Quoted from PoMC:applejuice I think quite a few of us Twister owners would like to hear your thoughts on why many of the magnet boards have failed over time. What have you found to be the most common cause of failure? And finally, could you explain the service bulletin updates and what those modifications resolved - such as removing the neon bulb and adding another diode in parallel with an existing one.
http://arcarc.xmission.com/Pinball/Stern%20Service%20Bulletins/sb/sb91.pdf
Great work you're doing and thanks for sharing your progress and your process.
No problem i will go through a few things and explain the designs once i get further along, as well as other things that i have learnt regarding how the boards work in twister (and golden eye). I should point out that my design does include all the service bulletin updates as well as my own improvements.
I have also been experimenting with what likely failure points there are and what happens if certain components fail or are missing etc.
As a quick reply though , i believe there are a few causes of pic failure., Firstly the age of the chip as it is much more complex internally than just a std logic ic. Microchip have improved the production of the chips over time, being flash ram based now, but most complex ics (processors, ram, i/o ics) are borderline when plus 20 years old. Roll all of them into 1 single ic and you have multiple failure points.
Secondly the pic ic is reset repeatedly during the game cycle (at every ball start) this is so the board can reconfigure itself and keep an accurate picture of the magnets/core for detection. The design of this circuit was improved in the service bulletin as if the chip doesn't come out of reset reliably then the board is disabled. This is another reason why i've included a reset led on board. The reset line on all cpu type chips is a key area.