pricing policy.

[AHOL]

Hello everybody!

Thanks to everyone who bought the display.
They sent everyone a replacement for a new (working) one.
Thank you for understanding!

Today I posted a new batch for sale on ebay. Link in the first post, there is also information on assembly.

On request, I add a schematic.

[strawberry]

unfair business practice https://youtu.be/Mt9jOJNHa88

[AHOL]

Thanks to all buyers!

Updated listing link (First post).
https://www.ebay.com/itm/154522071837

I draw your attention to the need to use anti static protection during assembly. Transistors are very sensitive to static voltage.

[bikeNomad]

I've also been working on a Fluke 8842A replacement display, and haven't yet got a diffuser that works well enough.

Going to an offshore supplier would require a MOQ.

AHOL, how many LED displays did you have to order?

I used the VFD supply pins connected to diodes, a 220µF capacitor, and a simple adjustable LM317 linear regulator (which lets me vary the brightness). I'm driving the LEDs with a 5V-capable microcontroller (with series resistors between the 30V logic inputs and the microcontroller digital input pins to limit input current). LED series resistance is provided by the ~50Ω microcontroller outputs. I did run a wire to digital ground.

I'm using 0603 white LEDs that have a forward voltage of 3.1V maximum (at 5mA).

I haven't looked at the effect on the various supply voltages yet, but it seems to work pretty well.

From my simulation, it looks like the 5V supply is providing about 100mA RMS through the transformer (for a load current of ~50mA).

A little switching supply would probably have been a better choice.

[AHOL]

Hi! I am glad to inform you that modules are on sale again.

https://www.ebay.com/itm/154545708329

[AHOL]

Let's raise for advertising.

[GigaJoe]

I have no clue about an original view of the bottom one , as it already was without the display, but here we are:
sense perception that bottom # a bit smaller, but not big deal. but letters def. are ; top one degrading , not even, but more less acceptable.

[philwong5176]

I am quite happy with the new display after using it for 2 weeks. The letters to the right of the digits are smaller than the original VFD but after a while it's not that bad.

If there's a Version 3 of the display I would recommend making the letters a bit bigger.