in 10 years you will be able to buy a 100% compatible board/mem/cpu combo for WoZ. it was built MODULAR. It doesn't rely on some board house to create a CUSTOM part for a small niche market. Parts are readily available at any point. And even the current linux OS in 10 years will work as WoZ is 64bit compiled with abi compatibility enabled. Welcome to the world of STANDARDS. Using standards allows as much flexibility as you can imagine.

So yes in 10/20/30 years a standard Motherboard/CPU combo will absolutely function perfectly. It will also take less than 10 mins to swap. A few connectors and 4 screws and the board comes out easily replaced by a new one. Welcome to the world of proper design and enterprise engineering. Where people see the forest through the trees.

BTW

$50 dollars for a harddrive failure? Where are you buying your parts?

http://www.amazon.com/NexgenCover-128GB-Internal-Drive-Solid/dp/B00IXCFP9I/ref=sr_1_3

P.S. an SD card will fail with 10x the regularity of an SSD due to no wear leveling. So your $5 dollar SSD will actually cost you $50 over the same length of time a Single SSD would cost you at currently 1/4th the price. So you're analysis is completely wrong.

Nice try though.

CAN'T, or wasn't convenient for you to do with your available tools that were designed to do through-hole soldering? There's a huge difference. People are going to have to play catch up and learn to work with SMT boards, and get the appropriate tools. It's inevitable. I know you're going to be of the opinion that ops and techs don't want to spend another 500 bucks on yet more tools just for one machine, but this is the present reality of technology. Through-hole has been on life support for YEARS.

SMT components are cheaper, more reliable, easier for manufacturers to use..it's not going away. I would place good bets on being able to find the appropriate SMT LED replacement in quantity for around 25 cents a pop. I found 100 5050 RGB @ 39.6 cents a piece from a hobby supplier. I would expect cheaper for non-rgb (doesn't seem like Stern made the full switch yet... /sigh) and for more bulk. The repair side in 5 years is going to be a non-issue once people actually develop the skills and start having the common parts on hand. Right now it's a mess because most of us DON'T have either, because we're stuck in La-la-land on old tech and haven't been forced to adapt.

The real question, like you pointed out, is "Does this system have a way around the serial LED issue?" I think a physical jumper pin would be awesome if they could implement that to allow the data to pass by the crapped led while making the system think that it's still functional, or have a way to disable it in software.

Just to be clear, a burned out bulb should not break the chain of the data. For a chain to break down you've fried the microprocessor on the board that processes the data and operates the LED. And now that microprocessor is either grounding out the bus or they've got some crazy protocol where each one has to forward the data or acknowledge the packet. The first is an interesting failure mode and the second you did to yourself by picking either the wrong bus or using the bus incorrectly. This happens a lot when the electronics guy picks the communications bus (prob cause its cheaper) without caring if its actually appropriate for the job at hand (ive had this happen to me on more than one occasion).

Choices:
CAN - not bad here..common to optically isolate the nodes from bus (fixes the short thing) and is very master/slave. And good for periodic messaging not for guaranteed protocols with lots of overhead. CAN is a little awkward to work with though. Weird standard protocols for CANopen or I've had to use CANkingdom.

Rs-485.. Also fine from a not forwarding but the chain is limited to 32 nodes I think and you should really be using the spec resistance cabling if you are going that high otherwisr you get weirdness (not CAT5) I don't think those are typically opto isolated, but you can do it.

ModBus too but I'm unfamiliar. Ethernet prob too expensive in hardware if you're doing some high level stack (IP based) plus unless you use a switch then this actually could be more prone to weird forward UDP packet situations where one guy goes out and it starts fing things up.

LOLZ.. good luck with those x brand SSDs.

They would work fine considering we don't do many writes. And at that price point they can be throw away. Even if they last a year or two.

I was stating an observation in general.

Yes SMS tech is the future, that is a given, and both techs and hobbiests alike will have to retool their shops.

It is also a double edged sword as it is cheaper and quicker to incorporate for the MFG. But as for reliability it has not been really tested in the Pinball environment so I wouldn't claim that it is just as or even more reliable. Is the Spike system using Mil spec components? Is it more efficient yes but that is not the same as reliability.

Why would the simple changing of a socketed LED now have to entail dragging out soldering equipment in order to change a dead one, especially in a crowded and mostly dark location environment. Ever try reworking SMS on location?
Again what looks good on paper is not what works in the real world.

Oh I see, I'm so dumb, I'm to replace the bad board with a new one by again bringing out the soldering/de-soldering equipment for the daisy chained small led boards ( they never use connectors from what I have seen) to remove the bad one and replace it with a good one. Then I can repair the old one in a place with a proper set up. For other applications sure for Leds The whole idea is plain stupid.

Where is this advancing technology? This is a huge mistake for sms applications.

In this scenario with LEDs in a Pinball environment socketed LEDs are what is High Tech

What I just wrote is scoffed at in the paper world where LEDs never fail.

The best High Tech is the correct tech needed for certain applications.

As someone whose job is to create fault-tolerant server systems, I'm a bit taken back at the idea that they've created a fully serial network, or at least not provided failover paths.

From my perspective, having one more wire for bypass seems like a no-brainer, but perhaps I'm missing something. Cost reasons?

Good point here. If you'd have asked me (or most of my current co-workers and friends presently) 3-4 years ago if anything on a circuit board was replaceable at all you'd hear a resounding 'no'. Now I can fix anything thru hole because I have the tools, knowledge, and confidence.

There are obviously limits to what is feasible with SMT as many IC's and processor contacts are far too small for hand-work, but things like resistors, transistors, relays etc... are still very repairable in my experience.

double post

It's been tested in automotive control systems for 20+ years, to the best of my knowledge. How many times have you had to pull your car computer out and reflow joints? I tried to do it once, it was laughable and didn't work, it was just bad components. Point is, I'm pretty sure the computer in my dashboard of my Civic is exposed to way more vibration, electrical noise, and temperature shock than any pinball board ever will be.

You're right about trying to do SMT in the middle of an arcade. It's not very feasible, short of having a spare board then taking the other one to the "shop" and fixing it. Were the LED issues on WOZ the fault of the LED components, or the volatge source/controller? Wondering the same thing about your AC/DC experience, but I would bet Stern isn't going to tell you, are they? =\

Clearly, these things have some kinks to be worked out, but ultimately, it's really hard to see it NOT being more reliable in the long run.

LOL yeah if there is one thing you never had to deal with in DE/Sega/Sterns it was a buzzing sound from the speakers eh?!! haha!

Pinchroma, please don't skip your answer this post. Please inform me to what it is so I can fix it.

if they really wanted to make the boards bullet proof and not serviceable at the same time theyd conformal coat them. That's what they use on military spec'd boards for high vibration (helicopters are way worse than pinball)

I would have to disagree with this analogy(car/pinball).Cars are rarely on for 12-16 hrs a day.Thats how long they are on at Modern Pinball.If you were to drive or just turn your car on for that many hrs per day how do younthink it would perform.It would have failures just like pinball machines.

Overall I think component level repair is almost obsolete.Board swaps are all the rage in pro audio.If you have a relationship with the manufacturer they will Fed ex a new board while they wait for the old one so you can be up and running quicker.

Stern needs to stock replacement boards for their customers.Their really shouldn't be any back ordered BS from the manufacturer.

Hard to say what it is? I don't have a buzz on my game so it could be a number of things. Does the buzz come directly from the speakers? Does it occur at a specific time? I can tell you its definitely not fan related which was my post above because the fans run at a static speed so they aren't an issue.

I delivered pizza full time for close to 6 years... never had a single computer-related failure with that car....my current car is pulling up on 200k miles, no computer issues..

It's not just the time in spans, it's cumulative. SMT *is* more reliable not just in theory, but in actual practice, than through-hole. The issues we are seeing with pinball-related SMT stuff isn't due to the nature of SMT, the issue is with the control systems that are running these components, like the drivers for the LEDs. SMT has nothing to do with the serial bus interface issues either. These things need to be address with ops, like Modern, to figure out how to get this symphony to play nice together, not to sound like a fourth grade class. We're not there yet.

There is a fairly loud buzz/whine that is always present, but you can hear it ramp up in pitch to the lamps in attract mode. Reminds me of a really bad grounds on system 11 machines only worse.
To me it sounds like the fan on the power supply supply, slightly changing speed as the load on the power supply is changed.

I agree it sounds like there is something not properly grounded or something normally shielded is sitting on top of something HV. Can you verify a few things.

1) The cover is on the chassis. I've seen people running with it open.

2) Verify the audio amp connections are properly seated both amp side and speaker side.

Only thing i can think about off the top of my head.

I'll take a look, but the game has had this since day 1 so I assumed that is how they all are.

Actually it was completely redesigned for The Hobbit and should alleviate all of your concerns.

What I should have said was WOZ was a good example of a new system that was good on paper.

As you just confirmed by my point until a system is used real world its perfect.

In this case JJP learned from real world use and has redesigned its system.

I never had any concerns that JJP would accomplish this.

My one concern is that the issue for operators and location techs will not be able to do a simple quick

blown LED change on new machines.

Correct therefore the drivers should be socketed

Comes from the speakers. It's loud enough to hear from anywhere in a quiet room. It varies in pitch and loudness in perfect sync with the LED'S in attract node.

All of this debate is kind of fun but also kind of silly. There are so many good and bad points to make for one system or another but to be worried about what's going to be on hand in 10 - 20 years borders on silly.

I work in IT, I'm a pinball hobbyist who does a shit ton of through hole and SMT soldering, and I've seen a ton of systems and been through all of the pro's and con's. Not one system is hands down better than another one. They've all got their pros and cons.

We are virtualizing like crazy in IT. Desktops are all now little Wyse boxes that have no hard drives, etc. Tablets are replacing PCs at home and with all of this cloud stuff going on it's not unreasonable to think that in 10 years PC motherboards will not exist as they do today.

True, there are servers, but a server motherboard does not cost $100 and does not have an ATX form factor that you can't just screw in to your PC based system and have working in 15 minutes.

But then again there are billions of current PC motherboards out there today, so it's not unreasonable to think that, even if PC mobos are designed into extinction, you'll be able to locate replacement mobo's for your PC based pinball machine for a few bucks off from some scrap yard.

I have Cisco equipment sitting in the harshest environment I can imagine, an Iron foundry. There is the craziest conductive dust floating around, temperatures swing from 20f to 120f. And we're not regularly replacing switches because their RJ-45 connections go bad.

System 11 is a proprietary boards based system. It's true that System 11 boards go bad and need to be replaced. But the ones that go bad, what went wrong? Some short inside the cabinet blew a transistor. The battery leaked, 20 year old solder joints cracked after being in and out of warehouses with no temperature control. If you have a well maintained System 11 game that wasn't exposed to too much routing hell, the original boards are probably still working. These things take a beating and by and large they work fine until you really do something terrible to them.

If some mook connects a coil to the switch matrix on a WoZ, I'm sure something's getting replaced that you can't get at the computer store.

Light sockets corrode. Ever have to dick around with a 44 LED bulb to get it actually lock into a socket? How about twisting the top right off that new bulb trying to get it to stay in that socket? I have.

SMT LED's do fail. On location you'll probably want to just swap the board. Is it harder than just replacing a 44 or 555? Maybe. It would be easier if the boards had connectors but Molex connectors are expensive, soldering two wires to a board is cheaper. Replacing a board that has a molex connector is easier - unless you accidentally pull a wire out of the connector because of a weak crimp. A 44 is easiest unless the little nib on the bottom of the bulb is too big and you have to file it down to get it to fit, or it's in a weird place and you have to unscrew the socket, or you bend the socket back and accidentally short it out. So they use 555 sockets instead which can easily pop out and are protected from shorts, but those things are always popping loose from vibration and sometimes the leads on the bulb need to be bent in order to get them to make contact.. and on and on and on....

There is no pinball system panacea. Incremental improvements, slowly making things more reliable over time. Some guys will willingly change to whatever is new and cool and be OK with suffering the pains of change. Other guys will sit back and use what they know to be solid [enough] and reliable [enough] for as long as they can.

I'd argue that anyone who makes light of engineering changes has never been through one. Anyone who thinks their engineered solution is absolutely the only way to get something done is closed minded and foolish.

It's good to see Stern moving on from SAM. It may be overdue, but I totally get why it took so long. Why didn't they use a PC? Because that change is huge. SPIKE is an incremental change based on what they know. It's probably the best decision given all of their parameters.

In 20 years, if the world doesn't go to hell, someone's going to be making money building old parts for all of these systems because they are no longer available off the shelf, no matter what they are.

No definitely not the norm. Can you quickly open a ticket and i'll point you in the right direction. That's not really my forte but i know who can assist.

In the end it's "which pro's does a designer value the most and which pains in the ass are they most willing to deal with". No system is without both. Anyone who says different is trying to sell you something.

DK ….Great post!!!!!!!!!!

Your point is moot as an SSD does not (never?) fail due to wear but to numerous other electronic failures.
Where I work we have thousands of SSDs in use and so far we have not seen ONE to wear out.
And in a pinball environment, I can't see any reason for wear problems even on cheap/small SD cards.

Wait does spike stand for Stern Pinball Inc. Knows Everything ????

Exactly. Out of thousands of SSD's we have had I believe 1 fail? Maybe 2? And thats an acceptable rate of factory defect. .02% failure rate? Sounds good to me.

When analyzing the SPIKE acronym, don't discard extreme possibilities. Don't forget that SPI is descended directly from the folks that brought us the TYFFASI flipper control board.

The game I'm familiar with also does this. The buzzing is loudest when the LEDs are all white. If you enter the menu system when the LEDs are white it is loudest. If you cycle through the colors, it is loudest when the LEDs are white.

https://pinside.com/pinball/forum/topic/pinball-technology-brief

It's like the FTS-8. http://www.dkpinball.com/DKWP/?page_id=396

And no, the F does not stand for Forget.

Fun fact, I am legally compelled to thumbs up ANY MST3k reference.

" SPIKETM is a modern, scalable and modular electronics hardware system engineered for the rigors and complexities of pinball applications."

This sentences uses a lot of words to say very little. Could have conveyed the same information with the sentence, "SPIKETM is new."

Just sayin.....

One of the best posts I have read in a while

Flash drives do wear out. It might not have happened to you yet but we're talking 10, 15, 20 years here and deciding what is acceptable to whoever it is deciding what it is that they like.

Sources:
http://searchsolidstatestorage.techtarget.com/definition/NAND-flash-wear-out

"EEPROM and flash memory media have individually erasable segments, each of which can be put through a limited number of erase cycles before becoming unreliable. This is usually around 3,000/5,000 cycles" - http://en.wikipedia.org/wiki/Wear_leveling

I know that NAND will potentially wear out, but in practice it does not.
The chance for an SolidState drive to fail due to wear is like 1:infinite, compared to all the other stuff that can and will break an SSD.
I've never seen or heard of an SSD actually wear out, except the ones in the endurance test on techreport.com

i hope they at least make it so you don't lose your settings when you replace the batteries (i.e., not really need the battery for anything other than keeping time).

In this case I'm sure the only time the card is written is when you update the firmware. Unless they're storing audits there. Hmm.

If they add network support, lots of cool possibilities with this besides network tournaments. How about remote audits and diagnostics? You could even have the pin e-mail the owner and support when something fails. Support could already have a part on the way before you know it.

seems like they would need a staff to monitor this though. couple thousand machines sending emails for each diagnostic could be cumbersome after awhile, and the initial implementation would definitely have bugs. I love the idea though. Would definitely help them assess what replacement parts they currently needed on hand.

Could be huge for operators. Logging game times and duration, even optimal settings based on recorded games. Lots of possibilities.

I'm not sure if you're assuming it's Stern. It can be anyone really, just the operator would be enough. This is standard on IT infrastructure equipment. I have servers, ups's, a/c units, and network equipment e-mail me all the time with health problems. Really I only get notified when there is an issue, so it's only a few e-mails a day.

I'd love it if I was an operator and I got e-mails telling me that a ball is lost or a switch isn't registering, or some other problem.

For one, you might be able to skip stops checking on machines and use that time to work on machines that are having problems.

Even better would be if you could poll all of the machines. Just responding and saying all is well saves a trip.

This actually opens the door to being able to accept credit cards. Imagine a machine on location that you didn't have to visit except to clean it every few months.

_{Post edited by dkpinball}

This already exists in our platform. We haven't exposed those functions yet but it's coming.

If there's one argument for using a PC at the heart of a solution it is having a lot of flexibility to add features.

... although with flexibility comes complexity and fail points... always yin and yang.

Is this discussing SPIKE, or JJPs system?
Leave the agenda at home Pinchroma

Yea, the thread kind of devolved into talk about surface mount vs. through hole and PC vs. Embedded.

Usually that happens when there's no real news. Does anyone have better pictures yet or actual specs and facts? Those would be cool right about now.