Browsing popular auction sites, I once came across an offer for a display similar to those manufactured in the 1970s and 1980s. It was described as our indigenous CEMI design, although the “NN” logo, denoting National Semiconductor, was visible on the board. Of course, this type of design was produced in communist Poland, although not on a large scale, it is much more common to find displays from Czechoslovakia, where production volumes were larger. So, I think one can forgive the seller’s marketing gaffe of labeling the element as a CEMI production, although this is also some misrepresentation. Nevertheless, the price of the item was quite low, so I decided to order it, and in this article I will tell you about its construction and the adventures of trying to get it to work.

From the outside, the display looks similar to other designs of the time. On the double-sided epoxy-glass laminate there were paths connecting a series of leads located at the bottom edge to the rest of the design. I soldered goldpin leads to the connector to make it easier to get the display up and running. My modification is also visible on the left side of the tin jumper that connects the power line to the lead, the “+” sign was also not there originally. Most of the front side of the display is hidden under a piece made of plastic, whose main task is to increase the visibility of the small digits. There are as many as nine of these, each of which is a separate seven-segment display.

On the other side of the module we can see the previously mentioned National Semiconductor logo – “NN”, and the designation M830 6213. Slightly lower was the black glut associated mainly with Chinese designs. The photo does not fully reflect it, but on its top layer we can see the outline of two silicon cores hidden inside.

It was quite easy to find the pins that supply power to the whole structure, as these are the first two leads. When a voltage of about 5V is applied to them, the display comes to life. Individual segments begin to glow, arranging themselves into specific digits, arranged, however, quite randomly.

In the meantime, I decided to take some microscopic pictures of silicon light-emitting structures. As you can see, gallium phosphide was most likely applied to the substrate, forming individual luminescent segments, or actually subsegments, as each line consists of five elements. The exception here is the dot, separated into two parts. On the semiconductor was a layer of metal, separating the luminescent lines, to which the metal electrodes connected to the PCB tracks were attached. The metal layer was also used to make the manufacturer’s marking – “N”.

You have to admit that the semiconductor layer does not look like something that can glow under voltage, well the magic of microelectronics. In the photo seen above, you can also notice dirt, small black particles that were found on the surface of the cream. I do not know their origin, perhaps they appeared after removing the display cover, although I carried out this operation in a cleanroom, so you can suspect that they are original dirt.
From the very beginning, this display was a bit of a mystery to me. Despite the designation, searching the Internet I did not come across anything special. National Semiconductor produced quite a few different kinds of displays, but I couldn’t find a description of this particular design. And thus the display ended up in a drawer for a long time.
A random photo that changes everything

Some time ago I decided to once again try to find some information about the M830 display. Unfortunately, this time the Internet was also silent, but while browsing National Semiconductor products, I came across an interesting photo. It shows the interior of a calculator labeled Novus 835. Whatever the case, the display board looks very similar in this design, as non-identical to my module. This photo fundamentally changed the rules of the game, because from that point on it could be assumed that the design I once bought was not just displays. And the circuits hidden under the black ground make it possible that it could be a fully functioning calculator.
The board in the photo shows the designation M835, not M830. I searched for information on whether there was such a calculator as the Novus 830, but the Internet is silent on the matter. In a few places it appears that indeed something like this may have existed, but this information cannot be confirmed.
As you can see in the photo, there is a keypad attached to the module’s leads, which presumably short-circuits the corresponding pins, thus triggering a specific function. Therefore, I decided to perform a similar experiment with my display. Besides, in the available descriptions there is information that National Semiconductor calculators were powered by 9V, so that’s the potential difference I decided to seed my module with.

Experiments in trying to call specific keys can be considered both successful and unsuccessful. By trial and error, it was possible to identify several keys corresponding to specific digits, the minus sign and zero. However, strangely enough, it happens that on shorting the same pins the module can react differently, several times eight will be displayed, for example, but with one time the digit one will appear. In addition, after powering up, zero is not always displayed, as is the case with most calculators. Quite often all digits are active, with randomly lit segments. This state of affairs and the fact that I have not been able to find a connection of leads, corresponding to some action makes me think that with this design, there may be something wrong.
Therefore, at this point I will not attempt to restore the module to its full functionality. Especially since it is not entirely clear whether the 830 calculator existed at all, making it unclear what functions it could support. I am glad that it was possible, at least to call up specific numbers.
Sources:
- https://wiki.calcverse.net/index.php?title=National_Semiconductor_835&mobileaction=toggle_view_desktop
- https://americanhistory.si.edu/collections/object-groups/handheld-electronic-calculators?edan_q=%2A%3A%2A&edan_fq%5B0%5D=set_name%3A%22Handheld%20Electronic%20Calculators%22&page=10
- http://mycalcdb.free.fr/main.php?l=0&b=267