What I mean is the brown connector located on the keyswitch PCB of the twin PCB keyboard assembly. The other PCB is the keyboard encoder which has a corresponding 25 pin long pin strip and this one plugs into the wanted connector.
I dare a guess it might be a MOLEX connector but this is suspect wetware memory recall from 45+ years ago when these connectors were popular (human brains = wetware do not have ECC ;-)
Anyone able to identify this connector and tell me manufacturer, part number, and where it's still in stock ?????
Not that I didn't try to find it by web searches but every "hit" I found only talks about "the connector" and no further information about it was found.
I think that others who restore Apple II may have had to replace this connector. They are not very reliable and probably were not meant for many unplug / plug cycles, so there must have been a need to replace them.
IIRC, this type of connector also was used in the early 8-bit Atari computers and the Atari 810 floppy disk drive was notorious for failures of this (or a lookalike) type of connector. Which typically happened shortly after the warranty expired. The Atari 400 and 800 power supply PCB also used this type of connector to plug into the motherboard. Where it also did fail all too often, especially if the machine had been disassembled and reassembled several times. So there is a pattern: these connectors do not like cycles. Maybe they went extinct for that reason. I hope they are still out there . . .
Any hints are much appreciated.
- Uncle Bernie
Mill-Max still makes socket strips like these. I would check 801-93-050-10-001000. This is the "fattest" pin socket strip that can mate with .025" x .025" square cross-section pins. There are other series that only mate with round cross-section pins up to .025", or stamped IC legs .010" thick and .018" wide, in a variety of lengths and heights.
The strips are 50 sockets in-line and can be cut or broken to 25.
I just looked and they are available in 25 sockets also, with part 801-93-025-10-001000 for $6 at Digikey
Hi 'robespierre' -
unfortunately the connectors you have pointed me to in post #2 and #3 are not the right ones.
The connector I seek takes a pin header row from below (!) and to allow this, the Apple keyboard has a row of holes drilled for these pins, just in front of the solder pads of the connector itself. There is only a 'single wipe' contact spring for each pin, and this may contribute to the poor reliability of these connectors. The key switches in the keyboard I have work perfectly fine, but the connector has intermittent contacts which get better with some wiggling / plugging the encoder card out and in, but this remedy does not last for long. This hints at poor contacts, possibly corrosion, but no magic potion such as Deoxit D5 helps. So I need to replace this connector.
Here is a photo of the connector sought:
A2_Keyb_connector.jpg
Some companies still make similar connectors which can take very long pins, these are meant to "stack" PCBs, and if the pins are long enough, they will come out on the top of these connectors, and a second PCB having this kind of connector can then be stacked on them. This feature is rarely used, as the pins must be very long, and get bent all too easily, but the key here is that the connector has an open top and bottom through which the pin(s) can go.
Alas, all the modern connectors of this type are SMD now, which is useless to replace a connector on an Apple II keyboard.
Still seeking for any hint where to find the right ones.
- Uncle Bernie
You can find some hits by searching "bottom-entry socket". I had some trouble finding an image that shows what the connector looks like from both sides. Finally, it looks like the pins enter through holes which are offset about 3mm from a row of PC solder tails that run parallel to the holes? The Hirose MDF7-25S-2.54-DSA(96) would fit the bill, then. $3.72 at Digikey.
There's also the Mill-Max 834-93-025-10-001000 "pass-through socket" with an "organic fiber plug" that protects the contact fingers during reflow. The solder connection to the board is made on the outside of a hollow tube which the mating pin passes through to contact the fingers. So there are no separate PC tails. This is out of stock everywhere, though.
Molex was a good guess: The Full Line Molex catalog from 1981 lists a number of bottom-entry sockets, including the 4455-B series on page 42B. 25 circuit with gold plating is part 22-14-2251 and tin plating, 22-17-2251. This series was then expanded to several related parts with brass or phosphor bronze contacts and different platings: 22-14-2254, 22-17-2252, 22-17-3252, 22-17-2254, 22-17-2250, 22-17-3256, 22-17-2257, 38-00-1415. The full 4455 drawings are here. The space between the sockets and PC tails is greater than the HRS parts: 3.8 mm or .150". These parts are all special order with a 820 qty minimum (a full box). Molex used a variation on their part numbers that prefixes "00" and removes the dashes.
There is another similar part from Samtec, BSW-125-04-L-S, with a socket to PC tail spacing of 2.54 mm or .100". Its drawings are here. This isn't a stocked part either, but it appears it can be factory ordered in single qtys with a 6 week stated lead time. There are a handful of related parts with different plating or low-insertion-force.
Not sure if this will work if you straighten the pins ...
https://www.ebay.com/itm/296506812686
To all who contributed in this thread -
seems you found them - both the Hirose substitute and the original Molex parts. Thanks a lot for helping me find them !
(Finding connectors by online searches always has been difficult for me - there are just too many hits, even from digikey and Mouser search engines)
With the original Molex part out of stock everywhere, and huge minimum order quantity with the Molex and Samtec, the Hirose suggested by 'robespierre' in post #5 appears to be the best viable alternative, although the hole distance is different from the original Molex part. I think this can be fixed by drilling new holes with a drill press and using some 2.54mm grid perfboard as a guide to get the tolerances right.
But still, I'm a bit frustrated about the Apple II keyboard situation. Replacing keyswitches is very difficult and time consuming, and then when all keyswitches work again, after a lot of time spent, it is so frustrating to find out that the connector also is bad.
I think the best solution would be to design a replacement keyboard PCB which is tailor made for the Hirose, and uses Cherry MX keyswitches. I had a keyboard project on the back burner for a long time (to make Apple-1 keyboards) but so far had no time to finish the layout. I just have placed the key switches themselves with all the little mounting holes, and never found a way to just get a PCB drilled with this pattern, and no copper whatsoever, so it should be a cheap test if the holes and their spacing is right, and will yield a good keyboard after adding all the routing. It seems that most PCB manufacturers won't do such a drill test PCB for less $$$ than the finished PCB with all the layers. This is what stopped this project. I'm not poor but I'm very stingy and probably worse than Scrooge McDuck, when it comes to penny pinching in my hobby projects. If I could make a gazillion in profits from such a project, I'd of course switch to a different mode of thinking, and pay whatever the prototypes cost. But no profit could ever be made, at least not with my Apple-1 and Apple II projects, so no big money will be spent by me on NRE, ever.
But now I'm motivated to do it, as it turned out, as evidenced by this thread, that restoring / reworking salvaged Apple II keyboards to turn them into Apple-1 keyboards is not really viable if more than one is needed. At the moment I have two such modified Apple II keyboards, and only one works fine.
Again, thanks for all your hints . . . you have nudged me into taking my keyboard project off the back burner.
- Uncle Bernie
https://github.com/schlae/replica-datanetics
One of the easiest keyboard replacement projects for the Apple-1 and Apple II without rare and expensive components.
In post #9, 'Macintosh_nik' wrote:
" One of the easiest keyboard replacement projects for the Apple-1 and Apple II without rare and expensive components. "
Uncle Bernie comments:
Oh, I know that project, but although it has its merits, it lacks some features I want for my keyboard solution:
- Apple II keyboard compatible form factor
- allows use of surviving Apple II keyboard encoder cards
- has optional on-board scanner electronics to allow use of a four wire interface
All this is deeply rooted in my past experiences with Apple II keyboards, all of the ones I ever had on my lab bench, had some bad keys but the encoder cards were good or could be repaired. Replacing key switches is tedious. And then the issue with the "Molex" connector, it seems they were not built to last an eternity ;-) . . . actually, the Hirose connector discovered by this thread also has a very limited number of cycles specified in the datasheet. And when debugging / reparing such a keyboard, the connector gets cycled very often . . . to replace key switches, the encoder card needs to be unplugged, and then plugged in again to test the result. Rinse and repeat . . . this is why repair work can wear out that connector, too. And my experience has been that it does not cause the connector to fail immediately, but after a while (several weeks or month) after the repair. Needless to say, such a refurbished keyboard cannot be sold to third parties as it might fail, and this generates bad feelings.
So I did start my own keyboard project a while ago but was not sure if I got the drills right - the slightest mistake made with positioning the key switches will mean a pile of useless PCBs that can only be thrown in the trash bin. So I wanted to do a "drills only" PCB but found no manufacturer willing to do this at a cost low enough. It might be just having holes in a PCB but nothing else might interfere with their manufacturing process, or takes extra caution, I don't know how they do it today, but back in the day (> 40 years ago) when I still was hand-glueing PCB layouts on mylar foil, getting a "holes only" test PCB to check if the components would fit, was reasonably cheap. Today, it's not so anymore. This is why I had put the project on the back burner.
But I made the decision to continue with the project and was working putting in the routing for the last two days and half of the nights.
Here is the current, still incomplete state of the work:
KB_PCB_draft.JPG
You can see that the form factor of this PCB is almost the same as the Apple II keyboard type with the metal frame (from which I took the photo in post #4). There are a few small deviations, though, indicated by the yellow ovals in the above screen shot, which were necessary to fit the scanner electronics. I had to extend the height of the PCB by ~6mm (the yellow line in the right hand yellow oval) but I hope it still would fit into an Apple II enclosure, the reason for this hope is that the original Apple II keyboard encoder card also sticks out by roughly that amount, so there must be some space in the enclosure in that direction (hopefully). Another deviation is in the upper left corner, I just need the added space there, to have a mounting hole in the corner. Plenty of strategically placed mounting holes are necessary to limit bending of the PCB under keystroke forces. The outer mounting holes to the extreme left and right side are useless for this purpose, they are just the equivalent of the mounting holes seen on the original Apple II keyboard, which has a metal frame to solve the bending issue. My idea is to use distributed mounting holes around the key switch array to screw the keyboard into a custom enclosure (Apple-1 use case) or, if used in an Apple II and it would bend too much, use these distributed mounting holes to screw on a metal plate on the solder side (with spacers of course) to stiffen the whole assembly. If these ideas do work, has to be found out.
Alas, I "sold" the last empty shell of an Apple II I had to some applefritter member, just the postage had to be paid. Now it bites me because I can't verify my current keyboard work with that shell I gave away. Lesson learned: there is no such thing as a useless piece of vintage computer equipment.
Anyways, if it does not fit into an Apple II, this is not my primary mission objective. I just want a new keyboard that works for each of my 15 Apple-1, so I can sell them at some later point in time. I plan to order 20 pcs of these PCBs without any further mechanical checks and no prototype. This is done in a hurry as I don't want to pay Mr. Trump's stupid tariffs he wants to slap on Chinese made components (JLCPCB). If he does so, this may kill a lot of small businesses here in the USA, who need to source their components and subassemblies from China. But this is another topic, not topic of this thread, so pleeeese, don't comment on my opinions on tariffs. Still, one final remark for those why are interested in economics: according to one of my professors I had in my MBA studies, the abysmal quality of Italian cars (and other Italian industrial products) of the 1960s and later was caused by the factories trying to make every component in house, if possible at all. This was caused by government policies (tax on every component sold between different businesses). Not to be confused with a VAT which does not tax the full billing price but only the "added value" (this was the solution many countries adopted to dodge the quality problem and inefficiency). I think that tariffs on components and subassemblies will have the same ill effect that caused said Italian misery. We hobbyists might be forced to etch our own PCBs again ... in the kitchen.
Now I must go back to layout work - it's rushed due to politics interfering with free entepreneurial spirits (again). Could be the outcome is too bad to be useful. But this is the risk I must take. Thank you very much, Mr. Trump. (If these PCBs don't work, I'll send them to the White House as a "gift" - same idea behind it as with the car owners who send the rusted off parts back to the CEO of the car company who made their P.O.S. car).
If you have any comments on the technical side of the proposed keyboard PCB, feel free to comment !
- Uncle Bernie
The idea is to salvage a PS/2 cable will all connectors from an old, worn out mouse. These cables have three wires inside plus the shield which is used as a ground return. Since old mice are essentially cost free (or a dime a dozen plus shipping on Ebay) , this is a great penny-pinching idea on how to minimize costs and still have a good, nice looking, robust keyboard cable. Also, much cheaper and nicer to have than a long 16 wire flat band cable. Hence, the encoder will sit on the keyboard input socket of the Apple-1 (or Apple 2), and have a PS/2 jack, unless a salvaged original Apple II keyboard encoder is used.
Here is how my lab rat looks:
Keyb_4wire.jpg
You might notice I had to add yet another IC, oh the horror, more costs ! But this was inevitable to put in a signal conditioning circuit to keep the output of the keyboard scanner from impairing the clock signal via crosstalk in the long cable. This add-on can be seen in the yellow oval in the above photo.
If the cable is shorter than 6 inches then the signal conditioner is not needed. So populating the PCB with its components can be avoided, saving about 70 cents from the BOM.
- Uncle Bernie
. . . my cost per piece is $6, which I think is OK. Shipping was almost half of this. So it eludes me why American PCB manufacturers can't (or don't want) to compete. They could take more for the merchandize and with the presumed lower domestic shipping costs still could arrive at the same price. USPS Priority Mail flat rate boxes have high weight limits so even bullet manufacturers use them for the cheapest way to ship those bullets to their customers.
Anyways, I'm anxious if these PCBs will work as expected. I did the layout in a hurry and used no schematic entry, working from my scribbled lab notes on the circuit. And there was some confusion with how to get non-plated-thru holes ("NPTH") from JLCPCB. The solution I used was to generate two drill files, one for NPTH, one for PTH (plated through) but still, lots of things can go wrong with these NPTH holes. Some people reported their NPTH holes were mucked up with solder mask in the holes. Hope this does not happen to me. Yet another risk factor is the tolerances of the holes for the plastic pins of the Cherry MX key switches. Hole too small, they won't fit into the PCB, hole too large, they will not stick by friction and fall out of the PCB again, and this is a major problem for hand soldering, as you would need three hands to do that. I had that issue with another PCB using MX switches.
Yet another risk factor is whether the whole thing is really mechanically compatible with the Apple II or Apple II+. Not having any of those machines anymore, I can't test this myself, and would need volunteers, preferably in the US (postage cheaper and no customs issues). I do have some sets of key caps ("Made in the USA" by Signature Plastics, who also made the original keycaps for Apple back in the day, or so I was told) and original Cherry MX keyswitches (Buyer beware of the Chinese copycat MX key switches, which are somewhat cheaper than the originals, but have lousy springs and contacts, and won't last, I wrote about those here on Applefritter, and provided photographic evidence).
Cross your fingers that this project succeeds despite of all the odds being stacked against me !
- Uncle Bernie
Here is the link to the new YAAK project thread:
https://www.applefritter.com/content/uncle-bernies-yaak-project-yet-another-apple-keyboard
- Uncle Bernie