-
Rather than hijacking the thread "Fluke 8840A/AF 8842A VFD Display Replacement" I am starting a new one to outline my work in creating a 256x64 pixel graphics display to replace the VFD in Fluke 8840A/AF & 8842A multimeters. I have posted some of the basic work on the Fluke group over @ groups.io. But there is a lot of interest here and seemingly little crossover.
A quick caveat on both this thread work and that of the referenced thread. There is a very well made blue LED replacement module made by a group in the Ukraine in eastern Europe. It directly replaces the VFD on the PCB and is powered from the Fluke 5VDC test point. I bought 4 of them to repair several meters immediately. They are quite nice except for a few issues I will point out. The blue color shows through the Fluke smoked bezel better than a brand new VFD (I have one). If you look close at the picture you will see that the annunciation glyphs are considerably smaller than on the original VFD. It is a flaw that should be correctable, but I dont know the restrictions the designers were faced with. At ~$35USD & $5USD shipping you cant go wrong.
It is a pain to remove an old VFD. Use lots of liquid or paste flux so as not to burn the pads or PCB. Suck out ALL the solder with a sucker or wick. Work from both sides. Once the old leads are loose, you have to carefully cut through the old mounting pads (4) with a box cutter, then clean the PCB. Nothing stronger than IPA or GooGone or you may damage the button conductive points. Then just follow the supplied assembly instructions.
KO3Y
-
If you head over to the buy/sell board the designer of this display replacement has a thread which should give you more detail etc.
https://www.eevblog.com/forum/buysellwanted/ebay-led-display-module-fluke-8840a42a-(diy-kit)-replace-broken-damaged-vfd/
-
If you just want a quick display fix, get the blue LED replacement. If you want to have some fun with a flexible graphics replacement (at a higher cost in time and money) read on.
I like building stuff with mono & RGB graphics displays. I have used several based on the ST7735 type controller. While not simple to use, the displays using this controller have directly addressable RGB pixels. This makes writing code using bitmapped font tables rather straight forward. However, I found no such displays that came near fitting the Fluke 884X front panel format. The Fluke requires a 4:1 display ratio to get all the information contained in the original VFD. However there are several OLED monochrome displays available in the $30USD range with a 4:1 ratio having 256 x 64 pixels with 16 gray-scale levels. These come in green, blue, yellow & white. (No red that I have found.) I'll jump to the punch-line and show the completed display. Then cover the development.
The following image shows the closest I could get to duplicating the Fluke 884X VFD and take advantage of the flexibility of a dynamic graphics display.
more to come - KO3Y
-
I actually think the LED replacement looks better, it's much sharper and a nicer color though I do wonder why the glyphs are smaller than the original display, it seems like they could be made in arbitrary sizes. Neat project just the same though.
-
Neat project, but an OLED display has a shorter lifespan (And burn in problems) than the original VFD which sort of defeats the point compared to the current LED kit?
-
Yea, there are a range of options for replacing the VFD. If AHOL ever fixes the annunciation size close to the original it is the cheapest fix.
The OLED lifespan & burn-in is a bit of a red herring issue in this application. Certainly the OEM VFD had burn-in issues and most displays are still good after 35 years.
This OLED display design is an attempt to make the Fluke 884X look as it might if Fluke used a LCD graphics display in the design in 1985!
Anyway, it was kind of neat to be able to control all of the units to display large and easily readable. Plus there is both code space in the controller and display to put programmable logos and model numbers within the display rather than on the outer plastic bezel.
-
I also, on occasion, have some alternate color display bezels for sale on eBay. They can give an interesting different look to a fading OEM VFD without replacing the VFD itself. The first images are of a green bezel with a bit of a machining defect! The part broke loose in the fixture and the tool cut a worm path into the bezel!
-
I think the best would be an LED display using LEDs of a color approximating that of the original VFD. I quite like the soft greenish blue VFD color, I'm really not a fan of the saturated blue LED displays, but I do like the nice sharp segments and glyphs. I have some greenish blue LEDs that are very close to VFD but I've never seen them in display form.
-
Yea, there are a range of options for replacing the VFD. If AHOL ever fixes the annunciation size close to the original it is the cheapest fix.
The OLED lifespan & burn-in is a bit of a red herring issue in this application. Certainly the OEM VFD had burn-in issues and most displays are still good after 35 years.
This OLED display design is an attempt to make the Fluke 884X look as it might if Fluke used a LCD graphics display in the design in 1985!
Anyway, it was kind of neat to be able to control all of the units to display large and easily readable. Plus there is both code space in the controller and display to put programmable logos and model numbers within the display rather than on the outer plastic bezel.
OLED Lifespan and burn in isn't a "Red Herring" issue at all, I've been using OLED's for years and unfortunately they do suffer badly with their brightness decreasing and also burn in. As impressive as your project is and I'm not knocking that, OLED really isn't ideal for this application.
Visibility won't be great in a lab as they are not that bright to begin with and most multimeters are left on for long periods of time often with many static digits.I think the best would be an LED display using LEDs of a color approximating that of the original VFD. I quite like the soft greenish blue VFD color, I'm really not a fan of the saturated blue LED displays, but I do like the nice sharp segments and glyphs. I have some greenish blue LEDs that are very close to VFD but I've never seen them in display form.
I think that could be achievable with some colour filters, most VFD's have a colour filter in front of them to get the desired colour.
Granted the Blue LED's are a lot brighter and not much green in the colour spectrum but I wonder if a dark deep green or a rose filter will tone it down a bit and give the desired colour. -
You're not going to be able to filter a blue LED to some other color. The VFD phosphor produces a quite broadband emission spectra, a blue LED is a fairly narrow spike. Since filtering is subtractive it doesn't really work with narrow band emitters.
-
OLED Lifespan and burn in isn't a "Red Herring" issue at all, I've been using OLED's for years and unfortunately they do suffer badly with their brightness decreasing and also burn in. As impressive as your project is and I'm not knocking that, OLED really isn't ideal for this application.
Visibility won't be great in a lab as they are not that bright to begin with and most multimeters are left on for long periods of time often with many static digits.
I accept your experience with OLED devices, however none of the replacement approaches come with any long term reliability data and certainly haven't had quality endurance testing as a big name manufacturer might do. These blue LED replacements may fade in time! And I dont mean to indicate that using a OLED or LCD module is going to in any way be cost effective compared to this blue LED drop-in replacement. The OLED alone costs as much as the blue LED module! and the OLED still needs a 7 sq in PCB packed with circuits to convert the 30V signals and drive the graphics!
But my goal was to free the display part from the button PCB so that nothing needed to be re-soldered to the Fluke PCB. The OLED display is held in place by friction clips and can be easily removed albeit the meter has to be disassembled. I really wish I could find a full color LCD in this form factor! I could really create an interesting display with it. But, alas, I have not found one suitable.
In the photo below one can see the display and ribbon cable floating in the button frame opening. And in the second, the friction mounting clips. The third photo shows the controller PCB that mounts to the inside side of the display PCB. The controller interfaces to the display PCB to convey the button signals by plugging into the inboard original display interface socket. The ribbon cables from the main board plug into the controller sockets as they did to the original display PCB. No other connections are required other than the 5VDC power that is derived from the main board 5VDC supply. This way the display & controller remain isolated from the case on the internal in-guard side.
The switches (3) control the bootloader, display format options (7), and a display simulation mode. The controller is held in place by the one 16 pin socket on the original display PCB (hidden), a friction clip and a fortuitously located hole through the button retaining frame (yellow).
-
You're not going to be able to filter a blue LED to some other color. The VFD phosphor produces a quite broadband emission spectra, a blue LED is a fairly narrow spike. Since filtering is subtractive it doesn't really work with narrow band emitters.
That's what I was wondering when I said it has not much green in the spectrum, maybe a white LED and coloured filter would be a better approach.
For what it's worth I can't stand Blue LED's -
A white LED would certainly work, although those rely on a phoshor and in my experience white LED displays have the same problem as VFDs only worse. The best is a non-phosphor LED emitter in the desired color which do exist.
-
Rather than hijacking the thread "Fluke 8840A/AF 8842A VFD Display Replacement" I am starting a new one to outline my work in creating a 256x64 pixel graphics display to replace the VFD in Fluke 8840A/AF & 8842A multimeters. I have posted some of the basic work on the Fluke group over @ groups.io. But there is a lot of interest here and seemingly little crossover.
A quick caveat on both this thread work and that of the referenced thread. There is a very well made blue LED replacement module made by a group in the Ukraine in eastern Europe. It directly replaces the VFD on the PCB and is powered from the Fluke 5VDC test point. I bought 4 of them to repair several meters immediately. They are quite nice except for a few issues I will point out. The blue color shows through the Fluke smoked bezel better than a brand new VFD (I have one). If you look close at the picture you will see that the annunciation glyphs are considerably smaller than on the original VFD. It is a flaw that should be correctable, but I dont know the restrictions the designers were faced with. At ~$35USD & $5USD shipping you cant go wrong.
(Attachment Link)
(Attachment Link)
It is a pain to remove an old VFD. Use lots of liquid or paste flux so as not to burn the pads or PCB. Suck out ALL the solder with a sucker or wick. Work from both sides. Once the old leads are loose, you have to carefully cut through the old mounting pads (4) with a box cutter, then clean the PCB. Nothing stronger than IPA or GooGone or you may damage the button conductive points. Then just follow the supplied assembly instructions.
KO3Y
It really is interesting and a cool thing you are doing, fair play to you.
I'm assuming these OLEDS are SSD1322 based or similar?
Is it possible to get a dot matrix graphical LCD in the same footprint and similar driver that would work too, as if you started making kits then the end user could chose OLED or LCD? -
I'm assuming these OLEDS are SSD1322 based or similar?
Is it possible to get a dot matrix graphical LCD in the same footprint and similar driver that would work too, as if you started making kits then the end user could chose OLED or LCD?
Yes, uses the SD1322 controller. Actually a rather PITA to work with considering that you cant easily address just one pixel. It requires a good bit of thought and planning and not easily amenable to bitmapped and scalable fonts.
I searched long and hard for a 256 x 64 LCD mono or color. I haven't yet found one. Maybe a forum visitor will see this and remember seeing one. I think, from what I have done, the minimum resolution needs to be 256 x 64 or higher with an aspect around 4 or 5:1. All of the small text in the display is one pixel wide fonts. I wish the OLED was brighter. I cant use the OEM bezel and be satisfied.
What is nice with this approach, though complex in firmware, is that the VFD grid and segment data is simply decoded on the fly. The bulk of the code is in decoding and extracting what should be displayed. The actual difference between the SD1322 and say a ST7735 display is just a couple of function calls. -
It would be great if you could just trim to size the way you can with EL strips. Maybe some day.
-
Once I get some new colored bezels made I'll post what the finished meter looks like.
-
While I have had this display design working for some time, I have had an insidious, subtle and consistent problem. While the 2 meters I use as testbeds (8840A & 8842A) are fully functional and pass all self tests with both a VFD and the newer blue LED replacement display, I would get test failures with the OLED design. Specifically tests 4, 6, 7, 15 & 22(8842A). These typically point to the A/D which I know is fine!
I at first suspected my level converter design was at fault for possibly overloading the +30V supply. But I found the errors remained after completely removing the level converters (no connection to the mainboard +30V grid and segment signals! They remained after removing the +5V from the OLED controller PCB. The only connection to the mainboard was the pass-through connections to operate the PB switches and the controller PCB ground to the "digital ground" by the power transformer.
Ok, maybe there is something obvious here, but it took some time for me to realize that there has to be a un-expected short between the analog ground and digital ground. When I removed the OLED controller ground from the mainboard the errors disappeared! But where was the short?
I went back over my design and the schematics of the Flukes looking carefully at the logic control and the VFD board wiring. On all of the schematic versions I have of the VFD, pin 16 of J203 on the display has a 7 in the grounding triangle. That is digital ground. But a careful look at the digital main PCA schematic shows J203 pin 16 with a 2 in the triangle! That is analog ground. A quick ohm meter check confirmed this.
What I had done is follow the VFD/PB schematic and on the pass-through wiring for J203 from the main PCA to my OLED PCB to the VFD/PB PCB I connected pin 16 to digital ground everywhere. Voila, analog to digital ground loop! I simply lifted the pin 16 circuit completely, since the VGD PCB ground plane is rather un-necessary now I think, and all tests pass!
The moral of this story is 1) these old hand-written schematics are not always correct, 2) analog and digital grounding issues can be subtle. The impedance and resistance between them may be small, but the currents can play havoc with such a sensitive meter.
kjo - KO3Y -
Here is a photo of my replacement display along side a couple of pretty good OEM VFD displays.
All three have the OEM filter screen. The OLED display on top is using green OLEDs. I have a
few others using blue, white & yellow that I am experimenting with.
The controller behind the display has quite a few user programmable options as well as a demo test mode.
Kind of expensive to build compared to the AHOL blue LED replacement, but you
do get much better annunciators and units!
-
The 8840A, 8842A & 8840A/AF are about 40 years old in 2022. What a great design to be still in demand after all that time.
But the VFD selected had not held up to the ages as well as the rest of the design. This thread is about the replacement
design I have been working on for quite some time. It cant compete on price with the likes of the AHOL Ukrainian LED design that was
available for a while, but it is a whole lot smarter!
In addition to decoding the grid and segment signals, I have added a feature function mode that was missing from the original design.
Because the controller knows exactly what units, mode, and scale the meter is set to, I have added a continuity test mode
that drives a small PCB mounted sounder. By selecting OHMS, M or F rate and 200 ohms range the function will activate
the sounder if the resistance is below 20 ohms. (the "S" rate has an unacceptable delay caused by the ADC sampling algorithm.)
Further, the continuity mode can be toggled ON or OFF by activation the "AUTO" and "EX TRIG" functions. (See photos upper right)
Also, the "B" indicates that the currently loaded firmware is using a bootloader. If standalone firmware is loaded the "B" becomes a "V".
Kevin KO3Y
-
Well I finally got some more acrylic display bezel lenses fabricated and mounted blue, yellow & green OLED
displays and controllers into 3 8842A meters. The camera does not do them justice because they are so bright!
All 3 look really good to the eye, but my favorite is the yellow.
This yellow OLED has a light amber acrylic lens, but looks good with orange, green or blue lens.
(Bottom is an OEM VFD that I use to think looked good. Second up from bottom is the blue LED. The glyphs are much too small.
The second version from AHOL was a bit better, but he is gone for now.)
-
Quite the stack there
, the yellow does look to be the clearest from the photo (obviously cant tell what real life looks like for each display though).
Looks really good, nice job
.
-
Nice..
I was looking at displays for a similar project found this one 4.5" https://www.alibaba.com/product-detail/4-5-137x39-MM-25664-256x64_1600300422418.html -
@KG7AMV
Looks like an LCD monochrome backlit display. It is the biggest of that form factor I have seen. May not fit 8840A but man are they cheap. I may just have to order some to learn the NT7107 controller.
Also, I would be interested if anyone has C code configuration data for the NT7107 controller.
KO3Y