The Model M13 is a very interesting variant of the Model M - it looks so clean and assuming like a standard Model M, but on closer inspection, you have got this little red nub in the middle of it all to play around with. As I have talked about these on forums and Reddit, more than a few people were surprised to discover that more variants than people were used to - ie, the beige, black and industrial versions - exist. So with that in mind and because I own both, I thought it would be interesting to take a little look the differences between your average Lexmark M13 and the rarely talked about Unicomp On-The-Stick.
First, we have an example from Lexmark. Officially named the IBM TrackPoint II Keyboard (or IBM Enhanced Keyboard with TrackPoint II, depending on the source), P/N 92G7461 is the bog-standard beige Model M13. My example was made on the 23rd January 1995, putting it around the middle of Lexmark's production period of this keyboard. You can read a bit more about my specific example here.
Second, we have an example from Unicomp. Officially named the Unicomp On-The-Stick, the keyboard was still actually made for IBM as a likely FRU order to replace an original M13. Made on 26th April 2004, this is from the latter half of Unicomp's production run. Despite coexisting with the EnduraPro since 2000, the M13 left production around 2009, thus making the EnduraPro the sole pointing stick keyboard Unicomp still makes. Anyway, you can read a bit more about my Unicomp M13 here.
Let us start with the obvious - the front casing. At a glance, the cover sets are pretty much identical. M13s (at the least the beige ones) are known for having a very over-textured finish compared to their older brothers, and that is indeed visible on both keyboards. The patterns are close enough that there should be no doubt these were made with the same moulding.
The fact that the Unicomp M13 retains the supposedly redundant oval badge insert from when these were largely made only for IBM reinforces the likeliness that these were from the same moulding. The only lingering question from this is why did Unicomp not use the mould for the Lexmark-branded M13s that have a rectangle logo (as they did with the Unicomp Classic and Classic Trackball).
Moving on to the lock-lights, another difference appears. The Lexmark M13 uses the traditional bottom-left aligned LEDs, whereas the Unicomp M13 uses the 42H1292-style top-middle alignment. The Lexmark's overlay is the traditional pebble (as Unicomp calls it) three-segment overlay, whereas the Unicomp's overlay is a more unique and clean panel with small icons and a simple IBM logo. And finally, the last observation here is on the keycaps - note how Unicomp added a groove on the keycap that flanks the underside of the pointing side:
I will make this part brief as there is not much to say. They sound really similar except that the Unicomp's is noticeably more pingy for some reason. I will leave it up to you to decide which one is better.
The polarising pointing stick differences are perhaps the most interesting for people as it pretty much decides which one works better than the other. Starting with the Lexmark's implementation, it uses IBM-developed TrackPoint II technology. TrackPoint II was the first implementation of the TrackPoint as we now know it, replacing the original TrackPoint that was just a name for the trackball/mouse combo for the IBM Personal System/2 L40SX laptop and the red trackball on the Japan-only IBM ThinkPad 220. If you have ever used the red 'nipple' on an IBM or Lenovo ThinkPad, you have used this. The TrackPoint is a stationary stick to you apply pressure to flag what direction you want the computer's cursor to move in - ie, a strain gauge that translates its applied force (the 'strain') value into a usable value.
The physical appearance of the TrackPoint II stick is fixed stick emanating from in between the G, H and B keys. As a side note, this sort of TrackPoint stick was used by Maxi Switch for their M13s too. However, being an older (revision 2) implementation of a technology that is currently in revision 4, there are some weaknesses that TrackPoint II suffers compared to TrackPoint IV found on current ThinkPads. Namely, TrackPoint II lacks the negative inertia feature that causes the cursor to react faster when it is pushed and ultimately makes the technology more capable with high-resolution displays.
By contrast, Unicomp takes a different and conceptually simpler approach to their pointing stick. The Unicomp Pointing Stick is based on the original patent that describes an early version of the TrackPoint by Lexmark, in which strain gauges and force-sensitive resistors are considered. As you might assume; if IBM and Lexmark ultimately used strain gauges, Unicomp used the force-sensitive resistors. I will try sparing you the potentially esoteric comparison between them (check out this article by AZoSensors for an actual comparison), but in short, the real difference is that Unicomp's stick implementation has to move when its pushed to generate a value.
With that last fact in mind, the groove on the "B" keycap as pictured earlier is a lot more understandable if the pointing stick has to move about in a very confined area. However, from my experience, it is still a far from a perfect solution - the Unicomp Pointing Stick falls to crawling speed when pushing it in the up-left, up-right and bottom directions. Software acceleration will improve it, but you can only go so far.
On the surface, both keyboards appear to have the same sort of mouse buttons to compliment their pointing sticks with. However, Unicomp in fact claws a few brownie points back here. It is quite well known that the standard Model M13's mouse buttons are quite 'gummy' when you use them, featuring weak tactility and a slight squish near the end of travel. The Unicomp's buttons feel a lot better, almost like a typical micro switch you might find on an actual mouse button or game pad.
As evidenced with that recording, you can clearly hear the difference in the mechanism. The Unicomp's button switches are much louder and stronger.
There is not a lot that can be told from looking at the back except the obvious feature the Unicomp keyboard lacks - a PS/2 port for a mouse pass-through.
I can imagine that this was a useful feature back in the '90s, giving you a way to change between TrackPoint II or a discrete mouse without reaching behind your PC. In today's world, you can live without it given USB ports on the front of your computer are a thing.
The stickers are what you would expect from both companies.
In the early days of Unicomp, whether or not they used Lexmark-era model designations was ambiguous. Here, the Model M13 designation is no where to be seen.
By the time that the M13 came out, Lexmark had already largely moved towards using fixed cables as we see here (the only SDL-based Model Ms still being made during this time were the M4, M4-1 and M5-2). The choice of cable is different though. The Lexmark M13 still uses a largely coiled cable that results in it being a manageable length most of the time and able to stretch when needed. Whilst not colour-coded, the plugs feature a bold symbol denoting whether the particular end is a keyboard or mouse.
Unicomp uses a straight cable with coloured plugs and shallow symbols to distinguish the two PS/2 plugs. The straight design of the cable will mean that it is more likely to kink and/or tangle itself. The cable is also extremely long and I did not have a reasonable sized and well-light stage to take a good photo of it from.
As I explained in the introduction, these very similar looking keyboards are in fact quite different if you take only five minutes to prod around. In regards to which one I think is the best, I would choose the Lexmark.
The Unicomp puts up a strong fight. They feel and sound similarly enough and Unicomp vastly improved the mouse button switches. But a 'real' TrackPoint like what the Lexmark offers is untouchable despite lacking negative inertia. The Unicomp Pointing Stick's design is limited by how close the keys are, and given that the device is supposed to be a keyboard first, there is no real way to rectify this other than software. To that end, I am working on such a solution but my advice to Unicomp would still be that of recommending they upgrade the technology as soon as possible on their EnduraPros.