Thursday, March 26, 2015

Nuts and 3d printed plastics


M3 and M4 screws are a very easy and strong way to connect 3d printed plastic parts that possible need to be taken apart again later. Gluing parts is another very good option, but only when you never need to take them apart any more.

The nuts can be either left on the surface of the plastic like here:


A bit more convenient is to embed the nuts in nut traps: either horizontal of vertical. They are then hold in place and don’t move when mounting the screws.


The drawbacks if the nut traps are various:

  • More work to design them.
  • Sometimes need to add extra support layers to hold the layers on top of them. Particularly is this needed when the nut trap is on the bottom of an object.
  • The horizontal ones do take away a lot of material, weakening the object. They can only be used in larger objects.
  • Need to have access to either the side or the back of the object.
  • Depending on the object the screws needed can get very long, because they have to go through part or whole of the object.

Then I came across these nuts you melt in the plastic. They solve lot of issues.


One is inserted into the plastic part.  They only need very little space and material, but still hold the plastic very strongly.

They come if lots of all kinds different lengths, diameter and screw sizes. The ones shown in the picture are M3 x 4mm high x 4mm diameter. They are mounted in the plastic using a soldering iron set to 250 degree for ABS.

Here is an example of use of them: hold the spring screw.




To make a pinball you need a surprising amount of things. Even when I making most parts of the pinball myself, you still needs lots of them. My desk is full of stuff.

I just got my parcel from Marco with rubbers and balls. I ordered the FirePower rubber set. Likely  most rubbers will match, but some will not because I will change some parts on the playfield.

Marco uses a strange company to handle the international orders : IGlobal. I paid 5 dollar extra for tracking. But do you think they the send me a tracking number?  Had to ask for it. First they email I didn’t pay for tracking, but then it occurred to them I had, and send me one. Still a useless tracking number: it was not tracked, even when the latest due date of the parce was just 4 days away. Suddenly it popped up:  It was only tracking the part in New Zealand ! What a waste of money for a tracking number! The only good things ware that the customer service answered very fast and the parcel did arrive.

Likely will not order much any more from the USA: the shipping cost is killing !


It was nice to see that the rubbers fit the flippers very well .



Also received where the WS2812 addressable RGB leds. These are quite nice because the vendor already soldered a 10cm lead between the Leds. So in most places this will be long enough to go from one playfield insert to the next. This will save me a huge amount of soldering.  In some places the distance will be too long, and I will need to extend the wire.


With use of a little bracket (still to design) the Leds will be mounted on the bottom and top of the playfield. Still need to try them to see how bright they actually are, but that is not straight forward because these Leds need to controlled with a special digital pattern.


Wednesday, March 25, 2015

Eject shield


Eject shield

The eject shield is the piece of plastic that hold the ball in a ball trap. This is the red square part in the pic.

tn_5H5A6358 - crop

For the design of the ball eject assembly, I first drawn up the normal eject shield. This can then be used to as a base for my own one ball eject..

Here you can see the eject shield plastic from the top:


The problem as always that I don’t have access to one. So have to resort measuring some pictures and then hope you have the correct sizes. With these make the following design:


The ramp is different, because the two little points in the original design are really to fragile in a 3d printed part.

Interesting is that the ball is not centered in the hole, but is offset by about 2mm. This allows the ramp to extend to the top of the 12mm playfield.

Some other views

t_eject2 t_ball

Viewed as printed part.



The eject shield  design is published in Thingiverse:  . Also the pinball ball you see in the picture is found there:

Saturday, March 21, 2015

Firepower Pinball


This website is dedicated to the firepower pinball: and has lot of info about the game. What is missing is a good scan of the playfield.

Firepower was one of the most sold pinball machines and was designed by the Williams company (that has since closed doors)

The FirePower playfield looks like this:


These are the active elements on the field:


1 01 Ball Release ( trough )

2 08 Ball Ramp Thrower ( trough )

3 04 Left Eject Hole 1

4 05 Right Eject Hole 2

5 06 Upper Right Eject Hole 3

6 07 Left Ball Saver Kicker

7 17 Top Left Jet Bumper (pop bumper 1)

8 18 Bottom Left Jet Bumper (pop bumper 2)

9 19 Top Right Jet Bumper (pop bumper 3)

10 20 Bottom Right Jet Bumper (pop bumper 4)

11 21 Right Kicker ( slingshot 1)

12 22 Left Kicker ( slingshot 2)

13 -  Right flipper

14 -  Left flipper



To do

The playfield contains 3 types of elements: active with coils, sensors and playfield objects .

First, I started with designing the active parts. Then the sensors, the playfield objects, the playfield cad and decals.


Active parts

2x flipper (l,r)

1x ball saver

3x eject

2x slingshot (l,r)

4x pop bumper

1x trough (2 coils)



3x 3Target

1x 1target

7x rubber switch

1x spinner

The rollover switches and other lane switches are replaced with induction sensors.


Playfield Objects

Playfield objects:  ball guides, pins, covers, inserts etc.

Not defined yet.

Glue ABS plastic


Glue ABS plastic

ABS plastic can be glued with acetone. This is a solvent of ABS and it temporarily makes part of the plastic liquid . Couple of drops and it is fixed.

Alternatively, superglue does also work well to glue ABS and PLA plastics.

Acetone is a bit messy to work with. It is very smelly and evaporates very fast. So I needed a way to minimise evaporation and easy application. I found a tip on a forum to put it in a syringe. I was afraid the acetone would dissolve the syringe, but that did not happen.

It is very easy to use with a syringe. Just by holding it in your hand small drops appear due to your hand warmth. So you don’t need to press the plunger.

Because the acetone evaporates very fast, the parts are fixed together within 30 seconds. Same speed as superglue. The acetone is very fluidly, so it fills tiny gaps very nicely.

Also you can store the acetone in the syringe. There is only a very small hole where the acetone can evaporate out. Even after storing it for a couple of weeks it looks like no acetone has evaporated. I normally store it upright, with needle pointing up.


I just took a normal needle and grinded the tip off:


The ball saver


Ball saver

The ball-saver is an assembly normally associated with the left outline. It can return a ball back into the field if the player has met certain conditions.

This is a picture of the ball saver from a FirePower game:


( again: source of picture unknown )

This is one of the few coils in a pinball machine that are actually on top of the playfield.

Because it has to hit the ball in the middle for maximal effect, the height must match the ball. A normal pinball ball is 27mm in diameter, so that sets the diameter of the coil too. The alternative would be to mill a hole in the top of the playfield  and sink  the ball-saver into the playfield a bit.


The model

Cad model of the ball saver in rest/idle position


Red is coil  yellow is frame  orange is coil holder. grey is plunger . Green is plunger stopper and blue the top that actually kicks the ball.

Not shown in the cad is the spring that is used to move the plunger back to its rest position, and the grommet to dampen down the return movement.

This is the ball-saver is active position. The spring is stored inside the green plunger stopper.


This is the plunger.


grey is metal plunger . Green is plunger stopper and blue the top, both plastic . The groove in the green part will hold the spring.

The plunger is hot inserted into the green stopper. Im always surprised how good a bond this does give between the plastic and the metal.  The blue top and plunger are connected with an internal screw.

The final result

The coil has about 600 turns of 0.5mm. I planned for 850 turns like the original but could not fit more windings. It still has more then enough power to push the ball away. Still have not found the test ball back .



Detail view of the spring and stopper.


I have a box full of springs, but do you think I could find anything matching what was needed?  I was already browsing the web trying to find something when I thought: why not make one yourself ?

And that was so much easier then expected! Just wind some 0.6mm spring wire around a 8mm former, and voila: the spring.


Only pull it out till the needed length. So much faster then trying to source it from somewhere.

So in the end about every part of the ball-saver is made at home. I estimate about USD6 in total cost.


Couple of space invaders came in to check out the spring.  Thanks to the honeycomb treatment they received, they are not aggressive any more.



Next stop: the ball eject.

The flipper actuator



The 3d model


A typical flipper assembly looks like this:


(sorry source of picture unknown)

I tried to design the 3d model similar to this with all parts placed on a baseplate. But that did not work out that well. It was not very strong and a huge baseplate.

So tried another approach with a beam that holds all parts. That did work out much better, and was using the capacities of a 3d printer and plastic better.



The yellow bit is the base of the flipper. The blue and green are the movable parts. Red is the coil and orange the coil holders. The vertical red cylinder is the flipper shaft. Cost me a week fiddling to get all the dimensions and movements correct.

The small light green rectangles slots are nut traps. You press a nut in here, and then you can use a screw from the top.



The coil holder


Interesting part is that complexity is no issue with a 3d printer. So I designed a thread between the 2 parts of the coil. You screw both parts together and then glue them forever together.So the screw function is only used once.



tn_IMAG1536 -crop



The flipper assembly

After letting my mendel90 run for a couple of hours here the result. A fully functional flipper made out of plastic where possible.

For the first test wound the coil with 0.68mm copper and 300 turns. The coil takes a huge amount of current. With 3 amps the plunger barely moves. With 5A it just decides to flip. On a 10A 24v supply it works much better. The plastic test balls fly everywhere. The on duration is hand timed and thus not short enough, so the coil heats up fast.


In contrast to the standard flipper assemblies, this model uses real bearings. No more issues with worn out nylon parts! You cannot see the bearings in the picture above, because they are inside the plastic.

This is the support seen from the other side showing the bearing



The plunger was made with a 10mm silver steel rod.  Really pricy are these here in New Zealand !  Bit of hacking with an angle grinder to make the slot.  The small bit on the right is the plunger stopper. This is melted into the plastic on the end of the coil.




Design issues

shaft arm

The arm holding the flipper shaft causes most design issues. The issue is that in order to hold the flipper shaft non slip, you need a lot of pressure. But the shaft and the plastic are smooth and slippery.

The needed amount of pressure is just on the edge of what ABS extruded plastic can handle. I made 4 designs. Some suffered from ABS bending to much, and preventing enough pressure. Others suffered from fracture, because the pressure of the abs was just too much.

See here the various designs that failed.


The top was too flexible and did not grip the shaft. The middle was better, but cracked because it was too stiff. The bottom one cracked between the screws.

The current one looks ok, but don’t know if it will holdup under load. Likely needs to be printed with PC (polycarbonate), a much stronger plastic.

press fit bearing

Also making bearing press-fit is tricky. Either the holes are too small and it breaks or it is too loose and does not function.

In this print the fitting was too tight, and the plastic did split when inserting the bearing





I came across the blog of the homepin manufacturer , and he had some drawings of their flipper mechanism: (picture below © homepin)


Surprise: They used the same approach for the flipper ! And he is also using bearings. Their design is more square, because they mill them out of metal.



Next challenge: the ball-saver.



Started with the flipper design. Thingiverse had some examples like , so could follow their size. ‘Dawn’ it in openscad 3d design tool. Drawn is a possible the wrong word, because openscad looks more like a programming langue.

I tried also freecad, but that is not mature enough at the moment. It has the potential to become a good cad tool, but there are just to many bugs in them to be usefull. Also I like a bit more parameterazable  designs. 

Openscad is very extreme in parameterazable  : it only works that way.

Anyhow. This is the result:



The 3d printed result:


The flipper is printed this with transparent ABS. I have a filament maker and quite a couple of Kg of this transparent ABS as granules, so most parts of the pinball will be printed in transparent.


The shaft is made with a 6mmx70 silver still rod, with a flat facet for extra grip. The rod is mounted in the flipper by heating it and melting it in the flipper. This gives a very strong bold.

Still waiting on my shipment from with the pinballs and rubbers, before I really can test it. For sofar the top playfield part of the flipper. Now need to design the actuator for it.

Custom Pinball


I came across this blog: . What a nice project to do!  I also like to build one!

But with a difference: All custom pinballs are still build with metal parts either from old pinballs or newly made. But why buy it when you could make it yourself? I like to make most parts above and below the playfield with a 3d printer.

This project I plan to do in 2 phases:

  1. First phase is to build a custom pall based on an existing playfield. The focus is on the creation of a framework for pinball . Designing the playfield thingies like flippers and ball troughs. Also designing the electronics and software. CNC routing and playfield printing.
  2. Second phase is to design my own playfield. Focus is on game design, rules, music etc.


So start with phase 1!

I like the relative simplicity of the Firepower pinball, so also going to use Firepower playfield.


This game does not use ramps, but still has a lot of interesting features like spinners , pop bumper , ejects etc. Also, I like the SF theme as I am a serious SF fan.


Some of my requirements/ ideas

  1. Because I not going to use standard parts, I can use all metric parts. Easier to get, and more attractive to design with.
  2. Also try to use modern sensors and electronics.
  3. Use a 24v supply. Try to use a single power supply.
  4. Distributed controller screwed on the bottom of the playfield
  5. Because I don't like the big size of standard pinball housings. It fills your living room so quick, WAF issues. I aim at a thin box I can easily store. So this also means that all the stuff below the playfield is restricted in height. (so no standard ball trough)
  6. So also get rid of the black box. The player has to do it with a couple of simple 7 segment displays for the score and ball count.
  7. Nice to have: Have the possibility to leave the balls in the game, when storing or moving the pinball.
  8. No need for coin changer. Idea: pay with bit coins!
  9. I also like to use the programmable serial LED of the WS2812B. This will save tons of wires .
  10. Nice to have : Wifi

Some of the consequences of making my own metric stuff is that I need to make my own coils and plungers. Not a biggie

Also plan to create my own inserts: need to test this, but likely can print them also on the 3d printer. Alternative cut them with a laser cutter or CNC machine.

For playfield prints I plan to use either waterslide decal or (pressure adhesive) polyester.

I still need to decide if I going to base the distributed control on an AVR or ARM processor. Also what to use as a main controller: Raspberry Pie or a AVR/ARM processor. What physical interface between them. But those are problems for later.

The issues I face at the moment:

  1. There are hardly any 3d models around of the standard playfield parts, so need to design them all.
  2. I don't have access to example playfield parts, so it is difficult to get dimension of existing parts. Likely end up measuring photos of playfield parts.
  3. I am unable to find a good quality scan of a Firepower playfield.

(and of course time)

So let get started with the plastic design first!



This is my brand new blog. First for me.

The following subjects will be addressed here:

  • 3d printing models
  • Filament maker
  • Custom pinball machine
  • Interesting thing  I come across

Google blogger does not have the option to sort the blocks from old to new. Only new to old. But that is not really handy if you like to follow the progress. Also blogger has a lot of ‘junk’ on the bottom of the blog that is difficult to get rid off.

If you would like to read the blog from begin to end I suggest the following approach:

1) select the welcome blog from the right –> do it now!

2) after reading the entry, scroll down and click on ‘newer post'