Cobrapin not forwarding switch events to MPF?

[Aziz Pinball]

Hey Folks,

I've been having this issue for years and it was never a big deal, but now I'd finally like to solve it: I run a bunch of Cobrapin boards that I bought from Thomas in 2022. I noticed quite frequently that when my slings or jets trigger, they do mechanically work, but the event does not seem to be forwarded to MPF, so the corresponding light show, scoring and sounds are not executed. This behavior is erratic but reproducible.

A second issue that may or may not be connected to the first one is that if I spin my spinners, the sound/light does not trigger evenly but is often intermittent, as if there are switch events missing.

Thanks for your support!
-zap

Here's my config for the switches in question:

switches:

# Main Board, J3, pink wires

s_spinner_center:

number: 1-0-10

# Satellite Board, J1, purple wires

s_slingshot_trigger_left:

number: 2-0-23

debounce: quick

tags: s_slings

s_slingshot_trigger_right:

number: 2-0-22

debounce: quick

tags: s_slings

s_jet_trigger_left:

number: 2-0-16

debounce: quick

tags: s_jets

s_jet_trigger_right:

number: 2-0-18

debounce: quick

tags: s_jets

s_jet_trigger_lower:

number: 2-0-17

debounce: quick

tags: s_jets

[Ben Tripoli]

I see this also, mainly with my slings. I do notice however its usually the ball barely touching the slings and it seems to half kick, like the coil is not at full power, and like you, I hear no sound associated with the kick. I never really bothered to chase it as my ACNC shows the exact same behaviour, ACNC is Proc hardware and I believe its MPF, So I thought it was something to do with MPF.

With your sound on your spinner, just check you aren't going over the simultaneous_sounds: limit on your sound setup. Also I put a small ignore_window on my spinner to smooth out the sound so it wasnt intermittent. A hard fast hit would create too many events for sound to keep up. 

[Aziz Pinball]

Hey,

I've not experienced half kicking, even when the sling fully fires though, the event is either not sent out or not received or processed by MPF from time to time. That is annoying, because I use the slings to switch the mode of my drop targets and sometimes you want to shoot them intentionally to switch the mode. And if that does not work half the time, its frustrating. You can see this behavior in this video:

https://vimeo.com/1070237363?fl=pl&fe=sh#t=1m6s&end=1m11s

Ignore the sounds though, they were added to the video in post. Both slings are supposed to toggle the target bank green/red and also show the green line on the slingshot plastic edge when they fire.

As for the spinner, I will look into the sound settings again. Might need to increase the simultaneous_sounds a bit.

sound_system:

master_volume: 1.0

buffer: 65535

channels: 2

tracks:

music:

type: standard # sound_loop

simultaneous_sounds: 1

volume: 0.5

voice:

type: standard

simultaneous_sounds: 1

volume: 0.7

sfx:

type: standard

simultaneous_sounds: 16

volume: 0.9

[cobra18t]

I will take a look at this assuming it is a hardware/firmware issue. There still might be something in MPF contributing to it, though. We are working on some firmware updates for CobraPin now, but it won't be a quick turn around. I have also seen this issue but it was never a real issue for me and I thought it was just the sound effects getting stomped on.

I am guessing at this point, but it might be that the switch inputs sent to MPF are further filtered vs what is used for autofiring the solenoids. 

As a quick hardware fix on the slings, you may consider a small capacitor across the switch. You will have to experiment with the value. I would start with 0.1uF and consider moving down from there if you see other issues. Remember, if you only have 0.1uF capacitors you can create smaller values by putting them in series, and larger values by putting them in parallel (opposite the behavior of resistors). The purpose of the capacitor is to make the switch hit appear longer to the firmware and smooth out any switch bounce that may be occurring. Knowing what capacitance improves the behavior may also help me find the root problem.

[cobra18t]

Also, for the spinner, I had better luck when I also set the simultaneous_limit of the sound. 8 worked well for me. It likely will not solve your problem, just an optimization that helped me.

https://missionpinball.org/latest/config/sounds/#simultaneous_limit

[Aziz Pinball]

On Monday, February 2, 2026 at 2:50:04 AM UTC+1 cobra18t wrote:

As a quick hardware fix on the slings, you may consider a small capacitor across the switch. You will have to experiment with the value. I would start with 0.1uF and consider moving down from there if you see other issues. Remember, if you only have 0.1uF capacitors you can create smaller values by putting them in series, and larger values by putting them in parallel (opposite the behavior of resistors). The purpose of the capacitor is to make the switch hit appear longer to the firmware and smooth out any switch bounce that may be occurring. Knowing what capacitance improves the behavior may also help me find the root problem.

@cobra18t just a shower thought: I know EEs have capacitors on some of the contacts to stretch the signal length, but those use switch matrices. I'm using direct switches though, which are low active, and closing a switch connects the GPIO to GND. This means there is no voltage for a capacitor to work with, right? 

[cobra18t]

There is a weak pullup on each of the direct inputs to 3.3V. If you had a capacitor, closing the switch would short out the 3.3V on the capacitor very quickly and the voltage would go to 0V. When the switch opened again, it would take some time to charge the capacitor up again through the weak pullup to 3.3V. So it would still have the pulse stretching effect.