Switching flipper coils with MOSFET

MisterBill2

Joined Jan 23, 2018
18,176
OK! Now you have a drawing that should be built from, AND you know more of the stuff that is seldom covered in classes, at least not directly. The things that I suggested can be deduced through circuit analysis, and even be proven. BUT certainly they are not obvious to most folks with recent degrees.
 

crutschow

Joined Mar 14, 2008
34,283
My design now looks as follows:
What's the purpose of C1?
That capacitor will tend to burn the relay contacts due to the momentary high current when the contacts close.

Is that the capacitor that you originally had from the power supply to ground (where you should have one)?
 

Thread Starter

Raymond175

Joined May 28, 2019
57
What's the purpose of C1?
That capacitor will tend to burn the relay contacts due to the momentary high current when the contacts close.

Is that the capacitor that you originally had from the power supply to ground (where you should have one)?
In older pinball machines the E.O.S. switch (S1, normally "closed") is used to switch the coil from low resistance (powerful "kick") to high resistance (weak "hold") by mechanical movement. Due to the fact that high voltage is crossing that E.O.S. switch, the contact points are known to wear out very fast due to arching/sparking. The pinball community reduced that wear by adding a capacitor (C1) to reduce that arching/sparking. But this E.O.S. switch is known to be the component that requires the most maintenance (filing contact points, replacing, etc.)

But your question started me to wonder whether or not to directly build the microcontrolled flipper mechanism version (known as Fliptronics II) instead. In newer pinball machines the E.O.S. switch (S1, normally "open") is not carrying high voltage (to disable low resistance path), but to tell the microcontroller that the flipper is up, having the microcontroller switch the high voltage from the low resistance path to the high resistance path.

I'll draw the schematics for this new setup and see if it is wise to indeed build this directly, instead of having that as a version #2. This would also allow me to implement a testing mode for the flippers to test for component/wire failure and/or wear by powering the flippers (output) and checking if the flipper is up (input) and the time between these two events.
 

Thread Starter

Raymond175

Joined May 28, 2019
57
So be removing the need for capacitor C1 and by introducing a microcontroller controlled coil resistance switch I've created the following schematic (I hope it makes sense):
Diagram (v2).png

* Digital output D1 will be enabled by the software when the player presses the flipper button (input for this not in this schematic!)
* When digital output D1 is enabled, MOSFET Q1 will activate the coil L1 (powerful "kick")
* When coil L1 is activated, the E.O.S. switch SPST1 will be closed by mechanical movement of plunger, enabling digital input D3
* When digital input D3 is enabled, digital output D1 will be disabled by the software and digital output D2 will be enabled
* When digital output D2 is enabled, MOSFET Q2 will activate the coil L2 (weak "hold")

Of course I will also build-in a protection measure than when digital output D1 is enabled for more than a few microseconds (e.g. 500?) then the system must be shutdown with an error on the screen for the flipper failure (e.g. inproperly adjusted switch, component failure, bad connection, broken cable, etc.) to prevent coil L1 to burn out.

I hope this makes sense?
At least it will be harder to have this easily on the PCB without crossing lanes :)
 
Last edited:

MisterBill2

Joined Jan 23, 2018
18,176
What's the purpose of C1?
That capacitor will tend to burn the relay contacts due to the momentary high current when the contacts close.

Is that the capacitor that you originally had from the power supply to ground (where you should have one)?
NO, the switch will not burn contacts when it closes because the voltage is off when it closes, since it is a normally closed switch that opens after the flipper moves.And this it does not need that protection diode across the the higher resistance coil, because the capacitor already prevents the spike by absorbing the voltage when the switch opens.
 

Thread Starter

Raymond175

Joined May 28, 2019
57
Interesting article, however I doubt that that does help out when the ball is hitting the flipper bat hard... the "weak" hold will break and the ball will drain. With the E.O.S. switch, the switch will re-engage the powerful kick as soon as the ball hits the flipper bat, and if adjusted well enough, the ball will not drain, nor will it be kicked up again in an unexpected direction, it will "just" bounce in the expected direction.
 

crutschow

Joined Mar 14, 2008
34,283
It's not really applicable here since the solenoid in question has two coils, but you can use PWM to efficiently lower the current in a solenoid to the hold level after the initial application of full-power.
An Arduino can do that of course.

Below is a simple Hit & Hold circuit that also works:

1612302501800.png
 

Thread Starter

Raymond175

Joined May 28, 2019
57
It's not really applicable here since the solenoid in question has two coils, but you can use PWM to efficiently lower the current in a solenoid to the hold level after the initial application of full-power.
An Arduino can do that of course.

Below is a simple Hit & Hold circuit that also works:

View attachment 229390
I know I can do it, but that would still not prevent the ball from hitting the flipper bat hard and "breaking" the coil "weak" hold. The Arduino doesn't know that the flipper bat goes down, unless there is something else attached to it, like the E.O.S. switch.

I know that every component can be replaced with something else, but in this case I want to use as much as possible the official Williams flipper components for the electro-mechanical circuits. For the rest I'm not using the Williams PCBs, so that is what I can (and want) to change.
 

MisterBill2

Joined Jan 23, 2018
18,176
I know I can do it, but that would still not prevent the ball from hitting the flipper bat hard and "breaking" the coil "weak" hold. The Arduino doesn't know that the flipper bat goes down, unless there is something else attached to it, like the E.O.S. switch.

I know that every component can be replaced with something else, but in this case I want to use as much as possible the official Williams flipper components for the electro-mechanical circuits. For the rest I'm not using the Williams PCBs, so that is what I can (and want) to change.
In addition to those reasons, an arduino would be way too slow for good flipper action. AND PWM will be way too complex for the number of instances used in a pinball machine. And why in the work would anybody choose to remove the embedded diodes in an existing coil instead of simply not adding the extra redundant diode that serves no purpose???
 

Thread Starter

Raymond175

Joined May 28, 2019
57
Most flipper coils have no diodes.
Almost every pinball coils I see in machines and in shops have diodes.
But there are substitute coils that don't have the diodes, so I might better add the redundant diode D3 to prevent issues when someone replaces the coil with one without the diode(s).
1612357473483.png
1612357770004.png
 
Last edited:

Thread Starter

Raymond175

Joined May 28, 2019
57
A good reason would be to use good diodes in place of the fake chinese ones.
I'm not sure why I should doubt the diodes on the coils that are used by the pinball manufacturers, but like I said in my previous comment, there are substitute coils that don't have the diodes, so I might better add (redundantly) to my circuit design to protect my boards.
 

Thread Starter

Raymond175

Joined May 28, 2019
57
You just have to look at how much diodes are good and you'll immediately understand.
Well "unfortunately" I don't have any experience with that, but I'll re-add diode D3 to protect the circuit, because in the future someone might replace the coil with one without the embedded diodes, and then the MOSFETs will blow...
 

Thread Starter

Raymond175

Joined May 28, 2019
57
So one last time to verify my design, now with the (redundant) back EMF diodes (D3 and D4) back in the schematic, just in case the coil assembly doesn't have the diodes D1 and D2.

Also two more questions:
1) does it matter if the diodes D3 and D4 are on the PCB and not "near" the coil itself?
2) does the digital input switch need a capacitor to prevent "noise" due to arching when the switch opens/closes?


Diagram (v2).png
 
Last edited:

crutschow

Joined Mar 14, 2008
34,283
1) does it matter if the diodes D3 and D4 are on the PCB and not "near" the coil itself?
Actually, if the diodes are connected directly between the MOSFET source and V+, they will help absorb any small kickback from the wiring inductance.
2) does the digital input switch need a capacitor to prevent "noise" due to arching when the switch opens/closes?
Don't see why.
The diodes are preventing any significant arching, so there should be little noise.
 

Thread Starter

Raymond175

Joined May 28, 2019
57
Hi,

I'm trying to reverse engineer my pinball machine flipper end-of-stroke "normally open" switches that provide input to the CPU about the flipper state (fully energized or not). The wiring consists of the following:
- black/blue (J208-12) +5V left flipper EOS switch
- black/green (J208-13) +5V right flipper EOS switch
- orange (J208-14) GND flipper EOS switch

So based on this wiring and pin numbers I know that the two input switches are using three pins on the CPU board (PCB): two V+ and one shared GND.

1613039363595.png

The design I originally made required dedicated return wires for the CPU inputs:

1613039370127.png

How would the one from my pinball machine possibly work if both return paths are connected to the same ground? It must be that the inputs are then connected to the V+ side.

This is probably something very simple, but I fail to see it...

Thanks in advance,
Raymond
 
Top