Typewriter platen re-coating

August 17, 2025 by Lukasz Kumanowski

I have one machine with platen re-coated by JJ Short. It is a benchmark quality and it is the most appropriate for such ancient machine.
The issue I have with doing it for all my machines is that, well: I have too many candidates and I simply cannot afford sending all these platens for professional service.
… and I am a bit afraid of the package getting lost.

Besides - where is the challenge in this?

I was looking for an alternative, some other process which I could do in home workshop without breaking the bank.

Cork

Inspired by Oliver with cork platen I had a brief adventure with making these. Started with Blick by using synthetic cork. It prove to be too soft.

I then imported natural cork stripes, bonded together with some modern glue. Unfortunately that too prove to be too soft.
It was certainly not cork itself but the glue holding cork bits together.

My conclusion was that it is a dead end. Modern PU-glues used for making cork sheets are simply too soft.
Trying to make them harder by soaking-in superglue or impregnating with varnish was a desperate and failed try.

Heat shrink tubing

This DIY method is widely used.
It kind of worked for me for long time although I was not very happy with surface finish. Maybe it was the material I used?

It is essential thou that one uses a lathe in the process, even if it feels unnecessary. Heat shrink tubing very rarely shrinks uniformly, throwing concentricity of the platen through the window.

Grinding this on a lathe gives such ineven surface finish. It can be, as stated, material with such defects. It can be that grinding I did was too fast and caused heating up the material and “bubbling”.

I saw no future in this to spend more time in fine tuning the process anyway.
There was another idea, much more promising, lurking on the horizon!

Vinyl

While searching for rubber hoses of adequate diameter and hardness I stumbled upon vinyl hoses. Transparent ones.
That reminded me that there is already someone who refurbishes platens using PVC, he called his product “turboplaten”. I found this seller on eBay: Writertypes
There are generally good reviews of his service, with some concerns regarding affecting escapement timing. Interesting!

Keeping that in mind I started experimenting with developing my process.

The first “patient” to treat was Corona foldable with broken platen.

Primum non nocere, I thought, it can only be better, even with experimental platen.

Using standard lathe tools I managed to turn this down to required diameter. Fine-grinding to obtain the final diameter.

Resulting platen became milky-whitish.
Not bad but could be better. And grinding process was a pain in the neck.

Adam was brave enough to send me his platen, for Erika 5, to recoat in this process.
I managed to minimize scratches on the vinyl and decrease “milkyness” of the result but still there was a potential for improvements.

As a comparison to heat-shrink-tube platen the surface is much nicer, even if still milky.

Only after I made my own tool for turning soft vinyl I could finally achieve the result I was after: transparent platen which can be customized with color/texture/text!

First “patient” for this platen was black&red Royal. It had broken platen when I got it, repaired with heat-shrink version but I was still not happy with the imprint.

Platen core is wooden on many old machines.

I used a newspaper from the day when the World markets went down, showing red charts everywhere. Matching nicely red body panels - at least one positive thing in this whole situation.

It prove to be quite sensitive material to turn with my tool. Some lessons were learned at this stage.

Another trial and finally the new platen is ready!

Test typing.

The imprint is better but most importantly the machine is more quiet. And I like the look much better than with heat-shrink.
Paper handling is superb, too - vinyl is grabbing the sheet like a bulldog.

After a few days of using this new platen I got confident that this is the way to go. Especially for mass-produced, non-unique machines like most in my collection.

Next specimen on the bench was early Halda which I bought while traveling to Norway. She got comic-book platen, with Swedish version of “The Phantom” from SvD.

A typewriter friend sent me his platen for early Facit P1.
Here we went wild and chose bright green color, to nicely match green keys on his Facit.

Measured hardness of this material is around 85 Shore A. Which is on the lower part of appropriate hardness spectrum.
I found another supplier of vinyl, this time a bit harder. We have not yet measured its hardness but I estimate it to be 90, circa. Which is optimal for me.
I already started making platens for my other machines using this new material. It is bluish and not as transparent as the first one so I use it mostly for single-color (or black) platens.

Here are some recent examples.
Facit P1:

Olympia SM4:

Print quality on this Olympia was good even with old, hard platen. With a new one it still is very good but also the machine types more quiet.
Not a ton but a tone lower ;-)

Results & discussion

So far, with my limited experience, I have not encountered any problems whatsoever with these platens.
They are not more “bouncy” than standard rubber, I would say they behave same as my proper rubber platens from J J Short.

Someone mentioned a risk that “bouncy” platen can affect escapement timing.
I am yet to experience that.
All the machines I did live perfectly happy life afterwards. But maybe my material is not more bouncy?

What I like about these vinyl-platens is that I can make them myself at rather low cost. Sure, one needs a lathe to make it but, honestly: how can one be happy without a metal lathe??? ;-)

To be able to customize color or background is a plus for me, too.
With two types of vinyl I use now I reserve clear one for older machines which would match warm look better. Blue vinyl is more suited to more modern machines and to solid colors. Like standard black or blue. At least how I feel right now.

Example of different “feel” with both materials shows sample below.
It will be a “musical” platen for old Erika 5.
I chose Bach notes, inverted to negative and with sepia tone for more warmth.

We are about to test Facit platen with bright green core, soon, so no hard rules here either.

Last but not least - this is perfectly reversible process.
The platen material is a consumable, once it gets damaged/hardens/owner-does-not-like-it it is easy to replace without harming the machine.

Poshed up Olympia SM carrying case

August 02, 2025 by Lukasz Kumanowski

I saw people remove paint from these and expose beautiful plywood from underneath rather boring, hammered-enamel gray paint.

The paint on my box was already partially scrapped, exposing promising beech-wood plywood.

I decided to remove all the outside paint. One on the inside is still in pristine shape - still nicely pretending to be metal.
Hot-air gun and scraper make the job very easy and assures that only paint is removed, not plywood.

Hand-sanded and vacuumed it looks much nicer already.

I especially like these streamlined shapes, as on sports car from 70´

I decided to keep the wood unstained. I like this case to be in lighter color, in contrast to all others which tend to be dark or totally black.

Woodworkers favorite moment: oiling! Brings out wood grain beautifully.

All hardware screwed back in place.

Handsome piece of office equipment!

Torpedo 12. The looker!

May 10, 2025 by Lukasz Kumanowski

At first I thought it is a ‘14.
Torpedo 14, that is.
Closer inspection (serial number) and some research - mostly Robert Messenger’s excellent web page here and here - showed that what stands on my bench is Torpedo’s first portable typewriter: Model 12.
Serial number is on right side of the carriage. Manufacturing date for my specimen is around 1928. Soon she will become 100 years old!

Now: the machine is not mine. She belongs to Sara.
I got an opportunity to work on restoring the typewriter to working condition and she will go back to Sara.

Nothing worked initially. Neither space-bar nor key-action moved the carriage - but the draw-string was attached.
Escapement was somehow not clicking. Well, I started disassembling.

Closer look and I found the reason: one small spring was missing.
I found spare spring which should match required elasticity, installed it and the machine woke up to life!

As always in so old machines - paper feed rollers were busted: flat spots, cracked, rubber petrified and brittle.
To remove them one needs to hammer out the axles. They sit with friction fit.

New rollers made with silicone tubing and heat-shrink:

Back in business - in situ:

Draw-string attachment was not looking kosher to me.

I replaced it with appropriate end-piece.
(This tiny nut comes from an old mechanical clock I found in the woods some years ago. I kept the mechanism and whereabouts, finally they become handy!)

This machine is a very early construction.
It does not have automatic ribbon reverse system - one needs to reverse it manually by pulling - or pushing - small levers on each side of the machine. On this one they were missing. I manufactured copies, based on pictures from Internet.

Threads M2.5, or close enough to match. Installed:

Rubber feet with threaded inserts - this machine is screwed to base plate, part of the carrying case. I managed to find matching rubber inserts.

Platen is hard, of course, but not cracked or excessively used. Interesting that axle rod is permanently mounted, riveted on both ends. No ratchet clutch on this machine.

To install this assembly one pushes left side of the platen (with platen rod) into central hole on the carriage. Line advance claw needs to be lifted up to clear ratchet wheel. Then the right side of the platen is dropped into slot on right side of the carriage.
Platen is held in place by a latch, screwed in place afterwards.

Screw head holding this latch was butchered with improper screwdriver. I hate it, especially on a machine like this one, with hardly any other screw damaged.
I made a new screw from an old stock I have, same pre-war era.

In situ:

Same for platen knob screw - the one sitting there was butchered. I made a replacement.

This machine has bi-chrome selector so she should be able to use duo-chrome ribbon. It was broken thou and locked in black position. That needed some attention then - I made a new lever from steel plate.

Installed:

Once having this back in operation I spooled duo-chrome ribbon and started playing with ribbon color selector.
To my horror - it did not work reliably. On RED position, when ribbon vibrator goes all the way up, it often stayed in this position and ribbon was hanging on type guide nearby.
I tried increasing tension on vibrator´s spring by replacing it with a stronger one but that did not help much.

I tend to think now that the selector lever was intentionally broken and left in the only working position: for BLACK. So that no one messes with it.
Well, I’ve already made a new one so it will stay but I don’t know how to repair the problem with hanging vibrator at RED position.
For now, I think, we will just accept the fact that she is typing only in one color. Let’s call it age & patina, shall we?
I spooled fresh, black ribbon and did some typing.

Apparently I still believed it is Torpedo 14 while it is, in fact, Torpedo 12!

New feed rollers take paper like a bulldog!
Not bad, no?

Apart from occasionally hanging ribbon vibrator I discovered one more problem.
This specimen was unlucky to have someone who very enthusiastically treated line advance mechanism.
It was not working reliably - sometimes advancing 1 line, sometimes advancing 1,5 line. Turns out it was bent.
I bent it back to shape - or rather “formed it” as used in the industry. Works like a charm again.

The weather was fine so we went out for a concluding photo session. She is quite a looker!

One can clearly see Remington connections, type arm mechanism is the same:

She reminds me other Remington portables, like Remette or Model 5 (streamliner).

Honest patina on space bar. Decals in very good condition!

Most controls are on left side of the carriage.

Line-advance is a pinch-lever (black arrows).
Two line-spacing settings: single and 1.5 line space (red arrow).
Ratchet release (but no clutch - yellow arrow).

There is also rudimentary carriage lock: small lever (yellow arrow below) with a tooth, going into carriage rail to lock it (red arrow).
Works fine as long as you don’t press shift accidentally.

Very handsome machine. Typing action is similar to Remingtons of that era.

I very much prefer to type when the machine is detached from its carrying case. This Torpedo has its own rubber feet so this is possible.
Lap-typing, thou, was not yet discovered back then - the whole mechanism is unprotected from below. Setting the machine on your knees means trouble!

So there we have it - another survivor!

Stoewer KsM. Spooky Scary Skeleton-shift machine

May 03, 2025 by Lukasz Kumanowski

Stoewer Kleinschreibmaschine (KsM for short). I worked once on its younger sister, Rheinmetall KsT although it was not my machine. Until now I never owned one of them so when I saw Stoewer for sale I grabbed it.

Adam claims that these are best of the best of the best (with honors) portable typewriters. I had to try myself.
First thing I noticed is that this is also a skeleton-shift machine. Same as Rheinmetal KsT.
It also has this cool feature: carriage removal is very easy, just flip two latches on the back and voila!

My specimen was mostly very dirty and lacked some screws.
Disassembled for cleaning.

Paper-feed rollers were in good shape, just hardened and dirty.

I removed 0.2mm of oxidized rubber to expose still elastic rubber underneath. No need to fabricate new, yet.

Everything is very robust on this machine. After removing the platen (standard procedure, central rod can be removed from either side) nothing falls into pieces and all stay in place, firmly secured.

Skeleton-shift has a main advantage of lower mass being lifted upon shifting but also: shift adjustments (fine-tuning) is right there, very accessible.

Apart from dust and some dirt my machine had some surface corrosion on chrome-plated controls. Since there is no trace of that anywhere else inside the mechanism I recon it was bad plating and/or previous users were sweating very corrosively…
I’d rather have no plating than corroded one so I polished away damaged plating up to clear steel. Satin finish is my preference in that case.

A few of glass keys have custom, hand-made legends. I could, of course, change that to one matching the rest of the keyboard but I decided to leave them in place.
They tell a story, although I don’t know what is that story.

With carriage removed one can study closely parts of exposed mechanism. End-of-line lock and the bell are a robust and refined assembly.

Ribbon advance is directly connected with spring motor advance.

All threads are metric on this machine so I could easily replace missing screws using my stock of old screws from the same era.
Machine put back together and waxed with Fulgentine.

Test typing with (almost) fresh ribbon. All seems to be good!

Typing action is very pleasant, the mechanism is tuned and very snappy.

Traditional photo session followed.

Serial number is clearly visible on the back. According to this source my machine was produced sometime between 1926 - 1930 in German Stettin (today it is Szczecin, in Poland).

Bar-Let. Tilt-shift, no draw-band gem from Nottingham.

April 27, 2025 by Lukasz Kumanowski

If I recall correctly it was Adam who sent it to me. The poor, little fellow (typewriter, that is) was reacting only to space-bar but carriage was not advancing when keys were pressed.
Turns out it was this spring which was missing.

After adding it the machine woke up to life, although reluctantly due to thick layer of debris everywhere.

Thin and lightweight as she seems - it still has an outer shell which can easily be removed. After unscrewing a few short screws the machine can be wiggled upwards and out.

Segment and a few other parts are in cast iron, all the rest is pressed plate. Hence light weight and, probably, lower production costs.
My specimen was too dirty to consider simple cleaning. I decided to pull off the segment and type bars for deep cleaning.

Order in which type-arms are installed. Makes life easier to document each step.

When removing the platen one is wise doing many pictures of both sides of the carriage. After the central rod is removed all falls into pieces. Here is my reference gallery, it prove useful when building up the machine afterwards.

Nice thing about carriage on Bar-Let is that it tilts upwards upon FIG and CAP and there is only gravity which keeps it from tilting all the way. One can lift it up 90 degrees and easily gain access to underside of the works.

Each type-arm link seems to be same as others but they are not. Each belongs to one and only one arm, they are custom-formed into these “kind-of” ornamental forms.
I found out the hard way that keeping links assigned to their arms is a wise thing if one does not fancy metal-puzzle games.

Paper-feed rollers were busted, as expected. They sit on brass wire, pinched in place, locking it. One must cut it to remove the roller.

I made new rollers from silicone tubing and heat-shrink sleeves.

One last thing was to adjust escapement dog spring. The original was becoming too tired.
When the carriage was at the beginning of line (spring at highest tension) it was too much friction on the star-wheel to pull back the loose dog.
I’ve adjusted the spring by making it 2 coils shorter.
That is, by the way, a cool mechanism on Bar-Let: there is no draw band. Instead the star-wheel is itself spring-loaded and it is this which drives the carriage. Works fine on such lightweight construction, I suppose.

And so the cleaning and basic adjustments were done.

I assembled the whole mechanism, spooled a fresh ribbon and tried typing. Type alignment was a disaster, many arms stopping in type-guide due to misalignment.
It took a good deal of time to make them work at least but there is still a good bit to go if one wants to achieve proper alignment.

Not sure I’m gonna try further with improving the alignment. It is not a machine I would like to write longer texts.
I took her home for a concluding photo session.

Same as Robert Messenger I am also under the spell of this slightly curved keyboard and ribbon spools turrets.

Simple controls on this machine.
No bi-chrome selector, instead manual reverse and no-ribbon-advance (for spooling) as middle position.

Dedicated, standard width ribbon spools.

Line-advance as pinch-it lever. Single and 1.5 line advance. Ratchet release (non-locking). That’s it for the left side controls.

Right side houses carriage release (pinch-it, again) and single platen knob.

Serial number 10360 places its production around 1933.

Barr Universal. Slightly different animal

April 19, 2025 by Lukasz Kumanowski

Two of these landed on my bench: one in Pica, one in Elite. Both had several problems, apart from dust and dirt.
Common problem for both were smashed paper-feed rollers.

It usually is no issue to remove these but not on Barr.
Here, for some reason, they decided to thread both ends of each axis, leaving no place where one can grab it and try unscrewing. Entire length of all axis is covered by rollers or spacers (also rolling).
Moreover - even if you manage to get hold of the axis - you cannot unscrew them. Thread is too short to unscrew on one side and it gets screwed into the other side while you try.
In short: they did not plan to replace these, it seems. To remove them one needs to disassemble almost the entire carriage.

Well, after struggling to find a way to do it properly I finally gave up. No way I’m gonna remove rivets etc to dismount them from the carriage.
I decided to cut them out and make new, modified for future service.

Measurements of existing parts. In case of rollers I needed to guess their diameter - they were so deformed.

Cleaned old rubber on a lathe. It turns out that big rollers were mounted on brass pipes while small ones on steel.

I leave spacer rollers with their patina (brass). Freshly cleaned rollers will be covered with new rubber so shiny metal will not be seen.

Turning two sets of axles, for both machines. I use steel rods from old IKEA bookshelf.

It turns out that metric thread M2.5 is perfectly matching UNC thread which is originally on the machine (I suppose). I threaded my rods so one side of machine supports can be used with their threads. The other side got its thread removed and will accept new axles as THRU. They will be secured with nuts on both sides anyway.

New rollers were made from silicone tubing with heat-shrink cover, for period-correct color and to fine-tune required diameter. Mounted on the machine.

Next problem: snapped draw band.
Spring motor on Barr is located on left side, under the cover. I made a new string from strong twine, re-used original fastenings used with cotton band.

To add tension to the motor one does not need to wind it by hand while holding the drawband. No, on Barr this is civilized: firstly remove the locking screw.

Then just wind the spring until it is enough to draw a carriage all the way. Service manual gives exact value for tension: min 2lbs when carriage on position 40. I eyeballed it on my machine.

Next problem: carriage binding when trying to move with carriage release controls. For that I needed to remove the carriage, suspecting problem down there. It is easy on Barr - just two screws, one on each side. After detaching draw string the whole assembly is simply lifted straight up.

It turned out to be slightly bent bracket which is being pressed by carriage rail gliding above, when carriage release is used. Formed it back into shape.

One of my machines was having problems with typing in red. It was either totally crazy and unreliable or causing the ribbon vibrator to jam.

Jamming was especially dangerous when in shift - since Barr is a basket-shift machine it was causing the whole basket to jam below derailed ribbon vibrator.

To adjust it one needs to find a sweet spot in two places, adjusting each one of them separately and VERY SLIGHTLY at a time.
It costed me many hours and lots of cursing to find the spot for my machine. It can be done but requires lots of patience.
First we need to adjust position of the plate below, by easing the pointed screw. We aim at having ribbon vibrator erecting links to form a straight line when the key is pressed. I cannot show it on the picture but service manual describes it.

Once this is done we move to another spot, under the machine. By adjusting these two nuts we increase or decrease the throw for ribbon vibrator.
A word of caution: it requires very small adjustments, like 1/2 turn of a nut, to reach the proper position.

Well, with that fixed all the rest was easy.
Both machines had busted rubber feet. Luckily our local typewriter-supply shop could provide me with exactly matching rubber feet replacement. Originals, hard as plastic, look like that:

Modern, state-of-the-art replacement from Biltema looks like that:

They do the job perfectly and fit in carrying case alright. What not to like?

Ribbon spools on Barr are slightly narrower than modern Pelikan, plastic spools. I managed to find steel spools fitting perfectly.
Fresh, nylon ribbon from stock roll:

… and both machines are ready to fly!

Barr-typewriter is one of these underrated and looked-down machines which is utterly unfair.
They are different from all other but typing action, at least on both mine machines, is perfect. I would compare it to old Remingtons or Royal Portables: snappy, responsive, very satisfying.

Most of controls are housed on the carriage. Barr has additional handle on the right side, used if one wants to just move the carriage but not advance the line. Line advance lever is on the left, like on most other machines.

Feed-rollers pressure release lever has additional, pinch-action lock. It hooks on the back of the carriage and keeps rollers in disengaged state. No other machine has such feature, as far as I know.

Line advance selector is an arc of holes:

Speaking of line advance: I admire shape of the line advance arm! Reminds me sophisticated lines of Groma N!

Ribbon selector is located in usual place, on the right, above keyboard. There one can chose between top and bottom position. To chose Stencil mode we need to draw a separate lever on the right, bottom side of the machine:

Small lever marked with yellow is ribbon direction selector. Machine will reverse ribbon direction automatically as well, of course.

Carriage lock is implemented by means of pressing down line advance axis. It looks solid and bulletproof.

Margin and Tabulator rack is hanged on hinges, spring-loaded. One can lift it up, to see better while setting controls. Rack´s digits are etched up-side-down for that purpose.

Will Davis has an interesting article about this type of TAB solution.

Keyboard is more or less standard 4-bank layout. What stands out is that Shift.lock is on the right side only (but released by both Shift keys) and that Margin Release button is vertical.
In my writing it works equally well.

All in all - very sexy machine.
Typing feel is perfect, controls are straightforward even if a bit different than what we are used to.
Robust, no-nonsense machine for heavy writing. I totally agree with Robert Messenger on that.

Sharp PA-1050. New ribbon for thermal typewriters

February 22, 2025 by Lukasz Kumanowski

I guess I am a bit late to this game of thermal typewriters. When I researched my machine I found articles and typecasts of Ted Munk and Joe Van Cleave from a decade ago.

Well, better late than never.
The reason I picked this machine was my need for a very silent typer. And one which will not cause me troubles when flying. Call it my bad luck but on my recent flights to Germany security control was a PITA, both in Sweden and in Germany.
Imagining that I need to explain myself to carry a contraption like Erika 5 on a plane just gives me a rush.
Sharp PA-1050, on the other hand, looks like an old laptop and - being mostly plastic, with standard D-cell batteries - should not cause any alarm-bells ringing.

Called baby-wedge by Ted Munk, you can see why from a side:

Mine came with manual (in Swedish), a few cassettes and an additional RAM-module which extends its internal memory and enables storing several A4-pages inside, to be printed later.

I was saving these cassettes for a later, typing on thermal paper instead (no cassette needed then). My favorite thermal paper became this Brother one:

It gives very nice, dark print and I cannot find anything to complain about but… I want to type on my all-time favorite paper, Sweden-Bond!
It is slightly off-white and has a delicate texture with a water mark. On picture below it is an ark on top, laying on Brother thermal paper ark (extra-white).

Well, thermal transfer ribbons are still made today and widely used. We use Brother printers for labels etc, thermal ribbon comes in plenty of sizes and lengths. The challenge is that my machine uses 6mm wide ribbon and the narrowest one sold today is 9mm and it is just 10m long.
The cassette can be opened easily. I checked how things are routed inside and how wide the new spool should be.

I could not help reading the old ribbon left in the cassette. Seems that the last time this machine was used was 1995:

The choice of bulk ribbon is very wide. I bought one of these and started making a jig to cut it. The idea was simple: a row of blades through which the tape will be drawn, resulting in simultaneous cut of many 6mm pieces.

It kind of worked for a while but inevitably was derailing after a while.
The tape is too thin and too fragile for that. Besides - so many threads become unmanageable.

I dumped the idea and tried something more radical: cutting on a lathe!

To keep the ribbon from accidental unwinding I taped it with masking (paper) tape.

It is mounted on a lathe so that when it turns and meets the knife it is “winding itself”.

I made the knife from rotary cutter blade, mounted on a simple shaft.
The trick is to use as thin blade as possible - at the same time as stiff as possible so that it will not diverge to side during the cut.

I could not avoid some problems:

first cut became too narrow because the blade diverged from an initial line (yellow arrow)
second cut was too abrupt: knife movement vs rotation speed (blue arrow)
the third cut yielded acceptable results but not perfect yet.

Unmatched rotation vs knife movement caused the ribbon to heat and loose parts of the resin, visible as reflective stripe above. It is a Mylar-base without the resin which is showing through.

Once I got these parameters dialed-in I continued cutting with no further problems.
Next step was to wind the ribbon onto cassette spool.

It is a bit fiddly to mount this back into the cassette. Routing this tape through all nooks and crannies is taking time but can be done with patience.

Finally time for test-type!

Well, it works!
So there is hope for these cool “baby-wedges” out there!

The procedure is quite straightforward and after going through all possible mistakes it becomes a routine. As always with everything.

At least the material is not expensive: I bought a bulk ribbon for 25 USD, it can yield around 30 such cassettes. Maybe more.

So if something goes wrong - stop crying and just move on.

Erika 2 foldable - platen removal, ribbon position adjustment

February 09, 2025 by Lukasz Kumanowski

Always a joy to work on one of these little gems.

There was not much wrong with this specimen, mostly very dirty and with cracked platen. I don’t mention petrified paper-feed roller since this is always the case on machines as old as 114 years.

Removing the platen is easy: one plunger screw is holding platen knob.

Now, platen axis-rod has a key which needs to be aligned with a slot to come out.
So: no hammering if things don’t move!

Platen.
Removed the old, hardened and cracked rubber. Wooden core - beech wood.

I recovered the platen with heat-shrink tubing, back to original 32mm diameter.
Turning on a lathe is a must for such operation, heat-shrink tubes shrink too non-uniformly to rely on that to be concentric.

Line advance is a pinch-action on this machine. The ratchet did not come to home position, the spring was too tired. I changed to new, stronger spring and it works perfectly again.

It is a small and very light machine so good rubber feet are essential for comfortable typing. I made mine from Biltema components.

Key-tops were so dirty that I thought that they are damaged. Luckily these are made of real glass so it was possible to clean. I use Fulgentine - carefully to not flood the key. Just enough to soften the dirt and scrape from the glass with toothpick.

Lastly - new ribbon.
Nominally these machines use 16mm (5/8 in) ribbon. I was buying mine from Holland but right now I had none left.

Well, there is a way to use standard 12.5mm ribbon on these, too. As long as it is single-colour ribbon.
One just needs to adjust how high the ribbon vibrator rises upon key strike.

Shifting this latch left or right (red arrows) is causing height change for the vibrator (green arrow). Tighten the screw once done so it stays put in this position.

And so she types again, after many decennials of idling in the attic.

Very pleasant typing action!

I took her out for a photo session (the plank on which she is sitting is just to protect her from snow on the table).

Playing with filters for a bit more “posh” effect.

Erika 8 disassembly - body panels removal

December 01, 2024 by Lukasz Kumanowski

Every now and then I need to work on Erika 8 or Erika 10 and I never remember the procedure to remove the body from the mechanism.

Searching the typosphere brought no relief so I decided to describe my way - so that I don’t need to search for it again and possibly some other desperate soul can find it helpful.

Of course, one can just remove the carriage and then all becomes easy but for times when carriage removal is not necessary - here is what I do.

Grab type-arms together, middle ones are in a way when wiggling with the body.

Unscrew four bolts on the underside (Erika 8)

Unscrew margin rack

Set Shift-lock, to lift the carriage. Move the carriage all the way to the right

It is worthwhile to put some electrical or masking tape on the back of the panel - this part which will be wiggled around the carriage. It is tight fit but with some practice one can remove the body without any scratches.
Lift the body from space-bar up and then left side: up. Mind the back side, to clear the carriage rear extensions

Voila!
Assembly in the opposite order.

This works the same on Erika 10 and probably on other Erikas with unibody, like Erika 9 or Erika 12.

Survivor! Oliver 9 aka "Rusty"

July 04, 2024 by Lukasz Kumanowski

I bought this fellow in UK when I was working on my other Olivers with intention to use it as spare parts reservoir.
It was in so neglected shape that no one else was bidding and shipping to Sweden was cheaper than shipping same thing within Sweden…
I got this rust bucket and never needed to shift any part from it. It was waiting in the workshop for its destiny until last week. At first I thought about scraping it.

Olivers are known to have poor quality paint- and chrome work and this machine was apparently left somewhere humid and dirty for a long time.
Apart from surface rust and dirt the mechanism was complete. It was not working but there was a potential.
I decided that it’s never too late to give up so I might as well try to save it.
Removed spider webs and dirt, blew with compressed air - to prepare it for Evaporust bath. Keys removed as I’m not sure how they will react to chemistry. All the rest is steel.

I doped the machine in Evaporust and left it there for a few days, only changing position to expose most of the mechanism.

Most crude rust was dissolved, next step is cleaning with high-pressure water.

Water blown out with compressed air and directly afterwards the machine is being dried with hot air.

At this point I decide to remove rests of paint on iron casting. It is anyway in poor shape, sloshed with some white paint here and there.
To gain access I remove carriage rail and escapement assembly. It is easy on Oliver, it’s just a few screws and levers to disconnect.
I scrape paint with a blade - it sits so poor that almost no pressure needs to be applied to chip it.

It starts to look interesting!
Carriage rail assembly is very rusty, ribbon color selector is seized with rust.

I dump it into Evaporust and leave for 24 hours. Cleaning with water and drying in the sun.

In the meanwhile I dope keys in Ajax and let them soak it for a few minutes. One needs to be careful not to leave them too log in this chemistry, especially black keys - they become soft and porous after a few hours.

Lots of “elbow grease” with wire brush and sanding pads. Finally all is clean so I put the machine together. I decided to leave towers and cover plate with original paint as it was in good shape and decals are well preserved.

Type-arms shield was bent on the right tower which also caused bend on a few type arms. I straighten it roughly and leave fine-tuning for later.

Time to take care of the carriage.

Disassembled as much as needed to de-rust it. I leave it in chemistry for 30 hours.

Cleaning with lots of water - most rust is dissolved into green sludge, water-solvable. Drying in the oven in 100 degrees C (there are plastic knobs on some parts).

Wire-brushed and clean, ready for assembly. Paper tables left with original paint - in reasonable (for Oliver) shape.

I’m always unsure if I will be able to put this together again. This time I cheated and used another carriage as a guide. All back in place and mounted on the machine. All rollers are clean and lubricated - carriage moves smoothly on clean rails.

Looks good.
Next step is manufacturing draw band clip. I use original one as a template and make mine from brass. Much easier to work on and strong enough for the job.

In situ.

Main spring is (surprisingly) not broken so I don’t open it, just lubricate through provided hole. After some exercise it works smooth.
The machine is assembled!

Ribbon color selector is back in operation. After carriage rail assembly removal I needed to adjust its span so that letters are not printed in duo-chrome. It is done in two places: rough adjustment under the machine:

Fine adjustment accessible after removing front cover:

And it types! What is left is adjusting every type bow to bring all signs into line - it is very often that these are bent, mine has most problems on the right tower where protection plate was bent. It takes some time and patience but is easy.

I take her out into the garden and use sunset for photo-session.

I 3D printed ribbon covers, they turned out a bit too small. I’ll need to adjust the model and print new ones.

Lots of “patina” even after removing most of the paint. I think it turned out pretty well. Seems to be worth all these hours of work!

I kept original feet, for a time. I’ll probably make new in fresh rubber or cork - this machine is so heavy that modern cork-composite works equally well as rubber and is easier to form.

One more survivor from the long gone mechanical era!