New typeshuttles for the Hammond - steps forward, steps back

It's a lot like regular new-product development. Steps forward, and a step back (or two). Forward was that printing the strip with DLP resin technology gave incredibly crisp letters on the strip. 

Whereas the filament-printed strips still show the layering, on the scale of typewriter-characters the resin-printed strip is absolutely crisp. In actual typing, it does show that the resin-printed shuttle is better quality. Because the actual typing goes on paper and through a ribbon, the difference on paper is not as large as the difference when comparing the strips.

The ribbon was a hand-inked 'home-made' narrow one. To check the impact of this home-inked ribbon, the Hammond Multiplex was temporarily fitted with a new, commercially bought half-inch ribbon.

That changed the whole look of the machine - it didn't however improve the typing much. (It actually made it worse - so back to one of the narrow home-made ribbons!)

The DLP resin-printed strips were unfortunately also less firm. They really need the metal flange to stay in shape. That lack of stiffness probably is the cause of these new crisp strips again binding in the wheel and jamming the machine. That problem had been completely solved for the FDM-prints, so that's clearly a step back again.

With all the 'mucking-about' and exchanging ribbons, the relatively weak aluminum type-shield broke and lost a small section. This is the cause of the 'echoing' under the characters in the above image. More new shields to be made.

From the experience so far, it may turn out that for simpler typefaces (low-cost) FDM technology is the best compromise to create a working Hammond type-shuttle that produces acceptable text. However for more complex typefaces (e.g. Attic), the crisp detail of a DLP resin printer is needed - the added problems of resin-printed parts yet to be solved.

It's definitely product-development - and continue to be impressed with the fine balance of parts and process-parameters that Hammond achieved in the 1880s!

Re-assembled carriage of Mignon 4 index typewriter

Not the most readable typeface, I'll admit - fun to be able to use though :)

A few more pictures would have helped find the spot for the small line-feed helper-lever, but it went together again and now all very clean.

Still have to manufacture two new springs for pressing the ruler onto the platen. Turns out that those springs are essential to keep the paper properly onto the platen, to solve the ghosting or shadowing.

The one spring that I left out here on the board. It's definitely an original part, but it seems way over-the-top to have three of these stiff torsion-springs. The pressure of two of these is enough - better to have a symmetrical load on the rollers too.

Little wonder that so many Mignon machines have flat feed-rollers; the spring-pressure is weirdly high.

This machine is surprisingly fun to use; a bit cumbersome, but fun. These were genuinely used as typewriters in their day; e.g. the keys of this specimen show a wear-pattern consistent with actual, serious use.

With a bit more tweaking, a 'cheap-n-cheerful' Mignon is turning out to be a fun-machine. Maybe even worth getting the platen re-covered (extravagance for an old index-typewriter!)

:)

AEG Mignon typewriter with Rheinmetall logo

In the process of giving a Mignon index typewriter a thorough cleaning, the carriage was taken apart. This particular Mignon was made by AEG sometime in the late 1920s and in keeping with the overall heft of the machine, the carriage frame is cast iron.

There is a shield with the machine serial-number riveted to the casting and the middle-tie has lettering "SMa" over the number 6001. The 6001 may refer to the pattern or drawing-number, the meaning of "SMa" could be a simple as "SchreibMAschine" or typewriter.

Castings sometimes also have a mark of the foundry, when not manufactured by the typewriter-factory itself and sub-contracted out. This casting indeed does have such a mark. Unexpectedly, the logo on this casting is the Rheinmetall logo.

Clearly marked with the circle and inverted quarter-circles, as also used strikingly on the front-panel of their typewriters of the same period.

(wordmark image from a user manual on The Classic Typewriter Page)

The factories of both companies were also 'next door' to each other, the Rheinmetall site at Sömmerda is not too far away from the AEG typewriter-factory in Erfurt. 

It's thus likely that in the 1920s Rheinmetall not only made their own typewriters, but was also a parts supplier to AEG, the later Olympia.

Refreshing a Mignon typewriter character field or keyboard

The 'keyboard' or character-field of a Mignon index typewriter is where the desired character is selected with the pointer. This map or field is a curved plate, covered with clear celluloid. After 90 years or so, on some key-boards, the celluloid is not so clear anymore - and what was worse on this specimen; the old plastic started to crack and crumble.

Because it was falling apart and catching the index-pointer, this 'keyboard' was taken apart for replacing the clear plastic with new. There are a great many tabs around the rim; after carefully bending these open, the stack of a fibre backing-plate, aluminium sheet with the characters and the celluloid top-sheet can be taken out.

A replacement top-sheet of about 81.5 by 128.5 mm was cut from clear plastic sheet. With this new sheet, the difference in legibility is very noticeable. Bending back all the tabs very gently, fortunately only one tab broke off. (If this new plastic sheet will last another 90-odd years before its starts to crumble and needs replacing, then the keyboard has enough redundancy in tabs to last it at least a few centuries ;-)

The index-pointer no longer snags on the keyboard-cracks; smoother sheet, smoother working of the cheap-n-cheerful Mignon :)

New typeshuttles for the Hammond - more iterations

It's getting there.

It's iterative development, almost every prototype uncovered something new to improve or a new fault-mode.

Now at version 14 of the reproduction typeshuttle, the mechanical function and interaction with the machine is fine. The design is now at the stage where precision of the type itself needs improving. The rapid prototyping strips can't create proper stems of e.g. the 'p' or 'h', as can be seen from the type-sample.

Original at the left, reproduction at the right.

The Hammond typewriter itself is an impressively complex (and finicky) piece of machinery - my Multiplex needs some work done too :)

Mignon index typewriter - oiling of roller wheels

The Mignon index typewriter is a pretty common machine still, many were sold and they are sturdy. (To quote Helen Howes: "They don't compost well, so there are lots still kicking about.") 

For an index typewriter it is also somewhat over-engineered - the cast-iron frame and overall sturdiness impresses. These machines do have several 'weak spots' however; it's prone to rust and the feed-rollers generally have deteriorated. The use of pins also is unfortunate, e.g. removing the platen can be hard. Nevertheless, as a cheap-n-cheerful common machine they're a good subject for tinkering. Well, actually these aren't all that cheap - despite being 'common as mud' overhere, their unusual appearance probably increases the market for these and they do sell for more than a comparable type-bar machine.

Anyhow - this late 1920s Mignon Modell 4 got a complete rebuild of the machine-base, but still did not work very well. The carriage moved very stiff, as it if was binding. That probably was the reason someone, decades ago, mangled the spring to increase the pulling-force. Bad.

Inspecting the carriage rail and guide, these all seemed straight. The carriage slides with its cast-iron frame over a steel rod in the front and rests on a roller at the back.

Could it be as simple as a stuck roller, that was being dragged over its race? The carriage of a Mignon can be taken off very easily; place the left margin-stop lever in an 'up' position to clear the central-stop, press down the 'space' and simply slide it off to the right.

This wheel was indeed firmly stuck. With a bit of oil, temporarily loosening its bolt and forcing it round it quickly whirred round freely again. One roller oiled.

The Mignon actually has two 'space' or 'carriage-release' controls. On the machine base, the space-key (a) will space the carriage when pressed. If it kept pressed down, it acts as a carriage-release. On the carriage, the lever (n) releases the carriage allowing it to be moved freely. Giving this carriage-release lever a single 'press' will however space the carriage.

When pressing down the 'space'-key on the machine itself, the carriage could now be moved freely - no binding or friction at all. When however pressing the carriage-release lever on the carriage itself, it still moved sluggishly - very stiff.

Some more peering into the mechanism revealed that the controls (a) and (n) act via a different path. The space-key (a) pulls back the ratchet, whereas the carriage-lever (n) pushes it back via a roller-wheel. It turned out that this roller wheel was also seized-up solid and sliding over the release-bar of the carriage.

Again applying oil and forcing the roller round made it unstuck - again rolling freely over the long bar that pushes the ratchet away. 

Now, with both rollers oiled, the carriage moves freely and the machine spaces correctly - even with its mangled spring.

Instead of mangling the spring, oiling these two rollers would've probably been a better remedy - something at least - for a next time a Mignon-carriage advances sluggishly.

Remington Portable adding machine on its feet again

The rubber feet of 1930s Remington Portable adding machines are nearly always 'gone'. That is - they will be present, but invariably are hardened and/or completely squashed. As in this wreck-machine below:

As a finishing step of reviving this machine, the gone-to-goo old feet were broken off the machine. A hole may need to be pried in the centre of a foot to get at the mounting screw. When removed, the foot-remnants should be completely broken apart to retrieve the washer. (This washer would originally have been moulded into the rubber foot.)

New feet that are a decent-enough fit were modeled in CAD. In this case, printed in regular 'plastic' and not in rubber. The conical washer is needed as insert - designing the foot with separate washer makes it possible to print on an FDM 3D-printer. It's not a difficult or critical part, any printer should be able to create this fine - STL-files available for download here.

The tapering top of the foot is needed to clear the various pins and screw-heads that poke through the base into the foot-pocket of the machine-base. The pocket allows mounting the feet with the original, salvaged screws and washer.

Original feet had a circular 'tread' - as shown below. This pattern would actually be fairly easy to create with 3D-printed feet in rubber, but in this case settled for basic, hard feet to be soled with felt (or alternatively, a pad cut from self-adhesive rubber sheet).

The printed feet here finished with a 'sole' from furniture-pad felt, great for sound-dampening. Adding some latex-glue to the felt made the feet grippy. Had grey felt, so blackened the sides.

The Dalton adding machine thus on its feet again and fully functional.

These 1930s Brennan-type Remington Rand adding machines are pretty common today on the usual auction and classified sites. They can only really add numbers (which they do very well and they print a list too). Perhaps not very exciting machines and indeed they are not very valuable today. 

Still, if you can find a decent-looking and complete specimen: they are brilliant late-1920s digital-technology and can basically last forever.

Oiling the level too

After topping-up the oil level in an adding machine, also added oil to the spirit level :)

At 2.50 a relatively expensive thrift-store find, but it looked intact and had a brass "JNij" medallion; so made by J. Nooitgedacht & Zn, IJlst. A well-known (at least here) maker of woodworking tools and skates, started in 19th century and now defunct.

The tool is made of oak with brass fittings. The protective covers over the levels are 'plexiglas', so probably a 1950s item.

By coincidence, also this exact item is pictured in the 1950s tool catalogue.

The description fits, including the optional extra 'feet'; brass end-plates. The description also confirmed that it originally was delivered 'oiled'.

When picked-up from the tools-bin at the store, it looked rather dirty, dry and tired. After wiping with soft-soap to remove much dirt, a few helpings of boiled linseed oil were applied with a rag. After a good rubbing, the wood regains 'depth' and is protected for another couple of decades use as a tool.

New typeshuttles for the Hammond typewriter - iterations

The current state of the (slow, intermittent) effort to be able to use the Hammond Multiplex with new typeshuttles; two new, improvised shuttles on the machine:

Last year, when ordering 3D printed parts for a reproduction Hammond spool, also ordered some first test prints for a typeshuttle. These were printed from a fairly simple 3D model, with straight 'slugs'.

This was mostly meant as a test for the dimensions. The Hammond typeshuttle has an awkward design without obvious angles or dimensions (at least, not obvious to me). Despite this first-version typeshuttle strip being very basic, it showed promise and with an improvised flange it actually provided (somewhat) legible typing!

This first attempt was printed on a resin-printer with high resolution. Very crisp, no visible layering or artefacts at all! The first-version shown below on the Multiplex; it's also obvious that the strip is too thick.

As it did indeed look promising, an improved 3D model was created with tapered 'slugs'.

To make the iterations in design a bit faster, this was simply printed on an FDM printer. With a filament printer in PLA, the printed part does clearly show layers. However, with a layer-height of 0.08 mm and great horizontal precision this is not too much of an issue. With FDM, the parts are easily printed and good enough to validate the design.

An example FDM printed reproduction typeshuttle shown below on the right, next to an original on the left.

Again, reproduction with improvised flange on the right, original on the left.

The thin flange with the take-up for the vane to actuate the typeshuttle is a tricky part - that especially will require more thinking and probably several iterations to get right. Even this FDM-printed typeshuttle does actually type (somewhat):

Even though the print-quality is terrible still, this Hammond Multiplex typed a bit in 'script'. The bad print quality is probably also caused by a badly-inked ribbon; another thing to try to remedy on this machine.

More tweaking to do - and to get a better ribbon (more ink). A slow and intermittent project, but fun to do and still shows promise. Exploring this also gave more insight in the actual workings of a Hammond typewriter and how some of the first shuttle-design aspects 'echo through' in the anvil & shuttle - fascinating machine :)

Checking the oil-level in an adding listing machine

An old Remington Portable adding listing machine (or Monarch, Torpedo or Dalton) may actually need its oil level checked. Like most mechanical lever-operated adding machines, the Portable has a dampener to control the return-stroke and prevent the machine's lever from 'snapping back'.

The dampener is the metal 'pot' visible at the back of the mechanism on the right-hand side.

The metal 'pot' contains oil, the damper-rod rotates a vane inside the pot against this oil (and a baffle) to regulate (dampen) the speed of the machine. After 80+ years, it may be that the oil-level inside the dampener is low - oil may have escaped (evaporated) via any small gap, e.g. the fit of the rod in the pot. When the vane is rotating in air instead of oil, it of course doesn't work as intended.

The dampener is fortunately designed to be topped-up. It has a filling-plug!

The screw on top of the top is the filling-plug - allowing a check on the oil-level (insert pin, dipstick) and for re-filling the dampener. This machine 'snapped back' with alarming speed, perhaps risking damage to the mechanism. The oil-level in the dampener was indeed low, perhaps half full. 

To re-fill the pot, new oil was inserted with a syringe.

After re-filling the dampener this Dalton (or Remington, Monarch, Torpedo) Portable adding machine no longer snaps the lever back when you let go on the return-stroke. It still returns quickly, but now controlled with an even speed to the rest-position - as designed.

Not an obvious thing to check and uncommon on e.g. typewriters; when an old mechanical adding machine's handle snaps back very fast and hits the end with a clear 'impact', it probably needs its oil-level checked :-)

C-key for the Remington Portable adding machine

The Remington Portable adding machine of the 1930s has a C-key protruding from the front. This C-key has a pointer to indicate the position of the next digit, pushing it to the right allows clearing an entered number from memory: marked C for Correction. These adding-listing machines can be branded Remington, Remington Rand, Monarch, Dalton, or Torpedo; and before the purchase by Remington of course Brennan. The C-key should all be identical, no matter what a particular machine is named.

(The C-key shown above is an original.)

The key itself is a press-fit onto a short stub of a stem. This stem slides into a pocket in the machine, attached to the machine's pin-bed memory. This key needs to be pulled-out to take off the machine's housing, it is only held in clamped by a spring. 

Being exposed on the front and also very easily pulled-out, these keys and/or stems are frequently missing from these adding machines.

A missing stem could be re-manufactured from a bit of mild-steel or brass of ~1.3 mm thick (probably 1/10th of an inch, but 1mm brass would also be fine). The tab that holds the key is bent 1 or 2 degrees up, tilting the top of the key slightly towards the operator. The V-shaped indents on the main ~11mm wide length that goes into the slot is where it is gripped by a spring. Below scan shows the overall shape of this stem with the main dimensions:

To replace a missing key itself, a CAD-model was made of this C-key's pointy-design:

With this 3D-model of the key and a good FDM-printer, this little pointy-key can be 3D-printed without any supports:

Printing with a small layer-height (0.1mm or less), the amount of visible layering is not too great. One or two layers of hobby-paint will mask any layering and will create a smooth, glossy key. The lettering then  picked out with white (or cream) latex paint.

(The C-key show above is a reproduction.)

Now the entry can be cleared again also on this 1930s Dalton-branded Remington Portable adding machine.