Fixing up the Bennett
Some quality of life improvements on a very old index typewriter, making it mostly usable, if not exactly suitable for novel-writing.
Lately I've been spending a good bit of my free time poking around with type element machines. Most recently that's included this teeny tiny machine; a Bennett! It's pretty difficult to share in words just how compact this thing is. Despite being absolutely tiny, the inner workings are relatively simple and it's pretty straightforward once you're under the hood. There's plenty more work that this one needs, but it felt like it would be a fun snapshot of the progress and to show it in some detail.
I had agreed to pick this one up as-is, so it needed a little bit of TLC before it could reach those blistering double-digit words—per-minute speeds. In case it isn't painfully obvious yet, even the smoothest and most tuned example of this style is unlikely to be much fun to type on. By comparison, the Blickensderfer might as well be a Selectric...
Undoubtedly there are things to praise here however, so let's not deride it too much before we even get started. When I first got this example there were three primary issues; the droopy frame, no "W" presses, and the element would get stuck when using special characters. By the end, all of these had been resolved or at least some— what mitigated. As mentioned, there's still plenty to do, I'm sure the eagle-eyed among you will have noticed the missing "7" keytop and the abysmal state of the platen, however these will get addressed later. The primary purpose here was to get this one at least mostly operational, enough to justify taking it out for brief typing sessions.
Starting with the simplest, when pressing some of the top-row keys, most egregiously with the W key, the bottom-row key of the next row medially would actuate along with it, resulting in the wrong character being pressed. In the case of the W this would mean that "when" would become "chen"! Chat an embarrassment!
Removing the two thumbscrews and looking at the key levers in action it was pretty clear what the issue was, as the W and C keys shared a contact surface. Depending on how you applied force to the key it would often carry the neighbor along with it. This is intentional for the bars in the same stack (ie, W,S, & X) but a little extra friction was tripping up the next bar. As is so often the case, a little crud in between the bars was the culprit. Just a little flushing with some solvent and bit of oil-solvent mixture to keep things lubricated freed up the two bars and now when pressing the one, the other stays put. Having recently mixed another bottle of solvent—oil I've been reminded of how handy this stuff is.
Next up was the droopy frame. The Bennett has a metal frame that is lifted from a sheet metal bottom by some washers. Without these (one of them had crumbled away and left a gap) the whole frame would flex down when typing on the left side of the keyboard. This was probably the hackiest of workarounds, but so far it's been ok. Using some scrap from the bumper pads (thin, adhesive rubber sheeting) I built up a makeshift washer between the bottom and frame. In the future I think I'll print a proper copy and replace all of them, but for the moment I just wanted something to get it working. In the photo here you can just see the edges of the clear rubber bumper sheeting sticking out.
Lastly, the indexing on special characters. The Blickensderfer has a clever system that forces the whole arbor and head to align and trusts that the element is aligned to it. The Bennett on the other hand has a series of holes in the element itself, using a pair of pins on the card guide. This positioning on the guide however means that it is very sensitive to being bent or distorted. In this case, that's exactly what had occurred, and the guide needed to be bent back a bit. Removing the carriage and a tiny bend later had the characters aligning correctly and the element returning to neutral as usual.
In the photo here you can see the two pins on the guide. Originally, there was a bit of a twist and if you looked down you could see the top pin wasn't facing straight to the element, adding some resistance when it would attempt to return to rest.
Bending the card guide back into the proper alignment evened up the paper and now allows the element to return to the normal position. Looking carefully, seems that someone may have had a similar idea in the past, as there are some minor tooth-marks on the left side. From this view, you can also see an interesting roller that is on top of the post, this is in fact a friction-reducing mechanism that engages the ribbon drive from the underside!
