• Please review our updated Terms and Rules here

IBM 5155 (Portable) RJ11 Keyboard Jack -- help needed

E

ENIAC

Member
Joined
Jun 11, 2020
Messages
10
Location
Milwaukee
Hi,

I am hoping to get some advice from someone who knows more about electrical sockets than me. Short version: what RJ-11 jack do I need to replace the one that the PC 5155 uses to connect the keyboard on the front panel?

Long version: My IBM PC 5155 had been giving progressively more frequent 301 keyboard errors on bootup. It works fine with two other XT capable switchable keyboards. I took the Model F XT keyboard apart to clean, look for corrosion or other problems, etc. That didn't help and the foam is in surprisingly great shape. Looked brand new inside. Wonderful build quality.

So then I got a multimeter from Amazon to do continuity testing and found that the problem was somewhere in the cabling. Pinouts at http://www.clickykeyboard.com/2013/apr13/012.jpg. The middle pin on the DIN ("data") is supposed to go to the white wire on the plug that connects to the keyboard PCB but there was no continuity. All the other pins tested fine.

Next challenge -- is the problem it the cable from the keyboard controller to the phone-style RJ-11 jack, or from the RJ-11 socket to the DIN that plugs into the regular XT motherboard connector. (If you want to know what I'm talking about look at this picture: http://www.clickykeyboard.com/2013/apr13/012.jpg. The RJ-11 connector at the top left is what the keyboard wire terminates in. To its right is the entire assembly that originally ran from an RJ-11 jack in the front panel around to a standard XT DIN connector at the rear. The picture shows the front panel trim that holds the jack as well as the jack itself.)

The pins in the RJ-11 did look a little bent, so I decided to remove it. There's a piece of trim at the front of the computer that holds what seems to be a standard RJ-11 jack in place with two very strong tabs. A cable runs down from the jack to a small hole in the bottom of the case, then around to the side and back to the DIN socket at the rear. That appears to allow for easy removal of the unclipped jack through the hole. Unfortunately the jack came apart when I was trying to free it.

Having pulled the remnant of the jack out through the base and freed the wires, I was able to confirm that one of the wires does connect fine to the middle pin on the DIN. Also, one of the pins on the phone-style connector has continuity with the white wire on the keyboard controller. So the problem was the now-defunct RJ11 jack failing to bring those wires into contact. It seems like the jack should a cheap thing to replace. Unfortunately I was stupid and did not photograph the jack before destroying it. And now I look for a replacement, I see there are many styles of RJ-11 jack for purchase and not all of them are likely to fit into the front panel piece that holds the jack in place.

I know I need 6 pins rather than the more common four. I see some jacks are called "6 position 4 connector" but as five out of six pins are used (no reset wire on my keyboard), I assume that I need to find a "6 position, 6 connector." So I think what I need is something like: https://www.zoro.com/leviton-rj-jack-and-module-jack-6p6c-usoc-la-41106-rt6/i/G4041905/. To fit optimally it should be not too deep and no wider in the back part than in the front part. For example, I think this one is too wide to fit. https://www.showmecables.com/icc-rj11-punchdown-keystone-jack-white

Then there's the question of how to get the existing wires into the new jack. It looks like the options are punchdown and toolless. I don't currently have a punchdown tool, but OTOH the tools are cheap enough. It seems like toolless tend to be a little longer and hence less likely to fit optimally in the existing clip.

So if anyone who knows RJ-11 jacks and/or the 5155 could take pity on me and give me a link to buy a jack that's likely to work I'd be very grateful.

A bonus would be if anyone knows which color wire on the cable from the DIN goes to which pin on the jack, as I (again stupidly) didn't photograph the old one and it used an old-school connector with two rows of connectors that doesn't match the new jacks anyway. But I can figure this out by contact tracing on the cable from the keyboard PCB to the RJ-11 connector and from the wires to the DIN plug, as I know the pinouts on each end.

Thanks for any help.
 
I am having the same issue occasionally after moving the machine or keyboard around too much. It is fine once the wire finds it's happy place, but I know this type of thing only gets worse with time.

Digikey has a ton of 6p6c rj-11 jacks. I've been trying to comb through them all, but it's not easy unless you are a wizard with their search filter. I'll keep looking, but definitely if someone has found a source for these before it would be a huge help. If you do find a part number of one that looks right, but don't like the website, definitely try putting it into Digikey or Mouser and see if they have it.
 
Looking closely, I do did find some more clues. I think an Rj-11 with six connectors may be designated as an Rj-12 so searching on that will bring the possibilities down.

The top of the jack survived intact, and it seems to be about 9.5mm across. I remember the sides being pretty much straight, so something around 10mm at the widest part would fit well into the housing.

Versus most of the modern jacks it is less deep because the wires were soldered into holes rather than being pushed into compartments. Going by the top, I think the depth is around 14mm. A longer jack would probably fit but the plastic tabs wouldn't snap shut behind it so it might be less secure in place unless glued.

There are small tabs at the top and bottom of the mount, but they are much smaller than the ones on the left and right. So the vertical dimensions only matter for the part that sticks through the hole at the front.

The other potential issue is that the cable didn't have much slack in it, and the old jack had the wires coming in at the rear on the bottom. If the replacement has the entry point at the top I might need to solder a small extension for them to reach. On the other hand, the most compact jacks seem the most likely to fit well, and they are intended for PCB mounting with pins on the bottom. So soldering the existing wires straight to the pins might be easiest. Maybe this one: https://www.digikey.com/en/products/detail/amphenol-commercial-products/RJE0166001/3083548.

I understand that people wouldn't want to take a strong position and say "this jack will fit" but any pointers to help us narrow down the search would be appreciated. The things are cheap enough I could order five or ten of them and see which fit best.
 
Before you go all replacing the cable, try this: open the keyboard and detach the cable from internal connector, spray contact cleaner, reattach. This often fixes the issue.
 
You know what would be nice; a converter that would allow you to plug in a standard 5 pin DIN keyboard. I've seen Portables without keyboards before. I have a Zenith Z-161 without a keyboard, but it has the standard female DIN socket.
 
Casey said:
You know what would be nice; a converter that would allow you to plug in a standard 5 pin DIN keyboard. I've seen Portables without keyboards before. I have a Zenith Z-161 without a keyboard, but it has the standard female DIN socket.
Click to expand...

You can plug in a standard XT keyboard. The 1555 uses a regular XT motherboard, so there's a cable that plugs into the DIN at the back and goes around to the jack at the front. Unplugging it lets you use a regular XT keyboard or, I assume, one of the converters mentions. But the regular keyboard forms the bottom of the case which is why I want to get it working.
 
I managed to figure out the pinout for the RJ11 jack. With the tab at the top, and pin 1 on the left:

1: Light Blue wire (earth – come from the shield of the DIN)
2: Yellow wire (+5V)
3: Green wire (ground)
4: Red wire (clock)
5: Black wire (data)

Verified this with continuity testing from the DIN plug, and with cross referencing with the colored wires from the keyboard PCB visible in the plug.
 
The hunt for a new jack was more frustrating. Good news is I managed to remove the clip that holds the jack. To do this, just use a small flat screwdriver to pop out the tabs that hold it on the top and if necessary (using the access hole in the bottom of the PC) the bottom. Pops right out. Don't try to free the jack from the clip while the clip is still attached, which is how I broke my jack.

I simplified the search by looking for RJ-12 jacks, which have six connectors vs. the 4 connectors of RJ-11. To fit in the clip, a jack should be: 14mm wide by 12 mm deep. Height isn't as much of an issue. Most modern jacks are much bigger, unfortunately. Keystone-style jacks, which include their own clip to fit into standard panels, are so big that they could fit into the case INSTEAD of the original clip. Pretty much fill the hole, and with some kind of putty might be held in place.

14 mm width allows for very little plastic on either side of the connector. 12 mm depth is even harder to find. The smallest jacks were all designed to be mounted on PCBs. Of the selection of jacks I ordered, this one came closest to fitting into the clip. Fine on width, fine on height, and the pins are accessible in the clip and point down. https://www.digikey.com/en/products/detail/kycon-inc/GLX-N-66M/9990103

The problem is the depth of the jack: 13mm. The depth at the top left and top right needs to be even less than 12 mm because the clip slopes in at the top. So I got out some sandpaper and sanded away a little from the front and corners of the jack so it would fit in the clip. Then I soldered the wires to the pins.

Didn’t work though. I assumed it was my crappy soldering, but continuity tests work fine from the DIN to the pins inside the jack. OTOH, there’s no reliable continuity from the keyboard PCB side to the pins coming from the jack. But there IS continuity from the keyboard PCB to the pins in the plug. Conclusions: no contact between the plug contacts and the jack pins.

Before I did the sanding and soldering I verified continuity from the keyboard PCB to the pins on the jack. But somehow that’s gone. Two theories: (1) even though the locking clip still holds the plug securely in place, maybe sanding off a small part of the jack messed up the angle enough to break contact. (2) maybe the sanding process covered the pins with dust. I’ve ordered some contact cleaner, which seems like a good think to have around in any case. Assuming the contact cleaner doesn’t work, I may give up on my quest for a jack small enough to fit in the clip and just look for something to hold a larger modern jack secure in the case hole.
 
I'm just returning to wrap up the story. It turned out that I'd forgotten to mentally rotate the pin order for the front of the plug vs. the front of the socket. The pinout i gave above is correct, read left to right from the viewpoint of the FRONT of the connector or the BACK of the jack. But when investigating the problem I also found that the jack I was using (https://www.digikey.com/en/products/detail/kycon-inc/GLX-N-66M/9990103) seemed kind of flimsy. Even testing one that I hadn't sanded down to fit in the clip, I could see that it put a lot of stress on the wires when pulling out the plug and the continuity to one of the pins was flaky. Only the very smallest jacks come close to fitting in the clip, and maybe they are just cheap and nasty.

So I went with plan B and used a much more robust jack held in place with adhesive putty instead of by the original clip. However the PC had already lost one of the strips of trim on the side of the front so wasn't going to win any prizes anyway. I still have the clip, and could always go back to it later if I ever want to make the appearance perfect.

The jack I used has barrel connectors and is about the right size to fill the hole in the front panel once the clip is removed. It was really easy to attach the wires and works great, though the original look is spoiled slightly. It just needs the putty because it's designed to clip into some kind of panel.https://www.digikey.com/en/products/detail/te-connectivity-amp-connectors/216000-2/2054272

Hope this helps someone. The real lesson is: if you need to work on the keyboard jack for any reason, just free the clip that holds it. Don't make my mistake of trying to free the jack from the clip without taking the whole assembly out of the computer first. If there's a current source of high quality jacks small enough the fit in the clip then I wasn't able to find it. Just don't break the original!

The other mistake I made was fully disassembling the keyboard to look for potential fault problems with foam, oxidation, etc., rather than just waiting for my multimeter to arrive so I could confirm that the problem was in the cables/plug/jack. It was fascinating to see the insides of the Model M, but getting the metal plates back together with all the parts in the right place took a while. So don't don't separate the metal plates in your Model M keyboard until you've eliminated cabling issues.
 
Are you sure your pinout for the 6P6C connector is completely correct? I'm just wondering because female/male back/front issues aside, the IBM 5155 is basically an XT in holiday dress, and according to https://old.pinouts.ru/InputCables/KeyboardXT_pinout.shtml, the pinout includes a RESET pin. So I'm thinking your light blue (1) and green (3) wires may not in fact both be GND, and one of them could be the RESET pin. I would be very surprised if those pins deviated much from what's on the ordinary XT keyboard connector.

Second thought, I've just now found this at -0°: http://minuszerodegrees.net/5160/keyboard/5160_keyboard_interface.png
Note that that pinout differs from yours but matches the 5-pin DIN pinout from my first link.

Also, the actual 6P6C registered jack of course has six pins, so it's still not entirely clear how that looks on the actual 6P6C connector as opposed to on the motherboard header. Is the sixth pin just NC (not connected)?
 
PS:
I've only now remembered that I actually have the male-5-pin-DIN-to-female-6P6C-jack adapter cable also. Presumably this came with the actual 5155 Portable PC – it allowed IBM to use a standard 5160 motherboard but still present a 6P6C RJ-style keyboard port to the user. (Though just why IBM preferred that isn't entirely clear to me.)
I got a little confused by the fact that the DIN connector's pins aren't numbered in order. Their actual order of 1, 4, 2, 5, 3 seems to be an evolutionary relic to do with a degree of backwards compatibility with an earlier 3-pin plug DIN design.
I also got a bit confused by the fact that the wires in the DIN-to-6P6C adapter cable have colours (blue, yellow, green, red, black) that differ from wire colours encountered in the actual keyboard cable. To answer my own questions, yes, the sixth pin on the 6P6C is NC, and your pinout is substantially correct (for the adapter cable).
To talk about the actual keyboard cable, from 6P6C plug to keyboard, its cable splits in two after it enters the keyboard: A black single-wire cable connects/is screwed to the keyboard PCB, presumably to reduce resistance for the ground connection, to really ground out well what is after all a capacitive keyboard design. The other lead has four wires that mate to a big blue Berg connector (branded Berg, but a different design than floppy connectors you may know as Berg connectors). NB that that's an internal connector, inside the keyboard, not normally user-accessible. We might call that blue Berg a 7P5C connector, but that's purely descriptive; I don't know that it was ever called that. Depending on which way you look at it, the pinout might be NC, CLOCK (black), DATA (white), GND (red), NC, VCC (brown) and NC, or the reverse. It's a little confusing, because the key side of the keyboard PCB is the opposite of its component side, so the connector and components are sort of mounted on the back and it's not clear which way is up. This way assumes that what looks like a tiny bar on the blue Berg is in fact a pin 1 indicator.
I've also found this picture: http://minuszerodegrees.net/5160/motherboard/IBM 5160 - Keyboard interface circuitry.png. It mentions that only a "type 1 version of the 83-key keyboard" uses RESET, which may explain why that's absent/NC on the 5155 keyboard, since again, the 5155 is a 5160 in holiday dress. Unfortunately the J9 connector shown on that doesn't match the blue Berg or 6P6C pinout; it's its own thing, probably to make the confusing 5-pin DIN a bit clearer.
 
Last edited:
You must log in or register to reply here.
Back
Top