HP/Agilent 53131A/53132A/53181A Conversion to LED Displays

[TomT]

Hi all,

With the popularity of these counters, I feel that might be prudent to look into designing an LED replacement. The lack of new displays, questionable life expectancy of used ones,  and overall short life of new displays compared to LED's has been of concern since I purchased my counter on eBay.  Fortunately, my display is in good condition, but don't want to wait until the last minute.

The original interface is relatively simple, as all of the decoding is done in the processor on the main board, and clocked into a 32 bit shift register/latch, with high voltage drivers (SN75518FN [obsolete]) on the front panel PCB for the VFD.  All segments and annunciators, as well as digit enable (grid drive) are included in this string.  Since the VFD requires a positive voltage to the grids to enable a character, and positive voltage for each segment/annunciator (plate/anode), all logic levels shifted into the shift register are positive true.

Due to space limitations between the tops of existing parts on the PCB, and the front panel, I am looking at 0.4" 14 segment SMD LED displays, but they only contain decimals, and the segment width is 1mm.  The addition of an upper dot and comma would have to be made.  I only have couple ideas here.  One is to contact some display manufacturers to make single/multiple character displays with the punctuation built in.  Obviously, that would take a rather large order, but if enough people got involved, this could reduce the cost of the displays.  The other option is to make a black polycarbonate/acrylic with translucent pieces for a dot and a full comma.  Logic would have to be used to turn off the decimal point when the comma is in use.  I am not a mechanical designer, so help here would be greatly appreciated.

As for annunciations at the bottom and right side of the display, I expect the best method would be rear or side lit engraved flat black painted polycarbonate/acrylic.

The right most digit is set up in a somewhat unusual configuration.  I haven't decoded exactly what-is-what, but there are more segments used for annunciators than can be accounted for in the outputs of the shift register.  Annunciators M, Hz, µ, S, Gate, and ExtRef eat into the character segments.  Three use the punctuation, and two are specifically for annunciators, so one of the character segments is required for one of the annunciators (i.e. the lower part of the comma is used for the Gate annunciator).

I have two interface design ideas.  One uses 4x 74HCT595's or HC595's, the other uses an MM5450/51 (34/35 bit LED display driver [shift register/latch/driver [w/brightness control]).  Due to current handling limitations of the HC595/HCT595, external driver transistors would be required, as the absolute max current through V_CC or GND is only +/-70ma, and the MM5450/51outputs only have a 15ma pull-down capability.  The TLC6C598 looks like a very good 8 bit x4 option, but none are expected until 2023 at Mouser, not indicated at Digikey, and Newark has SOIC-16 stock expected the week of 12/6/2021.   Other suggestions for IC's would be appreciated.

I am about to order displays from Digikey, and will breadboard it to operate in parallel with the current display in my counter.  I am using RS-485 drivers and receivers, so the counter can remain in its normal position while work is done on the circuit.

I have also looked into upgrading the 34401A 6 1/2 digit DMM's, but that circuit may be a nightmare waiting to happen.

Any suggestions, or current plans for a project like this would be welcome.

[TheSteve]

I'll be watching this - would be nice to use in my 58503B which runs 24/7.

[Qw3rtzuiop]

This guy already implemented an alternative display:
https://github.com/ragges/hp_display

[TomT]

Hi Qw3rtzuiop,

Thanks for the heads up.  That project looks like an excellent option.  However, what I am proposing is a display that fits in place of the existing VFD, with a minimum of additional interface hardware.  It's going to be a tall order with punctuation and annunciators, but I am hoping that with the help of this crowd, we can make it happen.  With all the LCD replacements for CRT's showing up, I feel that retrofitting a VFD with LED's is a very viable option.

[TheSteve]

There is no point to it if it doesn't fit in place of the original VFD.

[TomT]

eDavid, that is an interesting project. However, it doesn't work for the HP/Agilent counters, which are 12 x 14 segment characters plus punctuation between characters, and very different annunciations. The characters display text as well as measurements.  If the manufacturer of the displays in the Fluke project were made available, possibly a special order could be made with them.

[jadew]

I have been looking into developing a similar thing and the custom display is something I was looking into.

Another solution would be a graphical VFD, but I couldn't find one of appropriate size.

[TomT]

Welcome to the group jadew.  I'm glad to see the growing interest in the project.

My first thoughts were 5x7 dot matrix LED displays, but when I saw that the data was already decoded to drive the VFD segments, I decided to go with 14 segment LED's, and use the KISS* method.  A single IC not unlike the one driving the VFD would be the best option, or at the very most, 4x 8 bit shift registers.  Hopefully, the TLC6C598 will do the job.

*If you aren't familiar with the KISS method (Keep It Simple Stupid, or Keep it Simple and Straightforward, etc.), it is a proven design principle where the more complicated a design, the more likely it is to fail than a simple design.  I was first introduced to it in the '70s at Texas Instruments, Houston campus, that closed some years ago.

[coromonadalix]

If the decoding is done, maybe an bluepill   stm32 ...  could do the decoding and interface the leds ??  may need somew buffering

An oled would be nice ...   check the 34401 oled conversion here  made by qu1ck   ....

[artag]

I have been looking into developing a similar thing and the custom display is something I was looking into.

Another solution would be a graphical VFD, but I couldn't find one of appropriate size.

I'm considering whether 2 OLED displays could be used - one for the digit/annunciator area and a second for the units display on the right. There is a bigger gap between these areas than most others.

Also, note that some annunciators are big because each element needs a separate VFD anode. If mutually exclusive annunciators are re-implemented as OLED symbols they will take up less space without losing clarity.

[jadew]

I couldn't find any OLED displays to fit either. I'm currently in talks with some display manufacturers to see if it's possible to get a custom replacement that looks fairly similar.

About two weeks ago I got in touch with the OEM for the original VFDs and they said that they haven't produced those in quite some time and they can't produce them anymore either, because they don't own the IP.

To me, it feels like the whole thing is potentially going to be a big time sink, so I would prefer it if some VFD manufacturer in china decided to make a clone VFD. When those displays started showing up on eBay, I thought that that was the situation, but now I fear they might have been genuine.

[TomT]

I contacted the seller of the Fluke VFD-to-LED conversion, and received a reply last night.  He couldn't give the company name that made his displays over eBay email, but told me that most any LED display company in China would be willing to make special runs.  The only problem being the minimum order of ~1000.  So, I'll be looking into these companies to see what I can learn.

[TomT]

Hi all,

I have a request.  I would like to take better measurements of a dead VFD.  What I want to do is take measurements of the actual phosphor anodes to give the specialty LED manufacturer as much data as possible.  Currently, I have only been able to approximate measurements from the outside of the glass envelope.

I don't care if the VFD has a vacuum, but needs to be intact.

Please PM me, and I will supply my mailing address.

Thanks.

[pizzigri]

Im also EXTREMELY interested in this. I also own a 53131a and the VFD in mine is not bright as all the other Agilent/HP gear I have. Sure its readable, but I’d love to see an alternative display.