Fluke 45, replace dead VFD by 7-14 segment LED display´s

[Satbeginner]

Hi all,

I bought a Fluke 45 on eBay, "Not Working", "As Is", " For Parts", so when it arrived it did not work, this was not a problem.

But my goal is to repair this meter, and since I already had a Fluke 45, finding the problem was fairly easy: the VFD (Display) is broken.
By exchanging the two displays I  could confirm the meter is still working, just not showing anything.

Are there members that recently were able to buy Fluke spareparts?
And if so, where?

I live in Spain, but have family in The Netherlands, so I am flexible in terms of location.

It is this part (just the display) that I need.

Un saludo,

Satbeginner

[Ian.M]

Explain 'broken'.  Is the VFD glass physically damaged?  The filaments look to be intact - is that the bad display or the good one?  Are the Getter spots on the left the same colour on the good and bad displays?  If its lost vacuum they've probably gone whitish, but if they are both still dark or metallic looking odds are its still got vacuum.  If the filaments and the getter spots are good, odds are its a problem with its supply voltages or the driver chip.

[Satbeginner]

It´s really broken, the little tube on the top left corner used during manufacturing to suck the vacuum, broke off.

So inside mechanically it is still OK, but instead of a VFD it is now an AFD (Air Filled Display).....

[SeanB]

Rest of the board still likely works, so all you need is the actual VFD and solder it to the board.

[Ian.M]

Unfortunately it looks like Fluke 45 spares were put on 'Last Time Buy' back in 2009 so you are probably S.O.L. unless you can get another parts meter, or can find one N.O.S. at whatever boutique price is being asked, or want to take on the considerable challenge of reverse engineering the display board and designing an compatible replacement using a graphic matrix display that emulates the VFD + controller.

Its probably worth asking here if anyone's replaced a dim display and still has the old one available - if the filament voltage can be boosted a bit you'd probably be able to get the emission up enough to get a few more years life out of it.

[Satbeginner]

Hi Ian and others,

I am afraid you´re right, finding a ´new / replacement´ display is the only way for me.
I do know the rest IS working, because I have another 45, and I switched display´s, and everything seems to work.

So, my humble question you you all is: 

Is there anybody with a dim Fluke 45 display floating around, or maybe a 45 that was struck by a direct hit from lightning?? (pretty sure the display has survived  )

Thanks to you all for the help offered,

Un saludo,

Satbeginner

[Satbeginner]

Hi @Ian.M , #Ian.M ,

redesigning a display board would be a challenge because of the serial in- and output it uses, but you gave me an idea: Maybe no redesign is needed??

I could re-use the existing board, because a VFD has all the individual digits, +/- signs and comma-dots wired to it´s outside pins.
So I should be able to do some current limiting and level shifting and be able to drive a common anode or cathode (instead of the VFD-grid, have to look into this) 7-segment LED display for the numbers, and some LED´s for the functions?

I have seen that my board (my meter is from 1985) uses a different VFD driver chip than the one in the servicemanual.
The manual uses a 75212, I have an D7527ACU, but they seem to do the same thing, because all the signals from the main board to the displayboard are identical.

I´ll try to do some basic testing on this.

Un saludo,

Satbeginner

[Satbeginner]

replacing the Fluke 45 VFD by 7 (and 14) segment LED display can be done, here is the "proof of concept".

In my Fluke 45 with the dead VFD (Vacuum gone) Fluke used a D7527 VFD display controller.
In my copy of the service manual the use something else, but I think they are functionally compatible.

An individual segment of a VFD lights up when both Grid (digit) and Anode (segment) are at +5V, this is positive to the filament AC level.
When a VFD segment is not used, both Grid and Anode are at a negative level.

In real life, these signals move between -32V and +5V.

According to the datasheet of the D7527 all outputs are capable of delivering up to 30mA.
Based on this, I "fiddled" the following solution, where I use about 6mA per LED segment.

I use the diodes (1N4148) to remove the negative -32V, now only using the +5V part of the signal,  followed by a NPN switch transistor to drive the Common Cathode Digit´s.

The LED current is supplied by the VFD-driver chip, again, negative voltages are removed by the serial diodes, current limiting is done by the 470 Ohm resistors. (appr. 6mA)

The reason I remove the Negative voltages, is to protect the display LED´s against reverse breakdown.

I found the five "number" digits can be a combined display who share the segment wires, the minus-sign has to be a separate digit, because segment wire 10 is used for the minus-sign, but it shares the same digit wire as the MSD.

I have not yet tried the second display.

un saludo,

Satbeginner

[carl_lab]

Well done! 

[rsjsouza]

Satbeginner, thank you for sharing your investigation. I love these modifications!

[Ian.M]

Looks like you are getting somewhere.   Time to start designing a PCB and selecting parts.   If you use SMD 7 segment LEDs, you may even be able to come up with a daughterboard that effectively replaces the VFD with only a single flying jumper for Gnd.  That's a marketable product . . . .

[Satbeginner]

Hi Ian.M,

there is physical space available to make a "thin" pcb that would fit in place of the original VFD.
I am looking for parts right now, but I have a few concerns:

- Thickness, probably do-able with SMD displays;
- The huge amount of extra indicators, like AC, DC, Hz, (m)A, V, dB, etc. (Thinking of using LED bars with a silk-screen in front)

My first  test will be something like 2x a normal 5digit common cathode 0.36" display, with a separate digit in front of this for the minus-sign.

I had a look at how to do the secondary display: In the original VFD they use something like a (but not really) 14-segment display in three digit´s, but I could not find any application for these extra segments, so I just think of leaving them out. (I´ll use 7-segments, so, just numbers there, and simple characters like ´print´)

Keep you posted about any progress,

Un saludo,

Satbeginner (Leo)

[netdudeuk]

Good work.

Here's another example of a forum member overcoming the issue of a broken display on an old instrument -

https://youtu.be/H9_Jwy-2vXU

[Johnny10]

Hey,
You've got way too much time on your hands!

 

[bitseeker]

Satbeginner,

I like where you're going with this and how you adapted the existing drivers. The LEDs should look great when you have it all running. Looking forward to seeing how it turns out. If you do end up making replacement modules or kits, I wonder what will happen to the price of 45's on eBay.

[lowimpedance]

 Replacing the VFD is quite doable and your off to a good start. Keep going it will be most interesting to see the final result.

Satbeginner,

I like where you're going with this and how you adapted the existing drivers. The LEDs should look great when you have it all running. Looking forward to seeing how it turns out. If you do end up making replacement modules or kits, I wonder what will happen to the price of 45's on eBay.

Does the 45 VFD fair worse than similar vintage DMMs , ie 8840, or is the 45 more useful than the rather basic 8840 function wise ?.

I had a go some time ago with an 8840a that had very much seen better days ! . Had some rough treatment resulting in a broken display etc.
Only persevered with it as a learning experience, (and that I knew that it worked apart from the display).
 Had to come up with a suitable way to display the annunciators so I used a led behind a film printed reversed with the function clear.
The 8840 uses multiplexing, not sure what the 45 does, but I had quite some difficulty with 'ghosting' of LED's that were not supposed to be on.
A couple of pics of the display working. Note the display actually has better contrast better than the photo lets on. (except the MSD 1 which turned out to be brighter annoyingly even though I played with the average value of current through each LED, quite a pain with the multiplexing).
And yes I know its as ugly as a hat full  , but hey it works and was a "can it be done exercise" without huge effort anyway.

[bitseeker]

If you do end up making replacement modules or kits, I wonder what will happen to the price of 45's on eBay.

Does the 45 VFD fair worse than similar vintage DMMs , ie 8840, or is the 45 more useful than the rather basic 8840 function wise ?.

I haven't noticed it doing necessarily better or worse (nor have I kept meticulous records), but if there's an easy kit or module to revive the aging meter, people may be willing to invest a bit more in their bidding. Just speculation.

The Fluke 45's dual display is a worthy trade-off for its lower (but still very useful) resolution.

I had a go some time ago with an 8840a that had very much seen better days ! . Had some rough treatment resulting in a broken display etc.
Only persevered with it as a learning experience, (and that I knew that it worked apart from the display).
...
And yes I know its as ugly as a hat full  , but hey it works and was a "can it be done exercise" without huge effort anyway.

That's quite a good, and working, proof of concept if you ask me.

[Satbeginner]

Hi all,

I already have a Fluke 45 in perfect working order, and I love it for repair/troubleshooting, because I see both DC-voltage AND ripple without the need of changing settings of the meter.

Call me crazy, but I bought the broken one, just because it was broken........ and I needed a WRP.
This is what I call a Winter Repair Project.

A WRP may not be all that useful, but it keeps me busy and off the streets, and I need a challenge every now and then.

I think of doing the same with the indicators: an LED bar with a laserprinted film showing the text in B/W reverse in front of that.

In parallel of this Fluke 45 exercise, I am repairing a HP 3457A as a WRP as well.
That one is doing measurements OK now, but the selftest is still giving an error 64, I give info on that in another thread.

un saludo,

Satbeginner (Leo)

[bitseeker]

Sounds great, Leo. Keep up the good work.

Being able to see ripple along with DC is exactly why I use a dual-display handheld DMM when checking power supplies. My bench supplies are just single display.

[Satbeginner]

Hi all,

parts are ordered at AliExpress, but due to the Chinese New Year, there is some delay in this little project.
Will keep all you posted.

Un saludo,

Satbeginner (Leo)

P.S. since my HP 3457A repair was successful, I do need something else to do.... 
https://www.eevblog.com/forum/repair/hp-3457a-multimeter-repair-advice-needed/

[bitseeker]

due to the Chinese New Year, there is some delay in this little project.
Will keep all you posted.

Yeah, funny how that works. I guess we can take some time off for Chinese New Year. 

[Satbeginner]

Hi all, back again.

There is progress, although I still need two "one digit" 7-segment displays to create the 'minus' sign. (temporarily done by using red LED's)

I used 2pcs blue 5-digit common cathode displays for the two measuring values, and 4pcs blue 10-bar indicators for the functional indicators.
I am going to mask these LED bars with a text display on transparent polyester film, but I need to get some sheets to be able to print this on a laserprinter.

Instead of the original VFD I placed some single row female connectors on the back of the Fluke display PCB.
The additional driver electronics (14x segment and 11x Digit) are placed on a piggy-back board that plugs in the original display board, two thin cable bundles run to the actual replacement LED-displays.
All the signals I needed are available on the VFD connctor-pins, I just had to add extra pins for ground.

The LED-displays are placed in a way they just fit inside the window in the cabinet, so they are flush with the colored Fluke window.
This just gave enough room for a display-print using normal components with the wires on the back between the original print and the front window.

I choose to do it 'breadboard-style' because I only wanted to repair this one meter, obviously a much nicer result could be obtained using SMD components and a real PCB, but from the outside it does look kinda cool.

Un saludo,

Satbeginner

Here some pictures:

P.S. the intensity difference has to do with the camera, not with the display.

[george.b]

but from the outside it does look kinda cool.

It looks great! Well done!

[Johnny10]

Bravo!

[bitseeker]

That looks fantastic! Bravo!

Using single rectangular LEDs for the negative sign will probably be fine. They look round due to bleed. Just mask, using heat shrink tubing, electrical tape or black paint, the parts of the LED that should not show in order to get a sharp rectangle.