Datron 4000A getting at test points??

[doktor pyta]

Thanks branadic,
You just proved that Your idea is not only feasible but nice looking!

[branadic]

Dziękuję bardzo

-branadic-

[wolfy007]

Does look pretty good doesnt it.

Might have to look at this for my 4700 projects. I think one of them may have a dodgy display but havent started working on it yet to confirm the problem.

[feedback.loop]

The dim digit seems to be a real problem. Any ideas how to fix it?
Also there seems to be some light leaking through the divider into the adjacent segments.

[branadic]

I don't think it's a real problem. As far as I can remember you wasn't lucky with your solution either and it needed a lot of changes. With my approach I wanted the unit to be as original as possible, with only minor changes to the hardware.
So this solution is better then having no display at all

I will wait for the new display boards to arrive and install both of them to see what's the actual situation on my calibrator. I then plan to increase the overall current through the LEDs by changing the resistors.
The light leaking is due to the plastic frame made fabricated with DLP, the white material it's made of being slightly transparent for the light, the fact that none of the parts is glued together without space in between and that I'm currently using a printed paper on top, while I'm waiting for the repro film to arrive.

Solution:
It's possible to paint the DLP part black or to use a black FDM printed material to avoid light leackage. Furthermore I think with the repro film instead of a printed paper sheet things will improve as well. However, a black epoxy molding compound as the frame directly molded to the surface of the pcb would be best, but it doesn't make sense for such small quanities.
And last but not least, glue everything tight together.

-branadic-

[coromonadalix]

I don't think it's a real problem. As far as I can remember you wasn't lucky with your solution either and it needed a lot of changes. With my approach I wanted the unit to be as original as possible, with only minor changes to the hardware.
So this solution is better then having no display at all

-branadic-

yess indeed,  better than nothing,   great work 

[feedback.loop]

I don't think it's a real problem. As far as I can remember you wasn't lucky with your solution either and it needed a lot of changes. With my approach I wanted the unit to be as original as possible, with only minor changes to the hardware.
So this solution is better then having no display at all

You seem to be a bit defensive. I did not mean to diminish your approach, and I did not mean to compare which one is better. Yes, my approach has problems. I do not like the result. Your approach is arguably better. But this dim digit, in my view, is a problem. It would be nice to come up with a solution if possible. Of course, any working display is better than no display.

[branadic]

Please don't understand my response as an attack on you, I simply had in mind that the OLED displays where also quite dim, despite your big effort in reverse engineering the display protocol. I pretty much liked and followed your video with big interest. Sorry if you got my statement wrong.
However, if it turns out that it's really due to (wanted) timing of the multiplexed segments I have no chance to bring them all up to the same brightness with this approach. I can only increase overall brightness. But since the more dim digit walks depending on the setting I currently see no solution. I wouldn't call it a problem but a drawback. Need to investigate on that when the new boards arrived and both have been installed.

-branadic-

[branadic]

After waiting forever I today received new display boards as well as 3D printed black frames. Board now has mounting holes for the frame. Interestingly the resistor boards for investigation on NOMCA resistor networks for LTZ1000 circuit that have been ordered long before still haven't arrived.
Time to solder some diodes. 

-branadic-

[maginnovision]

Does that picture show the printed films as well at the top?

[branadic]

Yes, glued onto a white frame. The black frame will look much better.

-branadic-

[maginnovision]

Looks really sharp. As for dim figure might be practical to use some small buffers or latches if it is a timing issue.

[branadic]

Finallized two displays, that will be installed next week together with the buttons that should arrive.

-branadic-

[branadic]

Received my buttons/switches today. Many thanks. Seems like I have something to solder tonight.

-branadic-

[maginnovision]

Ever get one of the new displays installed? I received my new buttons a few days ago and now it's much nicer and easier to use without a flaky button.

[branadic]

Yes, buttons and both displays are installed, but I felt like there was no interest in any update here. The repro films are glued to the printed spacers and finally I glued a sheet of paper in between to get it diffuse. The Datron is up and running, but I had no time to further investigate on the dim digits and won't have the time before MM2019. I also need to test the difference between black and white spacer, but that's a story for after MM2019.

-branadic-

[maginnovision]

Ah, ok. Well I'm very interested. Not only do I have the calibrator but I have many Datron meters: 1 x 1082, 1 x 1072, 3 x 1062 LS101, and 3 x 1062. I already put my perfect spare display in the 1082 and the original version of your replacement display would rival 1 or 2 of mine for clarity.

[essele]

I'm also interested to see how it's going ... I'm in the lucky position of having three calibrators all with working displays, but I'm in the market for a 1081 so it's always good to know if there are viable options for replacing broken displays.

[branadic]

As I said, both displays are working. If you guys are interested in, I have boards, printed spacer with attached repro film left. So if you want to test the boards yourself simply leave me a pn.

-branadic-

[feedback.loop]

The dim digits seem to be a timing issue. Looking at the schematic I see that the inter-digit blanking strobe depends on the comma strobe. I noticed on your photos that digits with active comma (segment "i") are dim. Somehow it did not matter for the gas discharge display, but it does matter for the LED.

By the way, notice the capacitors between the segment logic-level outputs and the transistor drivers. Effectively they differentiate and produce short pulses during transitions. I guess that was done to start the process in gas discharge displays. Perhaps they are not needed for to drive LEDs and can be removed (shorted)? However that can change the peak and average LED current, and might require changing some other parts like current limiting resistors and such. I am not sure this can help with the dim digits though.

[branadic]

Thanks for sharing your analysis and good to know that you come to the same conclusion, that was also my interpretation of how this works.

-branadic-

[maginnovision]

By the way I will PM or email you about a display(or two) soon. Busy days though.

[feedback.loop]

Any pictures of the improved display? I would love to see how it turned out without the leakage to adjacent segments.
Any ideas how to fix the issue with dim digits yet?
Thanks.

[branadic]

No new pictures by now. With the new spacer there is no more leackage to adjacent segments. However, I'm not happy with the dim digits and I have no idea how to fix them without changing to much of the original circuitry, but I have confidence that it is due to a timing issue and most likely wanted or was at least no problem for the original gas discharge display.

-branadic-

[feedback.loop]

branadic, could you share your designs? As I understand, there are three parts: a PCB with LEDS, front window with transparent symbols and a plastic spacer-divider to stop light leakage between segments. Perhaps you have some drawings that can be submitted to some manufacturers to produce those things. Where did you order them? How much do they cost? Thank you.