The yahoo news article covering the WWE game references Stern's new platform which I assume they are launching at CES as well

http://finance.yahoo.com/news/stern-pinball-ready-rumble-150000003.html

WWE Wrestlemania is the first pinball machine to feature Stern's recently announced SPIKETM electronics hardware system. The new technology reduces product complexity and simplifies manufacturing, resulting in more reliable and easier-to-service games. SPIKETM is both modular and scalable, accommodating and supporting future designs. Players will quickly appreciate higher quality sound, enhanced illumination effects and a more engaging overall game experience

The Verge video shows a bit of the underside of the playfield..

you can see connected PCB boards on some of the brackets and on the PF itself.

http://www.theverge.com/video/2015/1/5/7493233/modern-pinball-machines-linux-wrestlemania-ces

Looks like addressable boards vs one big driver board. Also looks like modern surface mount stuff. Maybe we are being moved to the 'replace the board...' vs 'repair the board' model.

This is double edged...

The same could have been said of the old design too.. simply swap the major boards.

The new design may have smaller boards (so should be cheaper in theory)... but it leads to the question...

Will the boards be uniform and what kind of availability would they have? The advantage of a large driver board in a reusable platform is you have a unit you can be assured will be available in spares and for some time. The more unique a part is, the higher the cost and greater risk of limited availability.

A tradeoff of 'cheaper parts' -- as long as they are available.

From the PR on new features

New SPIKETM system benefits and features include:
Improved earnings:
• All LED lighting enables enhanced lighting effects, increases reliability and attracts new players. Included is the florescent light replacement by LED’s, which last longer, run cooler and draw less electricity.
• A new modern CPU and expanded memory allow the designers greater freedom to create new compelling game scenarios to attract more players and lengthen their engagement.
• A more powerful high definition digital sound system enhances the game experience.
Higher reliability
• New built-in automotive-grade circuit protection for LEDs, switches and solenoids reduces circuit failures from accidental shorts and environmental faults.
• Lower game heat reduces stress on components increasing longevity.
Easier serviceability
• Low-cost game modules are easy for any technician to replace and costly circuit board repair is eliminated.
• Reduced wiring complexity eases troubleshooting and worn component replacement.
• Improved system diagnostics make the system easy to learn.
Greater efficiency
• Reduced chances of failure from environmental conditions with robust solenoid, switch and lighting protection circuitry.
• SPIKETM system games use less energy than older models and generate less heat during operation.
• New controllable LED artwork backlighting reduces fade, saves energy and lasts longer.

So bus system... and built in burnout/blown path protection supposedly

They leave the door open without saying they will do it one way or the other. Clearly they built the system to allow it... the question is, which direction will they go? Sounds like the display board would just be another module on the bus... so their display technologies are open ended.

I'm sure the main board probably has a HDMI right on it for build/debug type stuff... but they could leave game display to another board too. Sounds like its a matter of 'what they chose' vs constraints or not.

Yup - cheap fast storage

It doesn't sound like 'one driver board' exists... but rather addressable boards that may be more specialized/segmented

I think that's just lay translation/marketing muddying things up. By 'computers' they probably just mean 'processors' and they'd be on a shared serial bus for inter-communication.

Imagine a platform processor... a video processor... a switch I/O processor.. etc

They would load relevant stuff into RAM memory. SD doesn't replace memory the processor and I/O uses, just the offboard storage. Its the video stuff one would have to be worried about for storage... no so much game code or audio.

Even (by today's standards) slow SD cards can read in the 6 megaBYTE/sec range. With modern stuff being 45megabyte/sec and up.

How much does that really matter?

It's probably gonna decompress and load a kernel... than it's various drivers. That all takes time, but one shouldn't expect anything that would be a hurdle.

Add into that any power up self-tests the game would likely do.. and its probably all worth it.

What the design does is change from lots of 'one big, daisy chaining wire' to lots of smaller harnesses.

The cat5 bus really just eliminates the bulk of the long hauls and cuts down on the daisy chaining. It cuts down on wire usage.. but increases the # of components and connectors.

You still gotta connect from your driver pins to the solenoids, from your switches to the headers, etc.

Think lots of small cables... instead of your really long christmas string of lights. Integrating things onto shared boards (like the LEDs on the big playfield boards) is where you really cut down on cabling.

The harnesses will be simpler to construct... they'll cut down on bulk wire usage.. and it should be easier to debug these circuits easier.

The new marketing guy told him obviously

Well that's the theory with Field Replacable Units... you stock and keep the swappable units on hand to do quicker repairs than the component repairs we are used to. Swap out "cheap" board.. send broken board out for repair or bench work. Yes you must keep parts 'on hand' - just like we do today... but instead of stocking .20c parts, we need to stock a $40 board.

But to make the model work.. the boards must be common. No one wants to inventory dozens of various boards for the 'what if...'. The sunk capital cost is too high.

Unless Stern were to help operators by making spares inventory available cheaper and readily... its going to result in the WoZ situation instead of the goal of 'less downtime'.

The AC/DC situation you refer to is a poor example because its not really production stuff. Once the platform is 'live' and in production, that's the real measure of support.

To me the big question is still 'common interchangable parts'. How are they going to keep the board count low that operators/owners will need to stock to keep the benefit of a FRU vs component repair viable.

At Modern I have heard more than my share of customer complaints about AC/DC being down for more than two months and its all due to one board being out and it is on back order from STERN think about that situation being repeated over and over with a Spike system with this sort of design.
There must be backup systems in place in Spike for a machine to work at least 90% with one failed board by enabling circuits to be rerouted with a jumper or through coin door adjustments etc.
A machine can't be down waiting for a board because (for example) an LED or whatever failed.
Just my thoughts and Im really happy that Stern is trying to do something with a product that is still not mainstream….yet

If it talks... there's a vector. If it has a network stack connected to the net... there's a vector. Hijacking outbound connections and fooling the system into loading hacked firmware is a huge vector depending on how code signing is validated before execution.

Obviously reducing surface area is step one... but don't kid yourself that means no exposure.

No you don't - your footprint on the net and the fact you are talking out over the wire gives someone the chance to redirect that traffic mid-stream to a malicious target that can then spoof your intended target pushing content back to you through your 'trusted' channel.

even if there was nothing 'of value' retained locally... zombies are not welcome.. and then can become their own distribution point of other stuff.

Your dismissive attitude is very dated.

Of course there is always the "who will bother?" angle.. but don't fool yourself that no listeners = unbeatable. The network stack itself is still the broadside of the barn you can't hide and you are active on the net.. even if only outbound.

I don't understand why people are comparing with JJP so much. The JJP system seems to favor a Hardware Abstraction theory for it's platform. Write to a software abstraction layer that is widely supported and has a very positive future ahead of it.. (linux). The hardware BELOW the platform can be modernized, changed out, etc with minimal impact to the software running on the platform. That's where Alex keeps going on about the 'swap it' future proofing.

SPIKE on the other hand looks to be another hardware specific embedded system platform that has chosen to be distributed and extendable through modules vs a monolithic board system. It's not clear how much Stern has focused on abstraction vs simply taking linux and other elements for their simple time to market needs. So much of this JJP vs Stern stuff just seems to be talking apples and oranges.

And people are reading too much into the PR guy stuff. The tips to the 'proven auto' tech is talking about the bus technology... not necessarily that the boards or components are taken from proven designs. We all have to remember we are getting our info from marketing/PR people... not the hardware engineer

What remains to be seen is just how Stern is going to interact with it's distributors and operators to leverage this FRU strategy (field replacable unit). Are they going to help people stock components or force people to buy everything up front? Are they going to offer improved support like timely advance replace, etc? We have a new product, but are they going to alter how the field OPERATES?

Just how re-usable are designs going to be from game to game? Are we going from a handful of game specific boards that were field repairable to a dozen game-specific boards per release that need to be stocked or advanced replaced, etc?

It was mentioned by stern in terms of future growth/possibilities the platform can enable

This is a classic example of using extreme examples to justify throwing out the 99%.

If they forget about calling you that one time... the other 300+ days were fine and you'll notice the game is no longer reporting.. which means on your next service call you address it. It's not like you never visit your locations for collections or service calls... :rolleyes:

And your 'what ifs' are the same for things far more important like PAYMENT SYSTEMS.. and that hasn't held them back.

Resetting to default is a non-issue since your box would keep doing the same thing where it was exploited prior. It's simply groundhog day next time the unit cycles and falls into the same trap.

Listen... I agree that the game isn't a high value target... but don't kid yourself as a designer that your system is fool proof or impregnable. That's the kind of mindset that makes you believe you are invincible... right up to the point where you got killed </topgun>

Companies like HP used to think the same way too "no one cares about hacking your printer"
And people would say "who cares about us.. we're just a tiny HVAC contractor..." (many steps later.. Target goes down)

Security through obscurity is not a defense strategy.. nor does it reduce potential. It's just a factor in probability.

I've been part of many embedded platforms for a long time where we've said "no one will care about this little device..." and it breeds complacency. Then when your product changes audiences, and gets all kinds of new scrutiny from new angles... many times people have to make an abrupt change of attitude once they move into the big leagues where people are actually motivated.

There is a huge difference between probability and possibility.. just because something is unlikely, doesn't mean it's impossible. You just need to balance how you prioritize things and manage risks. But to go around boasting you're invincible... well that's just egotistical and inviting scrutiny. And it's not just about egg on your face when you are talking about your business and your customer's businesses.

No, you see the bottom of one board that has through components. Why are you leaping to the conclusion that what you are seeing is driving transistors and they are all that way? What about the other components normally in that path?

Note from the photos the flippers seem to be driven from that board up near the apron area.. which could be entirely different.

I think your conclusions may be premature.

The board in the center you can only see connecting to LEDs.. and no 3 hole through holes I can see.
The driver board under the apron is clearly for coils... and you can make out at least three tabbed transistors.. and some caps but that's about it. What the rest of the system, the driver circuits are setup as, etc is all still vague.

This is just a software platform/framework type of question. For ease of coding and since the system would be reused so often.. they would have likely built up a platform that abstracted all that hardware interaction from the game code. Instead of the game code worrying about 'rowA,colB', it just knows 'sw19' and somewhere there is a mapping table for the friendly labels to the hardware positions.

How much abstraction there is or not is entirely up to them in their software development. The more there is, the easier it would be to port across hardware revisions.

There can be constraints of course... but this is more about choices.

And the trade-off is duplication. Islands of functionality need to be duplicated where needed instead of sharing a common output.

I think the real gain is in smaller, simpler, less loaded PCBs that are modular for serviceability and scaling. Modular vs monolithic.

There should be some savings from cutting alot of the labor of building up and installing the harness... which of course is balanced by having to make and install many more smaller connector cables. The harness may have been complex.. but it was well understood and cut down on parts to stock, build, etc.

In theory games should be easier to design as there is less engineering needed to 'make it all fit' as well.

People keep bringing up the wiring harness... I think that's a bit of fools gold. It seems archaic, but was built out that way for good reason. The real gains are elsewhere and not necessarily in cost of materials.

If you only use one board and replace the harness with all "home run" cable runs... where is your cable savings? You're adding cable because you are not daisy chaining. You just save some depending on where you install the board vs the old head location.

I'm talking about the wire harness and all its functions.. not just the GI circuit.

Again, the automotive tech comment was almost certainly about the *architecture*... not the implementation.

And the DESIGN (not just the single implementation you may have seen) is to be modular. There are tradeoffs with that, but you prioritize what you are after. Stern is trying to address servicability and (presumably) scaling. Instead of having separate designs every time a game wants more more drivers the system simply scales.

I think this is a bit of revisionist view. The boards and components have been moved around many times by different manufacturers with different levels of success. Modularizing the boards is not some breakthrough that wasn't possible before.. it's been possible essentially from the start of the SS era. Many many times aux boards have been put on the playfield and elsewhere in both old, and newer designs. The benefit of the previous design was consolidation and reuse. Don't have multiple circuits doing the same thing, consolidate needs into shared resources, etc. The components had dedicated space... you didn't have to play for them in your playfield design, etc.

Returning to the wireharness - The benefit of the wiring harness was avoiding 'home runs' and reducing materials by reusing common paths and common circuits. The trade-offs are complexity in design, more complex part to assemble/install, and the service/integrity issues that come with sharing common paths/circuits.

People keep referring to the shorter wires as some manufacturing god send. Obviously there is less wire involved when you are not running all the way back to the head... but that is also offset if you remove the common hot or ground paths for everything.. and having many more connectors. You have lots of small harnesses instead of larger monolithic ones.

It's a trade-off, not a straight up elimination. Gaining simplicity/serviceability at the expense of other things.

memory is so short... people already seem to forget the uncertainty thrown out when MMr showed up with a huge under playfield board. Concerns rooted in past attempts to put electronics in the main cabinet.

I hope stern puts out some more technical data.. and we don't have to wait until people start getting games in hand to look at them.

Because I love my lotr over my tz because it weighs less. Roll eyes....

Weight is nice but not something to swoon over

Ack stupid autocorrect hates game acronyms

what kind of place was selling them out of curiosity?