Fluke 8060a refurb... possible mainboard-to-micro hack and oled conversion

[JdaleH]

i found a few tiney box glitches in the  .eps file, but I don't know if they will be a problem.  There is one little box in each digit group.  I will try to clean it up tomorrow.  Dale

[ogdento]

great draft of the overlay!  I believe yours is a spot-on reproduction of Dmitris... but I was hoping to make a few changes (to mine at least).  I wanted to lengthen the segments so that the overall display doesn't look so "choppy", for lack of a better word.  How big I can make them will depend on the material I use for a diffuser (epoxy, hot glue, pcb, etc).

I also rounded my segment edges - and can't make them as close as I'd like - because a pcb cutting machine is actually going to be routing this out.  I also made the digits bigger/taller since we don't have a continuity bar at the very top.

I whipped up a quick sketch of your draft (labelled "A") vs. what I was thinking (labelled "B")...

[JdaleH]

I really like your layout of the digits (it certainly looks better and more original). Using the pcb router to make the slots might be cheaper than using a laser cutter on a metal overlay.  The thin pcb material would work great under the router.  I think that frosted mylar film would make a good diffuser between the two pcbs.  I have a 50 yard roll of the stuff.  I can send you some to try out  It would just depend on the temperature of the display.  BTW, I did not manage to repair the black splotches on the old display.  I tried adding a new polarized sheet with shiny mylar backing, but the splotches were still there.  bummer.  Looks like you have things going in the right direction, so I will hang tight to see what you finally come up with.  Dale

[ogdento]

hey thanks!  and thanks for the offer of mylar... that may actually come in pretty handy.  how tall were the digits in your original draft?

i was thinking about the round cutting bits the pcb manufacturers use, and depending on the minimum diameter they can do i might be able to approximate a more 7-segment like look, kindof like what i highlighted in the green box below:



...but i really have to stop screwing around nit-picking the final details and get going.  i have to put together the test rig, order some varous led types/sizes, figure out how to draw the solder trace board and it's cut-out mask on kicad, and then order some to mess with

[JdaleH]

I like your latest, but there may be issues of cost due to bit breakage or LED bleed-thru between segments due to close spacing.  I think your previous design will reduce both issues and make the layout less expensive at the fab.  The bleed-thru issue was on my mind when I did the first layout like Dmitri's, even though it didn't look very nice.  If the reverse-mounted LEDs can be used in a slot the same shape as the overlay pcb, the bleed-thru could be eliminated and you could use your latest design.  I think the smallest feasible router bit is 0.5mm.  The mylar diffuser should work great to distribute the LED light within a segment. 

[JdaleH]

Oh, I forgot.  I'll send you a processor board to use with your test stand.  I have a parts unit donor.  Just contact me privately at joe(dot)dale(dot)hardin at gmail.

[ogdento]

Cool, 0.5mm min... thanks.  And no worries on the processor... I've since collected more parts units too (I just counted and I have *gasp* 10 8060s, and an apparent hoarding problem!)  You should keep yours for when we get a decent display going

You're absolutely right about bleed-through, and I think my latest had the segments too close for even just the physical integrity of the board.  I think bleed-through can be managed by either ordering the board with the segment slots plated through, or by painting the insides of the slots black. Plating seems easier if I can get it right in kicad!

Bleed-through is also why I didn't want to use a 3rd pcb with no copper as the top diffuser layer... bleeding was terrible in the Hackaday post where I saw this whole sandwich idea.  It seemed that the light gets into the core and just spreads right through the to the other segments.

Here's a sketch of the overall concept:



and then a grab from the hackaday post showing only the two layers I want to use:

[JdaleH]

I was thinking about placing the mylar between the two pcbs.  If it works, it will be easier than other methods.  However, you would have to glue the sandwich together rather than solder.  I just looked at Kingbright and the 0603 LEDs are 1.6X0.8mm.  Thats only 0.02mm room for placing the part under the 1mm x ? routed segment.  Can a pick machine do that?  I would assume so.  But, I don't know if I could hand place that close for prototypes.

[ogdento]

ha, i bet i can't hand-solder those either... but maybe with a lot of practice

the hackaday display used 0802 leds... i'm thinking they might be right-angle (side display) ones that he just rotated and soldered upright?  i sent him a message to see if he remembered the part number.

i think the ones Dmitri used are SMD-2 3.5x2.8mm, which are too wide for the "slotted" design approach - but here's the picture he sent me anyway:



check out Dmitri's photo showing the overall thickness of his display and how it fits into the 8060:

[JdaleH]

How about overlaying your two layer pcb sandwich with a thin frosted glass?  That would almost look like the original monochrome display.  So you would have two pcbs, thin frosted glass (or my favorite mylar) and the original clear plastic protective lens.  If that is too thick, maybe the second pcb layer with the cutouts can be thinner.  Some of the nano smt LEDs have different heights.  The trick is to get the sandwich the right thickness.  I'll have to get out my micrometer tomorrow and start looking at options. There's not much room to work with.

[JdaleH]

Here are the pcb thicknesses available from JLCPCB:
0.4 mm
0.6 mm
0.8 mm
1.0 mm
1.2 mm
1.6 mm
2.0 mm

[ogdento]

I didn't realize jlcpcb had that many options... that's good because I measured from the top of the perspex to the bottom of the lcd glass on two meters and got surprisingly different results... 4.14mm for one and 4.7mm for the other!

I like the glass/mylar top idea... I was originally thinking if the diffuser filling in the spacer board looked good we could just use a super thin clear plastic, but glass or mylar will probably look better.  Not sure if the original perspex is going to fit... it is THICK

[ogdento]

I ordered 15 different kinds of LEDs from mouser so when they arrive I can start fiddling with them to see which ones look the best.  I ordered devices with the highest mcd I could find at 2mA and a viewing angle over 100 degrees.  The red/orange have Vf around 1.9v while the green/blue are 2.6-2.8.

I also scoped out the LCD pins and here are the basics:  The waveforms range from 0 to about -3.2v.  The segment waveform is a 156hz square wave, and the H1/H2 are stepped and out of phase with each other.  When H1 or H2 and the segment waveform have a 3.2v difference (this will be for half of the H1/H2 cycle), that segment is ON.  The other half of the cycle has a difference of 1.3/1.8v and the segment is OFF.

The H1/H2 signal looks like this (showing H1 in pink):


Here are some shots probing the "Rel" segment, which according to the schematic is controlled by pin 3 and pin 1 (H1) on the LCD pads.  I set the meter to ohms, so it will display "OL" with "Rel" initially OFF.  The H1 signal is pink, Rel is blue...

Here's a shot of H1/Rel(pin3) signals with Rel OFF:


Here's a shot of H1/Rel signals with Rel ON:


the other segment controlled by Pin 3 is the Audible indicator, driven by H2 (yellow)... You can see below that the H2/Pin3 signals are never 3.2 volts apart so the segment is always OFF:


**edit: I always mess up the image order!

[ogdento]

Dale, here are those photos of the lcd pulled from a parts meter...

First is just the glass (it looks clear-ish but I suck with the camera... I could see more detail through the lens that didn't show up in the photo):


Second has the polarizer and you can really see where the liquid went nuts:


Last is a decent shot showing the segment trace connections on the glass:

[ogdento]

Dale here's some pics of the lcd I sent off to you...

I peeled off the sunburned back polarizer and the glass was pretty clean so I put some paper, tin-foil (front/back) etc. behind it - the two photos using the dull side of the foil looked the best, but I needed something stiff behind it to smooth out the foil and keep it from flopping down.

Anyway, I think you'll be able to make it look great with the mylar, and I'm looking forward to seeing one of your meters back in action!  The others will have to wait til the LED display is prototyped.  I did just receive my box of LEDs from Mouser, so I've got some testing to do...

[JdaleH]

Thanks.  That display should get one working.  I just received caps enough for two.  The third will be a parts unit because the board is not in good shape.  In the meantime, I replaced batteries in a 8050A "parts only" unit I got off ebay.  Works fine now, except it is slow to settle down.  I got caps for it also.  When I finish, I will clean, clean, clean to help it to zero out.  If I get the 8050A working fine, I may do an oled display replacement to make it good for another 30 years.  Could you send me your latest schematic for the 8060A LED display?  I want to use it to learn kicad.  Thanks,  Dale

[ogdento]

Here's the schematic I did... it's not right so I wouldn't build a board based on it, but it's certainly good enough to mess around with Kicad.

[JdaleH]

I was hoping you had corrected it.  I read the other poster's comments, but really didn't try to understand what the problem was.  I'm going to have to set hobby projects aside for a few months.  I have some "honey-doos" to get done and some traveling to do this spring.  I'll still be around online.  Dale

[JdaleH]

I got the display.  It will be a few weeks before I can try it out.  I want to recap the board first.  Thanks so much, Dale

[ogdento]

Oh that's great!  Can't wait to see it back up and running

I still haven't soldered up a new test board yet, there hasn't been much time for tinkering lately but I hope to get some minutes to try out my new pile of smd leds

[ogdento]

Hey Dale did you ever get that display going?  I still haven't done a kicad layout for the board... lots of distractions!