Author Topic: Unicomp/IBM Franken M  (Read 5526 times)

0 Members and 1 Guest are viewing this topic.

Offline Magnusian

  • Thread Starter
  • Posts: 67
  • Location: Ontario
Unicomp/IBM Franken M
« on: Mon, 17 December 2012, 19:35:00 »
As I have both a spare IBM 1390120 and a spare Unicomp Customizer 104, someone mentioned in my barn-find of making a hybrid of the two boards, namely transfering the backing plate from the IBM board to the Unicomp. I decided go with my older Unicomp as the primary donor of hardware since it's had a double keypress issue that's been driving me up the wall and would require disassembly to fix. I've disassembled both boards and have from the onset noticed a large number of differences between the two, some purely cosmetic, others being more quality-of-build related.

Here's some bad photos (sorry, took them with a gopro and I have shaky hands) with captions documenting their disassembly.


Oh dear, two naked keyboards side-by-side, what follows might not be safe for work! Aside from the 3 extra keys on the Unicomp, the LEDs, and the location of the PCB, you can actually feel a difference in the weight of both sides of the shell between the two keyboards, the Unicomp's lower shell has a thinner material thickness and overall it's lighter. Oddly enough, and not pictured, is while the lower shell for the Unicomp has the Lexmark-era drainage holes, which is to be expected. The back of the case no longer has the divot in it for the SDL cable, the two posts which are meant to help hold the older PCB in place are still there, though the two tabs that hold the PCB are not. I wonder why they were left in place when the rest of the shell has been modified for the newer design.


Close-up of IBM 1390120, this one's birthday is 23JAN87, that makes it 4 months and 14 days younger than me! I'm not sure who's weathered the years better, me or the keyboard...


Closeup of Unicomp Customizer 104 (UB4044A), this one's birthday is 6/27/2007, dunno what that means, 6th day of the 27th month of 2007? That doesn't sound right...


Backing plate of IBM. Looks like someone spilled something on it and never bothered to clean it, nothing a wire brush on the end of a drill can't fix!


Backing Plate of Unicomp. You can immediately tell a difference between this and the IBM, there's some weird cutout in the backing plate, and the rivets look like they were melted by someone who shoved a nail into a soldering iron and used the head to melt the rivets (you can see the marks from a nail head in the rivets), they aren't as uniform as the IBM rivets. I'm not sure if I like the fact the keyboard has a more human touch to it than the IBM (which likely had a machine do the rivets), or if I'm non-plussed about the clear difference in quality of build. I know around the time this board was assembled that Unicomp was just about on death's door financially so I'm not going to fault them for having to do things cheaper to stay alive.


The two backing plates side-by-side, I know it's pretty blurry but you should be able to tell that the Unicomp plate (left) is almost half the thickness of the IBM (right) one. The 1390120 is a much heavier keyboard all around, and quieter as well, the difference in backing plate thickness must be part of that, we'll find out once it's been put into the Unicomp.

Time to remove the rivets from two keyboards!


Somebody puked IBM parts all over my living room!


And then they puked Unicomp parts! It will probably be hard to tell, but if you look at the IBM barrel plate, it has 3 rubber tubes located along the edges, the Unicomp lacks these and there is no way to fit them into the newer barrel plate. I suspect that these also affect the amount of noise generated.


IBM-made hammer and spring assembly.


Unicomp-made hammer and spring assembly. Identical in every way except hammer colour, which is white instead of translucent. It also has a different mould number on it. Some of them are a little rough around the edges, as if they hadn't spent enough time in the part tumbler getting cleaned up. The Unicomp had a lot more dust and random crap between the bottom of the barrel plate and the rubber mat, including some debris right around the ; key, I wonder if that's why I was having a double keypress issue with that key. Keep in mind that this particular board has been through the hell that is a college student's apartment, I often wonder how I survived, let alone this keyboard.


I use a razor blade to sheer off the plastic rivets, if it wasn't for the fact over half of the ones on the 1390120 had popped off of their own accord over the past 2 and a half decades my fingers would be weeping bloody tears of pain. The Unicomp rivets were much harder to remove, probably due to the extreme variation in thickness. Hard to tell, but IBM rivets are on the left, Unicomp ones are on the right.

Tune in next time where I start drilling out the Unicomp's barrel plate and assemble the hybrid board. Aside from the backing plate from the IBM, I will also replace all of the Unicomp hammers (less 3) with the IBM made ones. 58 bolts, nuts, and washers, why do I do this to my poor hands?
Model M 1390120 1986  | Unicomp Customizer 104 1-Piece Keys | Franken M IBM 1390120 1987/Unicomp U4044A 2007 hybrid | CM QFR Cherry MX Greens

Offline fohat.digs

  • * Elevated Elder
  • Posts: 6470
  • Location: 35°55'N, 83°53'W
  • weird funny old guy
Re: Unicomp/IBM Franken M
« Reply #1 on: Mon, 17 December 2012, 22:08:02 »
My prize hybrid Model M is an ancient 1390131 but I do have 2010 Unicomp springs in it.

I might drill holes in the 1390120 case for the LEDs to shine through, and use case and plate from the IBM, and most of the rest from the Unicomp.
"The Trump campaign announced in a letter that Republican candidates and committees are now expected to pay “a minimum of 5% of all fundraising solicitations to Trump National Committee JFC” for using his “name, image, and likeness in fundraising solicitations.”
“Any split that is higher than 5%,” the letter states, “will be seen favorably by the RNC and President Trump's campaign and is routinely reported to the highest levels of leadership within both organizations.”"

Offline wcass

  • Posts: 506
  • Location: Fort Lauderdale, FL, USA
Re: Unicomp/IBM Franken M
« Reply #2 on: Mon, 17 December 2012, 23:13:26 »
thanks for the report; good info.

Offline Magnusian

  • Thread Starter
  • Posts: 67
  • Location: Ontario
Re: Unicomp/IBM Franken M
« Reply #3 on: Tue, 18 December 2012, 02:04:51 »
I just finished putting the board back together. Typing this post out with it right now. I'll have to post a picture filled update in the morning because it's 3am and I'm gonna hit the sack.

Still didn't completely solve the double-press for the ; key, at least I think so. It did it a few times and now I can't make it happen again when it would do it at least 1 out of every 4 times I pressed that key.

The Franken M's louder than my bolt-modded 1390120, but slightly quieter than my stock Unicomp.
Model M 1390120 1986  | Unicomp Customizer 104 1-Piece Keys | Franken M IBM 1390120 1987/Unicomp U4044A 2007 hybrid | CM QFR Cherry MX Greens

Offline Magnusian

  • Thread Starter
  • Posts: 67
  • Location: Ontario
Re: Unicomp/IBM Franken M
« Reply #4 on: Tue, 18 December 2012, 19:48:20 »
I was gonna wait for someone to make a post in here so I could avoid double-posting but that didn't seem like it was going to happen.

Anyways, here's some pictures of the Franken M going back together.


I'd put all of the bolts in and started on the hammers when I'd realized that I'd forgotten to take a picture! I wasn't in the mood to dump the hammers out again so I could take a shot of the top of the barrel plate. Please ignore my brother and his attempt to become more like Al Bundy.


Sorry for how blurry the image is, my hands were shaking something fierce after having drilled all of those holes, driven all of those bolts, and placed all those spring and hammer assemblies. I ended up using 100 of the 1390120's springs, one of them appeared to have a dirty hammer but upon closer inspection it was actually burnt! I have no idea how that happened as there's no evidence of burning on the rest of the donor board. The 4 white hammers are Unicomp ones. You may notice that there's even bolts along the bottom-most row rather which differs from the typical bolt mod instructions. I've found that at least with the M2x8 bolts I use, they don't interfere with the bottom shell, though drilling them on a Unicomp barrel plate is interesting to say the least, as the rivets they replace are right along the very edge of the plastic.


Rubber mat and membrane in place, this is one of the hardest steps, as it's right before the hardest step.


The hardest step for me is putting the backing plate on without jostling the springs out of place. I've found it best to randomly pick 10 bolts that cover the top, middle, and bottom of the board, once the barrel plate is forced into the curve of the backing plate the hammers can't come loose and testing is possible. Up until that moment it feels like the slightest bump will make the whole thing go to pot... I completely forgot to take a picture of this, consider this particular setence a placeholder for now as I'll probably have the shell off again soon and will take a picture of all of those tiny bits of metal that hold it all together.

I have figured out how to reliably reproduce the double registration of the ; key, it happens the most often when I try to type : rather than ; though it'll occur for either. I'm not sure how to correct it. I may try the controller out of my other customizer to see if it's related to that, as the key itself seems to be working correctly from a physical standpoint. When putting the board together I made sure specifically to clean that area of the membrane and rubber mat, I must have overlooked something so I guess this keyboard will be coming apart, again if a controller swap doesn't work. If you notice in the picture of the membrane, there is a bolt that is to the top right of this key's barrel, if you look at the picture of the barrel plate pre-rivet removal for the Unicomp, this bolt is located exactly where the only missing rivet used to be. Bolt is tensioned to "finger tight" exactly like my bolt-modded 1390120 which does not experience this particular issue.

Any ideas?
Model M 1390120 1986  | Unicomp Customizer 104 1-Piece Keys | Franken M IBM 1390120 1987/Unicomp U4044A 2007 hybrid | CM QFR Cherry MX Greens