Hi, I'm following a similar path with thinking about and trying to write the code for converting a GTB sys1 to OPP + MPF. If I can get this achieved I thought I would write my own expanded ruleset for the game.

As I understand it the new ST processors do not chain serially, but will use the usb connection so you may need a hub to "star" connect each ST processor.

Are you aware of the plank boards for OPP - there are some restrictions with the newer ST processors. I think the plank only fits on one side due to different pin arrangements on the ST. I believe the new processor will be able to support serial leds which may change your thinking on lamps as it would allow a string of led to be run from just one pin.

Not sure how you're making your printed circuit boards, but some of the gerbers I got from the OPP repository didn't seem to work when I uploaded them to SeedStudio - but that may be something I'm doing. It's a fair way down the track for me, so I put it aside until I made good progress with the code.

Good luck with the MPF programming - I'm finding it has so much flexibility that there are many ways to achieve everything and it's hard to work out which way is "best", sometimes it's hard enough just getting something to work at all. I bash away and usually make some progress there but it's a steep learning curve and takes lots of time. The MPF documentation also seems to lean more to DMD style machines with an emphasis on slides which is a little different to what I, and I think you, will be trying. This is not a criticism of MPF by the way, it's great that it's available for us and a massive effort from those creating it which i really appreciate.

Good luck and I hope you keep this thread updated as you progress. I plan to start a thread on my game when I feel I have enough to say.

Sorry BENETNATH for the thread hijack, but it might be relevant to you.

Mezelmods are out of stock on solenoid and incand wing boards - I don't know how often they restock - can't imagine demand is that high and also typical shipping costs from USA to Australia are really high, so that's why I was looking at the board making houses.

I grabbed the gerbers from here: http://svn.code.sf.net/p/open-pinball-project/code/trunk/Kicad/

Below is a screenshot from the Seedstudio gerber viewer - it looks like the drilling mask is not located correctly? Maybe it's something I'm doing (I just upload the zip) or maybe a viewer issue? Panel-1019.zip and Panel-1020.zip look good, Panel-1013.zip, Panel-1014.zip and Panel-1018.zip all seem to have the misaligned drilling issue.

Thanks for any guidance.

panel-1013 (resized).png

If you configure the flipper as two coils then you take away the need for a functional eos which some may consider an advantage. For me I'd use eos and take less wiring. I guess you could just wire up the power winding and then use software to reduce hold strength, then it's just one coil and no eos.

Benetnath, if you run the GI through a controlled circuit you gain more flexibility for attract mode and special effects. If you used leds then the current draw would be lower so you could use them in strings by area. Anyway, just a thought.

openpinballproj. Thanks for the comments on gerbers. I hadn't thought about it enough, but with the move to the STM32 processor perhaps people wont use the plank as there is no need for the interface part. I guess it could be used and just not populate the interface bits and you gain the structural rigidity.

My take on the gerber drilling masks is:

1013.zip, Panel-1013.zip, Panel-1014.zip, Panel-1018.zip all look to have the drilling not aligned correctly.

1019.zip, Panel-1019.zip, 1020.zip, Panel-1020.zip are good

I don't know if this drilling mask matters or not - I just loaded the files into the viewer and got results that didn't look "right" to me.

Can you tell us which pins you are thinking are likely to eventually be used for led strings on the STM32 processor. I'm planning a layout and this may influence which wings go where. To me the leds will be a great addition to OPP - cheap, flexible, lowers pin and wing count and pretty cool. Thanks for your efforts.

openpinballproj, thanks for the comments.

On leds I'm guessing that it will be config switch to say if the MOSI pin is used for serial leds or as a switch input? With this in mind would it be an idea or possible to make it available on both MOSI channels? (I don't understand how the STM32 labels pins 17 and 41 both as MOSI1 - is this why you chose MOSI2?) I'd like to avoid long strings of leds for voltage drop and reliability issues and just design elegance. If one led in a string fails all down stream don't work. With say about 60 normal controlled feature leds on a pin with maybe three STM nodes it would still be 20 leds per string. That's if people don't go down the route of controlling gi and I'd guess bearing in mind the cost and flexibility of serial leds many will chose to put gi on the strings. I'm planning to not use any incandescent wings - just all leds.

The level shifting is easily done with logic level converter - about $1, plenty on ebay etc, typically they are bi-directional but you only need outgoing 3.3 up to 5V.

Benetnath - have a look at the pre made strings of Neopixel leds - saves lots of soldering!

On flipper coils, I think Bally used serially wound coils with a eos that shorted out the hold winding for the initial power kick, then the eos opens putting the hold winding in series with the power winding. Best to work out what type you have, both this and the parallel winding described in post #20 earlier could be used with a single output. If you wanted you could unsolder the winding wires from the flipper coil lugs and rearrange to suit whatever you wanted - parallel or serial or just one coil or both - the world's your oyster!

On gerbers, I had a quick look at 1013.zip, panel-1013.zip and panel-1019.zip these all seem to have two copies of the top and bottom copper and silkscreen files, this doesn't look correct? 1019.zip looked good.

I went through the gerbers 1013, 1014, 1017, 1018, 1019, 1020, panel-1013, panel-1014, panel-1017, panel-1018, panel-1019, panel-1020and have sent a PM to Openpinballproject rather than cluttering up this thread - best to wait until he comments.

Openpinballproject - on leds I guess I just have to get over my dislike of long serial strings - they just seem clumsy to me, but if that's my biggest concern then life is going well With your suggestion to use a solenoid for gi control, then you have pwm for fading but no colour ability. I had been thinking of gi colour changes tied in with modes which would be cool. I guess adding an extra STM32 would be an easy and cheap way to gain another led string output in the worst case (or to solve my tastes!)

Benetnath - don't stress over the flipper coils, most arrangement will just work and with software you can fiddle with funky modes etc

How is your MPF work progressing? I've found it a steep learning curve, but think after following the tutorial that coding an existing simple / old game is a good way to learn. You also just focus on the code rather than trying to be creative with rulesets - for me I plan rule extensions as a second step. As cobra18t says, it is so flexible it's hard to work out the best way, or even how to begin, and when it doesn't work as expected it can be hard to trace what is the issue. I also have moments of stupidity - I spent a few hours one night trying to work out why a led would not illuminate - tip don't set the default colour to 000000 The coding also makes you focus on the minutia of how things work - as an example early WMS solid state games light the bonus with all lower lamps on so 1 stays on, then 2 illuminates, next 1 and 2 stay on then 3 lights etc. GTB don't behave this way - only 1 or 2 or 3 is on at one time until you get greater than 10, then it's 10 and 1 etc. These things all take time and consideration. I'm finding the final little bits take a long time and it would really help to have a working game to play with your finger. Good luck and keep us informed as you progress.