Idiot building a MK14 thread

[SiriusHardware]

I think this was only done previously by Philoupat83 for his PIC14, which used smaller tact switches. As far as I know no-one has done this for the 12mm tact switches used on the issue VI, etc, yet.

I know you have already 3D printed items with fine detail in a second colour embedded in them. It will be interesting to see what you can do on the relatively small (10mm by 10mm?) area on top of a keycap. Speaking for myself though, I am very happy with the 'look' that mine has with the original-style legends printed on paper under clear keycaps, especially now that the 'Slothie' keyframe you printed for me has given it the final polished touch.

In an ideal world you would be able to print at a resolution which would allow clear rendition of the words "Go, Mem, Abort, Term" in that small space but they were frequently abbreviated to "G, M, AB, T" back in the day when key legending was often done with rub-down Letraset transfers.

[Mark1960]

Originally Posted by ScottishColin

Counting the wires there, I can only see 16 - is that right?

My original MK14 issue IV has an 18 way ribbon.

16 would be enough connections for an 8 digit 7 segment +dp display. Maybe earlier MK14 only had 16.

If you used a 16 way ribbon it would need a little extra care in assembly to make sure it was fitted to the correct 16 holes on the display and main board, so maybe this was an early example of poke yoke.

[SiriusHardware]

Counting the wires there, I can only see 16 - is that right?

Very observant! I had not noticed that. Like Mark's my original MK14's ribbon cable was 18-way.

My immediate thought was that the artist had made a mistake or employed 'artistic licence' but in fact if you look at the illustration really carefully the drawing shows two additional holes on the left hand side of the display which are not populated by the 16-way ribbon.

In the context of the MK14 there are basically three different types of bubble display which could be encountered:-

-A display with eight bubbles / characters
-A display with nine bubbles / characters
-A display with nine bubbles but only eight characters, the leftmost 'bubble' being a dummy with no actual display element behind it.

Here's a reminder of the pinout of a compatible display (attached), this is the pinout of a 9 way bubble or an 8-way bubble with a dummy ninth bubble at the left hand side.

The MK14 actually only uses the 16 rightmost connections out of the 18 on the PCB and the rightmost 16 connections on the display, so on a 9-way or [1 dummy + 8-way] display, display cell 1 is not connected or used, it only uses display cells 2-8 either way.

If your display is only 8-bubble then at a minimum the 16 rightmost connections on the MK14 PCB should connect to the 16 rightmost connections on the display. If your display has 18 holes like the MK14 PCB does it is OK to connect all 18 holes to all 18 holes on a one to one basis. As built, there is nothing connected to the leftmost two holes on the MK14 PCB but from a stiffness / support point of view an 18-wire strip is better than a 16-wire one.

[Mark1960]

I think the digits should be numbered right to left.

The ninth digit, if present, can be driven by adding a link on the back of the MK14 pcb to the 74145, this ninth digit will be driven using data from ram, but the monitor doesn’t initialise the ram for that digit, so it will display rubbish until the ram content is cleared.

[ScottishColin]

Mine has 9 - see attached photos. I assume I start attached the ribbon on the right hand side (looking at the front) and work left?

In case the photo isn't clear, it's marked as LTB-1498/T9148 on the back.

Colin.

[Mark1960]

Yes connect from right to left with the connections on the main board.

I haven’t connected one of those type to any of my MK14s but I did compare the pinout to a working display using a multimeter on diode test to light each segment of each digit.

One of the other displays of that type had one weak segment, so I’m not sure of the quality of any internal connections. Try and avoid any stress between the display module pcb and the plastic bubble lens.

[SiriusHardware]

I think the digits should be numbered right to left.

It is what it is, I'm afraid. I first drew that sketch (in pencil) in around 1978. I'm not changing it now . You are welcome to right click, 'save file as' and rearrange the numbers in Paint, or similar.

Colin, certainly on that particular display you are going to have to bring the cable up to it from behind (like on mine) because it only has solder pads on the front of the display PCB. (Most such displays including the original display are constructed on double sided PCB so they would have solder pads on both sides).

I seem to remember you do have a temperature controlled iron, this is one of those jobs where you will need to have it on significantly less than maximum, partly because you are going to have to solder quite close to the edge of the clear 'bubble' window which will be easy to melt, and partly because single sided PCB pads are easier to burn off with excess heat. Not to mention that LEDs themselves are notoriously heat sensitive.

Did someone mention that they had a lot of bother getting the pads on those 'Russian' displays to take solder? If so what was the solution?

[SiriusHardware]

While you are waiting for the display connector (Colin) you could give the display a quick test as per Mark's method - have your DMM on 'Diode Test', put the black lead on the '9 common' connection and hold it there while you touch the red lead to the a, b, c, d, e, f, g, h terminals. You should see each segment of the far right display light up (dimly) one after the other.

If you then move the black lead to '8 common' and run the red lead through a-h again this time you should see each of the segments of the second from rightmost display light up in sequence.

You can check the whole display that way, but note that your display may or may not have a working leftmost display cell. If it doesn't, that's OK as the MK14 uses the other eight.

[Mark1960]

They are nine digit displays. I’ll try soldering one and see how it goes.

[Mark1960]

Solderability is a problem. I tried fitting a connector without any abrasion of the pads, increasing the soldering iron temperature to 310C or adding liquid or paste flux didn’t seem to help. Mechanical scraping of the pads through the solder with the tip of the iron seemed to work, but not a good solution.

I think its best to very lightly scrape the surface of the pads with fine grade emery paper before fitting the cable and attempting to solder the contacts.

The good news is that it seems to give a slightly brighter and more uniform display than the one I was using before. Markings on that one were ANC 318A, not sure of the brand.

Edit: Just to clarify, the one I just soldered was LTB-1498, from the same batch as the one sent to Colin.

[SiriusHardware]

I remember that Colin has a fibreglass pen, would that be less harsh than fine sandpaper? (I suppose it depends on whether you count the several days afterwards trying to pick tiny bits of glass out of your hands).

[Mark1960]

I think a fibreglass pen would be better, I never thought to ask as I don’t have one in my own toolbox.

[ScottishColin]

All 9 tested ok as per your post - thanks.

I have my fibre glass pen at the ready. Thanks for the tip about lowering the temperature - I probably wouldn't have thought of that.

Colin.

Originally Posted by SiriusHardware

While you are waiting for the display connector (Colin) you could give the display a quick test as per Mark's method - have your DMM on 'Diode Test', put the black lead on the '9 common' connection and hold it there while you touch the red lead to the a, b, c, d, e, f, g, h terminals. You should see each segment of the far right display light up (dimly) one after the other.

If you then move the black lead to '8 common' and run the red lead through a-h again this time you should see each of the segments of the second from rightmost display light up in sequence.

You can check the whole display that way, but note that your display may or may not have a working leftmost display cell. If it doesn't, that's OK as the MK14 uses the other eight.

[SiriusHardware]

Using proper Lead/Tin solder will also make it a lot easier but hopefully that's what you always use anyway.

[ScottishColin]

Ok. Ribbon arrived, soldered into the PCB and the LEDs. I've tested again with multimeter as before.

Are there other tests I should do before populating the PCB?

Colin.

[SiriusHardware]

Since you've already established that your 5V supply is 5V as it should be, I would say just go for it. Just be patient, as long as everything goes into the right socket the right way around it should 'just work'.

If it does work it should come up with '0000 00' on the display when first powered on.

One potential cause for error might be the placement of the reset switch which can be inserted in any of four different orientations due to the square pinout, but only two of the four orientations are correct. Hopefully you noticed that the reset switch has a 'flat spot' on only one of the four sides. There would have been a corresponding orientation marker or key on the screen print.

And... if necessary, change the 'memory type' link from the default '2111' to '6561', if that is the type of memory you are using.

[ScottishColin]

OK - the PROMs you burnt me; green sticker in U2 (top right of board) and blue sticker in left hand side (U3) - is that right?

And of course I'm a chip missing from the BoM - 74LS173 so I'll get that ordered and hold my breath.

Colin.

[Mark1960]

Fit the minimum components first, don’t fit the 8154 yet or the two ram chips closest to the keyboard. If the display doesn’t show 0000 00 as soon as you press and release reset, turn it back off straight away and check for assembly or soldering issues.

[SiriusHardware]

Yes, Green dot is U2, Blue dot is U3.

I would have chosen it the other way around so that Green was IC three and blue was IC two, but no, SOC had to go and do it the other way so we are stuck with that convention.

Obviously, don't power it up until all of the necessary ICs are fitted - you can leave out the 8154 and two of the RAMs for the initial power up, as Mark mentioned, but that missing 173 needs to be in there.

[ScottishColin]

While we're waiting for the missing IC to arrive (this thread was well named), could you explain what the convention actually is please that makes one blue and one green?

Ta.

Colin.