I cannot speak every homeowner, but the 25+ technicians I work with regularly or converse, prefer not to fix Stern node boards.
This does mean they do not possess the skill, education, or proper equipment to do so.
Private techs are becoming more and more rare for many reasons, outside the scope of this discussion.
We certainly repair when we have no other options, especially when SPIKE boards are NLA, which has already happened.
I don't know many private collectors with hot air stations / microscopes and training, but I do know quite a few that pay others to maintain their games, averaging $75-120 / hour in labor which is realistic pricing.

SPIKE schematics and documentation remains insufficient in comparison to any previous pinball operating system, and electronic diagnostics remain practically non-existent. Essentially we have to reverse engineer Stern board designs to determine the shortfalls in design in order to replace SMT components, if possible. Direct suggestions to Stern for board improvements from SPIKE I / II where ignored in favor of reduced costs. All of this directly infers lack of interest for the design to encourage repairs. I do not know how to quantify this concept with statistics. Throwing numbers at the problem will not convince anyone, so this seems wasteful.

BTW, SPIKE I / II are not fully backwards compatible with changes that were not fully documented, and made intentional to promote obsolescence, not significant improvements. I am not speculating, but rather comparing the systems benefits and major features. Nothing wrong with technology advancements, but SPIKE II is not a step forward in integrated SMT.

What has held up a potential landslide of otherwise large number of failure complaints is the positive support of Stern customer service.
The technology overall is not new (SMD was opted by many industries in the late 80s), and the Stern SPIKE system is still in its half life stage presently.

I recognize the changes in today's pinball market and collectors may not care, and that is not in my realm of control.

This is where your argument takes a hard turn south. Stern is not known for its engineering prowess and extensive hardware testing pre-release.

Not really. As you, I don't have access to Stern's SMD derating guidelines, so no reason to speculate. Anyone could reverse engineer if they were interested, but I haven't seen that happen.

We both don't have any idea to the margins they design too, but I would suspect it's within the manufacture's guidelines for warranty concerns.

Don't get me wrong, my comments were not to defend Stern, it was in response to your SMD statement (not SPIKE 2 in general). But my engineering side of my brain says "Show me". I'll be the first to rake them over the coals if it's proved they are doing something incorrectly.

Can i send you an invoice for my iron maiden led board that had the smd led break itself off the board? It sucked buying a board instead of dealing with a socket or mount.

Only if I can send you invoices for all the thru hole parts that end up in the bottom of my cabinets.

Stern uses Cree LEDs on their boards (according to their schematics). Assuming they use Cree throughout, then the RGB LED on the board was likely something similar to this:

$0.57 at Digi-Key

pasted_image (resized).png

pasted_image (resized).png

Yes, and it broke off one side from it's pads. It was on the finger tip portion of the LED board that floats and doesn't have much support. It broke (and took the pad with it) after about 2 years old. I wonder why....

Dunno. I've dealt with a lot of boards and SMD parts pulling away is not that uncommon. Probably at least as common as cracked throughhole solder, but a whole lot more of a pain to take care of. Can't say whether thermal issues or vibration is the cause, but it's not uncommon in video game and pinball stuff. Auto and aerospace probably hold to a higher spec.

Yeah, guys - didn't mean to imply SMD is the Holy Grail (well, maybe some ) or is infallible. But with few exceptions, I'd still prefer it over thru hole. Definitely helps w/RF designs - I couldn't do my designs w/o it. I do think a lot of people are uncomfortable with it when soldering, but for most part it's not a big deal working on (BGA's being one big exception). I honestly don't think I've ever had a SMD part come off one of my designs, but our product goes thru a lot of vibe testing.

Vire - That almost sounds like a layout issue, not so much a SMD issue. SMD manufactures usually give recommended pad sizes for their parts. Perhaps the drafting dept cut some corners on that product? Your right about thermals - meaning if the expansion coefficient of the PWB and part are out of whack, that can be an issue.

Flynn - OK, duly noted. They could have made the pads bigger, used a dab of silicone, or mounted it better. Could even be the type of PWB material that is used (some boards are stiffer than others or the copper bonds better on some). Was the board loose?

The only gripe I really have with the newer boards is if the software is proprietary, making aftermarket boards a lot less likely.

Think I'd rather have socketed or twist mounted LED's for my ease of use also, but I understand why they don't do it. At least with the whole RGB aspect and programming. Looking at the bright side at least its not Apple or Tesla, Gomez is trying to give people the tools for it. (at least they started)

I am getting my first Stern spike game (Jp).
I repair pinball machines for 42 years.
Now I have to get used to the fact that I can't repair them myself anymore.
Never seen a spike game inside.
But when I read this thread.
3 thing we need.
One, more standard (backwards compatible) node boards so we can swap boards to check before order a new one.
Second, better manuals and test applications.
Third, lower prices for the spare node boards because they are ridiculous expensive.
Let's hope I won't regret it.

Most problems you will still be able to handle yourself with your experience.

Stern ships this with every machine. Node 8 and 9 are the same board with different addresses. You can change the address via dip switches and swap the boards. If the problem moves, you know the board is bad.

100%. There's NO EXCUSE for better diagnostic messages and tools on the system menu not being implemented this many years in.

That ain't happening, but many issues can be repaired, and pinsiders like Borygard can repair many node board problems for reasonable prices.

Congrats!

There are limited Spike and Spike 2 pins schematics along with a Spike manual that discusses the overall operation here: https://sternpinball.com/support/

What I don’t see being discussed is that Surface mount and boards being mounted all over the playfield introduces a lot of mechanical vibrations that the board may not be engineered to the tolerance levels needed to withstand this. Additionally, having the necessary equipment at home to to surface mount repairs may not be practical for an inter Merida test DIY person at home.

These machines are designed and marketed predominantly for home use, which means it won’t get commercial level play and the average home use owner will probably be ok for the most part. Drop them in a barcade? Good luck!

So, for Stern minor issues, you have to send out for repairs? Can’t DIY fix? That sucks....

That is literally one of the subjects of this discussion.

What is the failure rate? Things break, and for various reasons. If you’re claiming the failure rate is excessive and/or it’s due to an engineering design flaw please share your data. Was a more robust design considered and rejected due to other considerations? We don’t know.

What do you mean by “predominantly”? Please share your data supporting that claim.

I was going to mention this in a post, but I was afraid the thread would go down down a rabbit hole on comments. Some people mentioned installing grommets. On the surface, that might sound like the right thing to do. But the issue you would want to avoid is making a resonance (i.e. the frequency of where things want to naturally vibrate and fail apart) occur at a worse frequency. Adding a grommet would likely lower the resonant frequency and then the question becomes "Does that make the issue worse or better?" And to complicate things more, what axis is the culprit? I don't have an answer for that, so that's why I didn't feel like speculating. That's better answered by a ME and some expensive software. We had a huge vibe table at work and if I'm not mistaken it could vibrate in all the different directions and you could also change the frequency and amplitude. Louder than all get out too when it was running.

Flynn mentioned earlier he had a RGB LED pop off the end of the board that probably vibrated. So would the fix be perhaps lowering the frequency with grommets, or making the board stiffer (raising the resonance)? Don't know.

Terrible. All manufacturers should use standard controller and computer stuff.

Looking for help with an issue if any of you have ideas. My Deadpool randomly freezes at different stages. It could just be sitting there in attract mode and the screen will freeze, lights will shutdown, and you can no longer play the game. It happened yesterday while I was adjusting the settings. Sometimes it occurs during game play. The odd thing to note is that I could play for 2-3 days without issue and then it'll randomly occur 9-10 times in one day. I've checked every connection and replaced the SD card twice. I'm looking for suggestions before I replace the board. I'd hate to replace the CPU node board and have it not be the issue.

Reseat the SD card. Did that I see. I had a processor board go out on SW with the same symptoms.

Yep, did that about 10 times. I'm thinking it is the processor as well. Just looking for ideas before we spend money on it.

I think the main issue with Spike is not whether surface mount or through-hole components are used - with proper tools both are repairable. What looks like a problem, is Stern ignoring testing the effect of vibration on playfield mounted boards. We have already seen a diode and a capacitor shaking itself off the board. Mounting the board with rubber grommets will not necessarily be a fix, since that may just cause the board to vibrate at a different frequency.

Heavy or long leaded parts should be secured to the board, and all connectors should be vibration resistant. Maybe that would also cure the problems with SD cards getting loose and misbehaving.

You can pretty easily still DIY fix the Node board coil transistors, etc, bit most people don't have the equipment to replace surface mount parts. That's the main problem (aside from almost complete lack of diagnostics).

REPLACE the SD card with a good quality one. Stern uses junk cards.

Sounds about right from what people post.
What brand do you recommend ?

Go to Stern's site and look up the Spike manual (under Support somewhere). There's LED's on that board for the power supply's status. Check to make sure those are all lit according to the manual when the issue crops up again.

These are the ones I stocked up on this past BF for $9.99:

https://www.bestbuy.com/site/sandisk-extreme-plus-32gb-sdhc-uhs-i-memory-card/3990822.p?skuId=3990822

Even at $16, not too bad. But Sandisk is one of the most counterfeited brands in the world, so do not buy it from Amazon or Ebay, only a brick and mortar (based) store that's part of the manufacturer supply chain to ensure you don't get a fake.

....Thats with thousands of games made, possibly 100's of thousands of light boards, seems like an OK fail rate.
I will chime in the update from a couple years ago,... NO new failures, reliable as hell really. LESS problems than I've had with most games in the past.
Stern must be upping the game on playfields if we are back around to bitching about SPIKE.
Where is @dr. frightener to tell us all we have worthless machines within 10 years of production?

Yep, did that last week and they are all showing fine. At first I thought it was power unit. I'll do it again just to be sure. Thanks.

Don't mean to derail the thread more - but you did this when it was messing up?

Yep, right after it locked up. Sorry to derail the thread. Figured someone in this group might know. I appreciate the help.

could be the power supply, could be the processor, could be a connection, could be an the SD card, could be something else.

I replaced the Power supply on TNA, because it kept freezing and resetting over and over (sometimes it would go days without issues). I checked the voltages, it was 5.25 one time and went to 4.80 another... not sure why it was fluctuating so bad, but it made it hard to pinpoint. Replaced it and haven't had an issue since.
Spooky and stern both use power supplies from "mean well", I went with the next size up. Cost me $40 bucks from mouser.

duece

The little AUX power supply, right? Because the main TNA power supply is one of the more expensive ones meanwell sells - hundreds of dollars.

Interesting. Might be worth a try. Thanks!

A new $15 SD card, reimaged is your cheapest entry point and most-likely culprit, given Stern's history. An RSP-500-48 replacement power supply for Spike is like $90+shipping.

I've tried two different SD cards outside of the original card. I had issues with both of them. That's a cheaper replacement than I expected.

Were they both class 10? What brand?

The first SanDisk was not. It was around $20. I believe the one in there now is PNY? I would have to check. If you have a particular brand to recommend I could try it again for sure.

Class 10 is important. If they were not, then they will not work reliably (or maybe at all?).

These are the ones I stocked up on at BF for $9.99:
https://www.bestbuy.com/site/sandisk-extreme-plus-32gb-sdhc-uhs-i-memory-card/3990822.p?skuId=3990822

I haven't checked the stern model on mouser, but the one TNA uses is only like 26 bucks. I went one size up which cost 34ish plus 10 bucks to ship it. I'm assuming if you got it from the pinball manufacturer it'd be more expensive.

I'm quoting mouser/digikey/jameco numbers. The small TNA aux power supply is cheap. The main power supply in Spooky machines is a SE-600-48 is not. It's about $80-$90. In the last post, I was thinking about the American Pinball power supply which was close to $200 before they changed it for a no-name cheap one.

Spike power supply RSP-500-48 is almost a hundred bucks.

Many people call repair guys for small issues. They'll have to pay for node boards and the repair fee. Your speculation (what you call a guarantee) about him ripping people off is far worse logic than any of his. Gomez admits they cheaped-out on the PSU. That is a fact. Spike is overpriced and is difficult to diagnose and repair. Facts.

I can and will dispute your logic about costly system 11 board repairs as that's not what he's arguing. He's arguing that small faults (as small as a fucking LED failure) involve node replacement.

He's not an old Luddite. He's a guy pointing out cost cutting and shit quality. If you can't see it, that's your failing, not his.

I don't think he said that, and the 500W Meanwell in the Spike2 is actually pretty robust (the 300W Meanwell Stern abandoned is another story). I believe he just admitted that it had an annoying power supply fan that was too noisy. That's all.

My fear is that by finding another PSU vendor (which is unnecessary as Meanwell has said they would use a different fan if Stern requested it and they committed to order X quantity), Stern will use the opportunity to change to a no-name Chinese vendor with a much crappier power supply like American Pinball did. That would be terrible.

Time will tell.Spike games will and are failing.
Buyer confidence is starting to dwindle.

The Spike system is rubbish

Cheap and nasty.

The replacement node boards are over priced to give the consumer a false sence of value to cheap game they have purchased.

I have owned numerous Spike games and never had a problem. I believe this is being blown out of proportion for no quantitative reason. I will continue to enjoy the best era of pinball and keep buying Spike. Back to your regularly scheduled whine.

The SE line from Meanwell is one of their economy lines.

The RSP line is a decent upgrade from the SE.

Given that Spike 2 began in 2016 and the UHS-III standard wasn’t released until 2017 I think it’s unlikely a class 10 card is required or beneficial. The UHS-II maximum speed was only 312 MB/s, meaning a class 4 card would be fine.
https://en.m.wikipedia.org/wiki/SD_card

Do we know what SD host controller the spike 2 boards use? Are they all the same or have the controllers changed as the tech has evolved?