Reproducing tiny plastic gears for vintage 'gear'

[Jeff_Birt]

I love the 4-pen Alps printer/plotter mechanisms that were used by dozens of manufacturers back in the day. Repairing then and keeping them working is a struggle. The #1 failure is the pinion gears on the stepper motors. Being a simple spur gear, it was possible to commission getting a large number made at a reasonable price. These brass gears will last a lifetime.

The second most common failure is the Y-Axis (paper) anti-backlash gear set. This is a set of two interlocking gears that are spring loaded to reduce backlash on the Y-Axis. The outer gear is pressed onto a steel shaft and cracks (like the plastic pinions do.) Being quite complex machining replacements if not practical commissioning a mold is far too expensive.

Thankfully now we have relatively inexpensive 3D printers that use a UV curable resin. I picked up a low cost one recently, on sale, and have spent the last two weekends trying to reproduce this gearset. I finally got what looks to be a workable set after about 10 tries. It took a while to get familiar with printing with resin as it is unlike FDM (filament) printing.

Attached are a few pictures with a ruler for scale. The resin used is no ideal for this application but does produce very good detail and is inexpensive enough to experimentation. I also have some 'ABS like' resin which I'll try next if the design proves feasible. It is tougher and will make a better long-term replacement.

I'm excited about being able to reproduce small mechanical parts for out old calculator and pocket computers. This opens up new possibilities.

   

   

   

[AndiGer]

Amazing you can 3d print so small parts in very exact form nowadays. I didn't expect. Cool.

[Jeff_Birt]

The resolution is amazing for a cheap consumer grade device. The newly printed gear was tight & I broke it while trying to remove it. I did get it pressed on far enough to see it does mesh with driving gear. The paper feed roller shaft is stepped 1.88mm to 1.98mm making drilling to size difficult. The bore was increased 0.05mm & am printing 2 more for testing. Also added a chamfer to the neck of the hub where it broke to increase strength.

   

[John Ioannidis]

'home' 3d printers (meaning, any printer that costs less than a home ) may be accurate enough to produce those tiny gears, but I don't think the material is strong enough to take the stresses. There are other options, however; places such as emachineshop.com will machine real involute gears for you out of the material of your choice. But before you do that... have you checked whether McMaster-Carr or MSCDirect carry the gears in the sizes you need? They are expensive, but definitely less expensive than having them machined.

[Jeff_Birt]

Indeed, the range of affordable 'engineering grade' resins is rather limited. Years ago, FDM printers where in a similar spot with the available plastics suitable for printing which also had desirable mechanical properties when printed. Filament has come a long was and likewise resins are improving too.

As I mentioned in my original post the pinion gears for the stepper motors I had machined out of brass. They are a simple spur gear and making them from brass was an economical option for a reasonable quantity (a thousand gears).

The Y-axis anti-backlash gears being quite complex machining replacements is not practical and commissioning a mold is far too expensive. Thus, 3D printing is a viable option. A gearset that lasts a few years with hobby use is a better option than a non-working printer.

[rprosperi]

(07-17-2023 07:53 PM)Jeff_Birt Wrote:  A gearset that lasts a few years with hobby use is a better option than a non-working printer.

This seems to be the point most naysayers either don't get or perhaps a fact that they choose to conveniently ignore. You can literally print a new one every month for less than the cost of a 'properly-made' part. And although that could mean that you also have to do extra work to replace it once a year (or decade) this seems to not be an issue for most folks with the skills to make the parts.

[Jeff_Birt]

Here is a quick 2 minute video of the 3d printed Y-axis (paper feed) gear set in operation. There is still some tweaking to do but it works! In the video I mentioned that the gear fitment was too tight to allow offsetting the two gear halves. It turns out I was trying to rotate the inner gear the wrong direction. The image shows the gears are now offset by 1 tooth.

This resin is quite brittle. I was able to remove a small chip from a tooth with my fingernail. I have an 'abs like' resin while I will try next to see how it compares.





   

[ctrclckws]

Not remotely calculators, but as an example of 3D printing to replace old parts.

https://www.popularmechanics.com/militar...ing-parts/

I suspect that many collectible items will have printed parts in the future.
The last C-5 Galaxy is of a similar vintage to many of our favorite calculators.

[Jeff_Birt]

Must have been one of those fabled 30,000 toilet seats

One of the challenges with 3d printed parts is how to properly inspect them. Several years ago a colleague and I got to wondering about mm wave scanning of FDM prints. MM/microwave inspection is his field of expertise, and my job is more of building apparatus for various research labs.

I printed some samples with 100% infill, a.k.a. solid, except for intentional spherical voids of various sizes. I have attached a few slides from the presentation. Voids as small as 0.5mm were detected. The diagonal stripes were an artifact of the sling process. When it came to a void it would not 'jump over it', rather it would come back at the end and fill in all the missing sections behind the void. The strips are due to the difference in temperature of the previous layer over time, i.e. by the time it came back to fill in the sections skipped the previous layer had cooled slightly. Being able to see this was a surprise.

The ME department has also done research into how to test parts 3D printed out of stainless steel for military applications. Sort of like the toilet seat example some replacement parts are only needed in small quantities and some aerospace parts can be very expensive to make.


   

   

[jjohnson873]

Very nice, Jeff! I've printed some absolutely beautiful gears with my 3D printer. However PLA isn't flexible and durable like a nylon OEM gear. This printed gear was for a shredder. Needless to say, the PLA gear was shredded after I installed it. I used Taulman3D Alloy 910 filament, which was supposed to be a more rugged filament, but it wasn't up to the task of replacing a nylon gear for a mid-range shredder. I'm interested in your findings for cured filament and how it compares to nylon or other materials. Replacing old gears in small printers is a great endeavor!
~ Jim J. ~

[Martin Hepperle]

BTW: commercial 3D printing services often offer Nylon as a material to laser-sinter parts. These are made from a fine Nylon powder which is melted layer-by-layer by a laser beam.

This process produces very rugged parts, but these have a slightly rough surface finish (like maybe 200 grit carborundum paper). If you need heavy duty parts, this may be a solution, even if it a bit more expensive. But not for very small gears.

Martin

[Jeff_Birt]

After testing multiple resins, I found that the one I had originally tried was by far the best. This is rather ironic as if I had not started with it, I may have declared the project impossible.

I am looking for a few beta testers, people who are in need of this gearset for their Alps mech. Feedback is vital. I am in the USA so other folks in the USA are preferred as it makes shipping less expensive. I have run the gears for a few hours of printing time, so I think they will be fairly robust but I'm sure they will not last for decades.

[Jeff_Birt]

These are pictures of the ALPS Y-axis (paper feed) 3D printed gears I have been working on. The image with the black mark on two teeth is a gear that broke after about 800 printing minutes (V9). The teeth are holding up just fine which was my initial worry, one of the spring perches broke on this gear. These break often on the original gears as well. You can see it broke right at the top of the chamfer that I added to strengthen it. There is another picture of a new V9 gear so you can see what it should look like.

The other picture is a V10 gear which has a chamfer that runs top to bottom. The base of the perch is now about 1.8x as wide as the top. The continuous chamfer prevents the concentration of stress where the V9 perch broke.

The 4th picture is the inner gear, the side closest to the printer's build plate. On a resin printer this is the side that has a lower quality (less detail) than the side away from the build plate. The number '10' embossed into the plastic is 1mm tall and 0.1mm deep. I did this originally just to see if the printer could resolve this fine detail on the build plate side. With just the right resin, with the part at the correct angle to the build plate, etc. it does a good job. Amazing results on a $300 printer.