Low Force Joystick Grommets

jlpmauldin

jlpmauldin

Well-known member

Donor 5 years: 2017-2021

VAPS


Joined
Apr 26, 2017
Messages
1,029
Reaction score
533
Location
South Carolina
Earlier this year, I was incredibly fortunate to pick up this beautiful Robotron for a crazy low price from someone who wanted it to go to a good new home. It had been in his private collection for years (many of those in storage). It even still mostly worked. It was a super lucky get for me.

The game needed many common repairs: power supply rebuild, ribbon cables, RAM, CPO, and some new plugs for the wire harness. No big deal. Here are some pics after I finished the repairs:
full


It also needed new joystick grommets. The old ones were completely dried up:
full


No worries. reproduction 3/8" Wico grommets are easily available.

However, after installation, I was surprised how stiff they played. My older Robo had lighter grommets and I generally got better scores (pretty lame to blame my low scores on the controls but yeah). I figured they would soften up with play, but they didn't. I did a little research through KLOV threads and found that there are no sources for lighter force grommets right now.

I thought it would be a fun project to make my own. This thread will document my results so far.

First Attempt
My first idea was to 3D print the whole thing in one piece. It would have flexible members to allow stick movement, and rely on the inherent pliability of the plastic to re-center it:
full


I really thought this would work. A simple design that practically any printer could produce. Even better, the flex members could be made thicker or thinner to suit any player's needs.
full


This design actually does work and even plays well... at first. But, they break after only a couple of games (at least with PLA). I tried several variations: different member shapes and thicknesses, bushings/no bushings... They all failed during play:
full


Maybe there is some flexible 3D print material out there that would work. If I find some, I may revisit this design.

Second Idea
My next idea was a design more like traditional grommets, only I would use silicone instead of rubber. I would use a printed plastic piece for the mounting plate and press in a bushing for the joystick shaft hole.

My plan was to 3D print a mold for the silicone. The intended shape of the new grommet is shown below:
full


I made it as simple as I thought I could.

Here are the 4 mold pieces needed:
full


All these pieces are easily printable. I'd never worked with silicone, so I would at least learn how to work with that material.

More to come...
 
I first attempted to print the forms using my janky old FDM deltabot:
full


The result was decent. However, print times were super long. Also, all the extrusion lines were very prominent. Even worse, the middle pieces were too thin to get reliable prints. Some elements of my design were just too precise for this printer. I needed to use my SLA printer.

The results using the SLA printer were much better. Smooth surfaces and ultra-precise dimensions:
full


Even better, the prints only took minutes. I could print dozens of forms in a day without rushing if needed.

Another bonus: the resin is semi-transparent. So, I would be able to see inside as silicone filled the cavity.

Here is a form fully assembled. I used four 10-24 screws and wing nuts to hold it together:
full


My plan was to use a large 60mL syringe to fill the mold. The tip was too big to fit into the mold fill hole, so I had to print an adapter nozzle and glue it on:
full


Silicone for mold making is available in a variety of durometer levels:
full


I had no idea what would be appropriate. I ended up choosing a shore 40A silicone, as the hardness of a pencil eraser seemed like a good fit for this application. Also, it was available on Amazon for not much $.

Shore 40A Silicone kit on Amazon

full


This is plenty of material to experiment with on this project.

These posts haven't caught up with my real life progress yet. So, to be continued...
 
I have not tried NinjaFlex. It certainly looks promising. I remember Ben Heck used it alot in projects over the years.

I did try this stuff:
YOYI TPU 3D Printer Flexible Filament

It has the same Shore rating as Ninjaflex. It was too squishy for my printer, and it under-extruded so badly I didn't give it another try.

The main problem is probably my old printer.

Have you tried NinjaFlex?
 
Here are the steps to make the silicone mold:
  • Stir silicone mix part A thoroughly
  • Shake silicone mix part B thoroughly
  • Mix A and B (10 parts A to 1 part B by weight)
  • De-gas silicone mix
  • Inject into the mold (pre-coated with release agent)
  • Cure time 24 hours
Seemed straightforward enough. The only potential problem was the de-gassing step. The old oil-less vane pump I was using in my rig was on it's last legs. It can't produce a very good vacuum anymore. I'm not sure how good a vacuum I need. It was all I had so I went with it.

full


Mixing parts A and B:
full


Part A is a thick white goo, and part B is a green liquid. I made a small amount this first time (50g part A + 5g part B).

De-gassing:
full


My pump was only able to pull the vacuum down to ~11 inHg. This step should produce an eruption of gas up from the solution. I only observed small bubbles slowly coming out. Not good. Too late to turn back now...

Petroleum jelly is supposed to be a good release agent to prevent the silicone from getting stuck to the mold. Before I assembled the molds, I smeared some on all the inside surfaces.

Injecting the silicone into the mold with the large syringe actually worked well. The custom tip started leaking a little, but it filled the molds OK:
full


It was cool that I could see the cavity fill through the mold plastic.

So, here's my first attempt. Not horrible:
full


Unfortunately, the inadequate de-gasssing caused lots of voids in the mold. Also, I did not stir up part A well enough initially, so the cure time was twice what it should've been (48hrs instead of 24). I think not stirring the A well made it weaker than it should be too (although I'm sure all the voids didn't help it's strength either).

Here is it fully assembled. The fitment in the joystick housing was good. I just need to do a better job with the molding:
full


Improvements for next time:
  • Stir part A thoroughly
  • Better degassing
  • No release agent (probably not needed for a simple form like this)
  • Thicker mounting plate. This first one seems too flimsy.
Grommet.3.4.jpg
 
Damn, where's the next part?! I found myself instantly disappointed when I got to the end of this thread. Excited to see where this goes
 
Glitch said:
Hopefully centering spider for 49-way joysticks.
Click to expand...

You read my mind. I think that will be my next project like this.

PassivePac-Man said:
Damn, where's the next part?! I found myself instantly disappointed when I got to the end of this thread. Excited to see where this goes
Click to expand...

Thanks for the interest. I will have another update on this project posted tonight.
 
mattosborn said:
I would design the mold such that it held the mounting plate and the insert. That way when it cures it's one solid piece.
Click to expand...

That is a really good idea. I can probably implement this with the present design. I could just sub the mounting plate for the two center mold pieces. I think all the holes line up. I'm away from my computer right now. Will check tonight.
 
Last edited:
At this point I figured I just needed to make process improvements.

Better De-gassing
Here is my improved de-gassing setup. It has a small vacuum pump used for HVAC work. It doesn't have any trouble pulling the pressure down to <1inHg. I also made a bigger chamber that could fit a 16 oz. plastic cup, which I found convenient to use for mixing:
full


Now my de-gassing step looks just like the instructional videos. Much more effective.

Syringe Holder and Mold Breather Holes
The big syringe was difficult to hold while applying enough force the squeeze the silicone out. This simple holder make from 2X4s held it steady during injection. Also, I drilled small holes in all the molds to help allow air to escape while filling the cavity:
full


Scale Up
Even this small silicone kit had enough material to make alot of grommets. So, I printed more molds. This stuff has a limited shelf life once it is opened and I didn't want to waste it. Also, I just wanted to see how many I could make:
full


Results
This new batch came out MUCH better than the first ones. No bubbles or voids. Just flash to cut off:
full


Here's one with the flash trimmed off. I'm really happy with the result:
full


I also made the mounting plates a little thicker to add strength.

These improvements boosted my yield to 100%. I've got too many now:
full


I may need to get some other players to beta test these in their games.

Installation
full


These installed easily and play really well. I'm quite pleased with the action now. Not stiff at all, and they self-center just like they should. Shore 40A turned out to be a good durometer level for this application.

I think I would like to see an actual force measurement comparing these to the repros. I may be able to make a rig for that. I would also like to explore Matt Osborne's idea to cure the silicone around the mounting plate and bearing. I could also make a version of these that work with other joystick shaft sizes.

I've still got a little left. But, it's been alot of work and I need a break from this. I will pick it up again in a few days/weeks.

In the meantime, I'm going to play Robotron alot and to see if they're reliable. They are holding up fine so far.
 
Last edited:
Damn clever approach, man, bravo! If you need a beta tester for a set, I'll happily volunteer. I'm playing with two of the repro 4" Wico sticks and the grommets are definitely robust. Really curious to see how yours hold up.
 
geech said:
Damn clever approach, man, bravo! If you need a beta tester for a set, I'll happily volunteer. I'm playing with two of the repro 4" Wico sticks and the grommets are definitely robust. Really curious to see how yours hold up.
Click to expand...

OK cool. Let me know how they work out. PM sent.
 
This is valuable information. Molding is probably the best way to go; however, it may be interesting to see if these grommets(shock mounts) can be printed directly with a flexible 3d-printing material like TPE or durable/tough resins. What SLA 3d-printer did you use for the molds?
The last ones I bought (for a Gun Fight) were Lord Corporation shock mounts that I found on eBay, but they weren't cheap. They were sold as helicopter engine shock mounts with the same part number as the parts in my original panel for like $30 apiece. Expensive, but they should last for decades.
 
Glorp said:
What SLA 3d-printer did you use for the molds?
Click to expand...

This one:
ELEGOO Mars UV Photocuring LCD 3D Printer

It's an entry level model. Very low price. But, I've found it to be quite accurate and reliable.

I would very much like to continue looking at the 'print it all at once idea' I started with. I just need to figure out how to use the flexible filament reliably (I've got a whole spool of it and have barely used any). Maybe I need to use a bigger extruder nozzle.

Here's a flexible UV cure resin:
Monocure 3D Rapid FLEX100 3D Resin

Something like this might be good. It's more expensive than regular resin tho, and I have no experience with it, so I don't know if it works with my printer.
 
Last edited:
You must log in or register to reply here.

Similar threads

Galaga joystick refurb
Replies
0
Views
166
JohnV
Lumo
Joystick hole size for Pengo CP?
Replies
2
Views
177
Lumo
Lumo
SuperSpy45
Atari Centipede Cocktail Restoration, Low and Slow.
Replies
16
Views
1K
SuperSpy45
SuperSpy45
BeatEmUp
Joystick Grommet
Replies
2
Views
360
PrairieDillo
PrairieDillo
Sega Afterburner Joystick Cover Artwork reproduction
Replies
0
Views
249
cyberrob666
Back
Top Bottom