Is this the best thread to share node board troubleshooting experiences?

I recently tried to play my IMDN Pro, and found that Node 9 was not found.
The 6V LED was not lit, so I took the board out of the game and measured a short from the 6V rail to GND.
Since this could be any component that sits between the 6V and GND rails, I started removing components one at a time, trying to hit the most likely ones first.
The D9 diode was shorted after I removed it, but there was still low impedance on the board. It did have the zip tie installed per SB189, and the legs were still intact.
I removed a few capacitors, but mostly because they were easy. I thought maybe C28, the electrolytic, could have failed (it hadn't), but also removed a couple of ceramic caps before resolving to remove U9. Since it has a thermal ground pad, it's more work to remove.
After removing U9, the short on the board was gone, so both D9 and U9 were confirmed dead. Neither showed physical evidence of damage like I've seen in other pictures here.
I connected a bench supply to inject 6V, and the board sprang to life with the 6V lit and status LED blinking.
Replacement parts for all the stuff I removed are on order.

20220906_181947 (resized).jpg

So that is probably the catch diode in a buck SMPS regulator. Your probably lucky it failed short instead of open, the output voltage likely to spike up to the input level roasting stuff with over voltage.

I wonder what their logic was not use something like SS54 in a DO214 package versus that TO220 type package flopping around every time the flippers hammer. SS54 should be rated well enough and less likely to shake loose. I mean they had space for the big SMT inductor and caps, but needed a TO220 like Schottky diode??? I get why the used the through hole MOSFETs, that saved a lot of space.

I made a standalone post but maybe here is better. I'm working on a Stern Kiss and it can't find node 1. The cabinet node. It's lit up and seems ok. Fuse is good. I swapped in a Rush cabinet node and the Kiss is fine and dandy. So, we know it's the node 1 board. So, I removed it from the game and noticed that R59 is not populated. Now I know this could just be by design, so I wonder if any owners of spike 1 games could check their cabinet node and see if R59 is populated or not. Pretty please. There's also 4 solder pads directly above it that don't have anything on them.

Assuming these parts are just normally not populated, what might be a next step to see what's going on?

Thanks in advance for any help.

RC

20220906_181501 (resized).jpg

20220906_181506 (resized).jpg

I ran across this photo where it looks like they are using a SMD package diode. I don't know where it came from, board revision, or anything like that. I don't really know, but it doesn't seem like vibration caused mine to fail. The diode was still intact, and like you say, failing short at least gave the parts downstream a possibility to survive.

449c7152735c92a1d5955972e59e167599aec98f (resized).jpg

Person above just posted an older version (spike 1) that looks like the catch diode is in SMT DO214 package. I can't imagine they want to do the nylon standoff with zip tie mod forever. Hopefully Stern goes back to using a chunky well rated SMT diode.

switching regulators is like just one transistor away from dumping input voltage on the output. I guess the same can happen with the older linear stuff too.

Dunno what the problem is, but if you can't figure it out, try hitting up Borygard - he fixes node boards, and may be able to fix it for you.

In the future, check the 3.3V rail for a short if the 6V regulator is blown. My node 9 board for Star Wars shit the bed ... the 6V regulator failed ... once I replaced that, I learned that the 3.3V rail was shorted the hard way (I am a dumbass ... I remember thinking "I should check the rail real quick just in case" too). The frigging processor itself was kaput (making the board completely worthless since there is no way to get firmware onto a "blank" CPU). I removed the processor to verify that it was, in fact, the component responsible for the short.

Today I learned that there is some kind of problem where node boards will enter the 1Hz yellow LED blinking mode permanently if you powered a fresh node board up in a game that has some other fault that is connected to said node board. I still have more to debug ... plus I still have to find a way to get some kind of upgradable firmware onto my original node board (thanks to Pinballlife for telling me about it). It turns out that the 1Hz yellow LED 'blinking' doesn't necessarily mean that the board is ready to listen to the CPU board.

I can't believe Stern encrypts their damn firmware for these boards. I get that they use certain node boards for toppers and that they don't want people pirating toppers ... but can't they make a frigging topper specific node board with super-secret firmware and just let us have access to the raw binary file we could use to program a repaired node board.

I just noticed the fuse clip on that spike 1 board. They are super cute. Looks like universal clips that takes either 6.3x32 or 5x20mm fuse. Never seen or heard of that double fuse size holder.

Nice job! Typically when I have a power rail short, I find it by applying power and using a FLIR to see what components/traces get hot. If you don't have a FLIR, you can also spritz the board with alcohol; the hot area should evaporate the alcohol around it relatively quickly, so you can spot it.

I completely agree with this. Any time I power up a board after doing work, I probe all the power rails to avoid any shorted power. It's a matter of habit from my day job.

I don't have a FLIR camera, but I've been tempted to buy one. I end up balking at the price, based on how often I would use it. The alcohol trick is a good idea that I've never thought (or heard) of. As long as the converter doesn't go into thermal shutdown, in which case, neither approach would detect anything.

I keep getting targeted advertisements of what looked like a weed vape pen and chunks of rosin(?) to "find shorts". I am still not sure exactly what they want you to do with it, but I figure it is something the procedure is something like watching for solvent spray to evap quickly.

If you dont care much about the board, using a voltage and current limited supply raise up the voltage until you see smoke form =D.

Adjustment instructions:
- Set smoke color to white
- Set amount of smoke to minimum

interesting... So you look for the residue from the rosin smoke to melt.

Yeah I got a handheld unit about 6 years ago, when they were a bit cheaper than they are now. Though they have the "FLIR One" now, which just plugs into a smartphone, for around $300. I use my handheld about 4 or 5 times a year for actual electronics work. Otherwise, it's great for finding the coldest beer in your fridge.

Parts arrived today, and after installing them, the game is up and running. It does take a different skillset, but these boards are not impossible to repair, as long as it's not a part that requires programming.

I have a question regarding failure of a node 10 board on a Rush Prem. If there is some type of issue with the node board then wouldn't I receive an error message in the diagnostics? I checked all wires and they look OK.

Not necessarily. If the CPU board can still communicate with the Node 10 board, it may not throw an error. In most cases the failure on RUSH is the chip that operates the servos for the clock and ramp blowing. Usually it has clear physical damage, but not always.

I've been binging on Stern SPIKE 1/2 information trying to figure our where we are with failures and repair of the various node boards.
It has not been a pleasant journey.

Lack of a complete set of schematics for all node boards ever made.
Anyone have the SPIKE 1 Node 0 schematics for "Transformers (or Avengers) the Pin"?
They are not on Stern's site AFAIK.

Lack of good answers from Stern.

Rapid obsolescence of parts used on the node boards. Ya...I know parts obsolescence is a fact of life but damn.
i.e. A8498 buck converter on the SPIKE 1 Node 0 board.

Dubious (IMHO) explanations for the various failures.

Thanks to quite a few posters for providing bits of information, herg mbwalker barakandl megadeth2600 RobF G-P-E KenLayton (RIP) Wolfmarsh Peak-Pin ... and others.

If anyone would like to contribute to what we know about the SPIKE systems, I would encourage you to either obtain a PinWiki account or, if easier for you, mail information to me and I'll shove it into the Wiki. An account is easy to obtain...see upper right corner of the home page www.pinwiki.com .

--
Chris Hibler - CARGPB #31
http://www.ChrisHiblerPinball.com/Contact
https://www.youtube.com/c/ChrisHiblerPinball
http://www.PinWiki.com - The Place to go for Pinball Repair Info

Much respect to you Mr Hibler. I truly hope you can reverse engineer the nodes as your help will be needed when stern refuses to help the thousands of buyers who have these homepins in their homes with no access to spare parts. Hope you find out all the answers you need. When that day happens spike systems may be ok to use in pinball machines for longevity’s sake otherwise these are disposables… not made for long life commercial use.

How fast some parts go obsolete is a bit frustrating.

I can't find a direct alternative for A8498 right away. Maybe go to LCSC or Digikey and sort all buck regulators by the SOP8-EP package around the same voltage ratings and check each datasheet. Perhaps Allegro would have a recommendation. I clicked on a few likely suspects before giving up, they didn't match the pinout. I would guess there is something out there close enough, that or some kind of mod board to fit a different more ubiquitous regulator.

Will there be a way for 3rd parties to replicate or replace node boards with something more stable or readily available than Stern’s factory boards?

The excuse for the parts shortage is dragging on years now and repair seems very limited to just a few individuals on here.

Even if the board can be replicated there is the issue of the microcode. We’ll need both.

barakandl It looks like Stern switched over to the MP24943 to serve the same purpose as the A8498. However, it is not pin compatible.

https://www.monolithicpower.com/en/documentview/productdocument/index/version/2/document_type/Datasheet/lang/en/sku/MP24943/document_id/1966

I've been talking with another PinSider, mbwalker , about hacking an Amazon available buck converter in. This seems like it might work. Ugly perhaps. But owners of these older "SPIKE 0" games are SOL without some solution.

https://tinyurl.com/4zh3zurd

--
Chris Hibler - CARGPB #31
http://www.ChrisHiblerPinball.com/Contact
https://www.youtube.com/c/ChrisHiblerPinball
http://www.PinWiki.com - The Place to go for Pinball Repair Info

someone needs to dump the SD cards on the pin games / spike 1 home games. Just to have the backups.

Joe_Blasi Do you mean that Stern doesn't have the game code for those games on their site? If so...that is really, really, poor.
Edit: Yep...I see that it is not posted on the Stern SD Card site. Crimeny.
--
Chris Hibler - CARGPB #31
http://www.ChrisHiblerPinball.com/Contact
https://www.youtube.com/c/ChrisHiblerPinball
http://www.PinWiki.com - The Place to go for Pinball Repair Info

Does anyone know the screw / spacer part numbers to mount under the playfield for the Stern SPIKE & SPIKE 2 Node Board (520-7017-72) ?

I wonder if a Spike 2 CPU could run a Spike 1 game. Theoretically nothing stopping it. Might require a recompile of the Spike 1 code though. Might be a longer term solution for Stern rather than supporting/supplying Spike 1 nodes. Of course right now their approach seems to be not supporting it at all, so faint hope...

I had thought about that, but the DMD is the issue... I don't think Spike 2 CPU board has the DMD connector.

However for the other node boards (ie not the CPU) , I would tend to think that Spike 2 boards could be used. When taking a look at the available schematics, the ucontroller portion is pretty much identical between all, with just a node ID defined through switches. The boards are reprogrammed by the CPU board at first boot, and the rs485 protocol is electrically identical. Of course there might be some specific node boards which do not exist with spike 2, that could be a problem

Ah. Good point. Never thought about that.

Maybe a mod board that replaces the entire regulator circuit. The PCB manufacturers can cut a board in half at plated a drill hole giving you a way to solder to a SMT pad.

pasted_image (resized).png

That doesn't help if there are boards dead today. Looks like you can get new old stock A8498 for about $15ea. Likely salvaged pulls on ebay/ali for around $9ea. Damn that part got costly fast. If there is high demand for that part number on the grey market, may not be a good sign there is a drop in ready replacement or else people would be buying that.

barakandl If you mean build a PCB within the footprint of the A8498, damn...that would be a small PCB.

I thought the same thing about the price of the A8498. Never going to be anything to replace it.
--
Chris Hibler - CARGPB #31
http://www.ChrisHiblerPinball.com/Contact
https://www.youtube.com/c/ChrisHiblerPinball
http://www.PinWiki.com - The Place to go for Pinball Repair Info

I was the one that mentioned to Chris about using a buck-regulator from Amazon, etc.

Like you, my first thought was a daughter board of some sort, ideally have it fit on the A8498 pads. Nice, clean, and done. But the more I thought about it, more it sounded like it would turn into a complicated endeavor. First thing that came to mind was the daughter board may need a thermal path to ground (the A8498 had a thermal pad on the bottom). Then I thought since the new part might not have the same topology for the supporting parts, you're adding parts to the daughter board (making it bigger). Maybe you'll need to take A8498 leftover parts off the Spike CPU to make room for the daughter board. Then you have to layout a PWB, order, test. And it might have to be 2 layer for the thermal path. Just seemed to turn into a mess real quick.

You can probably see this was likely going to be harder than simpler. So that's where I thought of just buying a buck-regulator off of Amazon, attach 3 wires (Vin, Vout, Gnd), mount remotely, secure the wires, and move on. Then try to find a A8498 to eventually swap out at a later date. Sounds like the pin owners have been waiting a long time, so this method (albeit a kludge), seemed like a more practical way to get things up and running as easy as possible.

Just my 2 cents.

I meant the entire circuit, not just the SOP8 chip. The flyback diode, inductor, and capacitors are external. I haven't even looked at the board to see whats around it the chip, but you may not necessarily have to use all the pads at the chip, the nearby inductor and diode pads could give you room to spread out.

No DMD display out may an issue but spike 2 seems to have the DMD system still there so it can be done with an LCD display in some way.

But may need to VM the older CPU as direct boot of an spike 1 SD card may not work with the new hardware.

Now is the plan for spike 2.5? or 3.0 still just an better CPU more cores?

And if so then newer cpu board should run older games (may need small software updates to the base os)

I was hoping they would retro fit the spike 2 to work for spike1 stuff so it wasn't eol so early. I forgot to look at GB if it was the same NXP imx6 as a spike 2 or different setup

I think this is spike 1 (picture)? cpu was integrated into the board so there was probably lack power/features for them to expand.

IMX6 which i think is the exact cpu they use, was EOL a while ago, however there was a huge stock of it. I would assume we would see a spike 2.X or 3 w/ imx8 soon ish.

pasted_image (resized).png

Conceptually, it's far less of an engineering effort to port Spike 1 code to run on modern Spike 2 CPU board, DMD issue aside, than creating another CPU board with a more recent CPU just for the sake of spare parts.

DMD could be USB driven (after all, that's what PIN2DMD uses in virtual pinball), so from the HW standpoint they would not need to redesign any new board, everything is pretty much out there already.

Now whether they want it or not is another story... But if they don't have a solution for existing Spike 1 owner whose CPU board fails, this is a customer PR disaster waiting to happen. So really... they should.

Yeah, the only thing stopping Stern from being able to support their older (and newer) stuff with replacement parts or substitutions is Stern themselves. Third party license the boards if necessary, get it down to $150 for a CPU node and $75 for a 8 driver node board, have ample supply out there and suddenly this becomes a non-issue.

Not sure if those prices are feasible, don't know what Stern's cost per board is on this stuff, but in a perfect world they would sell them to verified game owners at cost + reasonable handling fee + shipping. Providing items needed to fix broken machines shouldn't be a profit opportunity.

A Spike 2 CPU board from PinballLife is $399.00 USD. That is $100.00 less USD than a new Xbox Series X. I know economies of scale can come into play here, but....

Ashram56 Agree with you and I'll say that it's already a disaster, and Stern's lack of attention to it is making it worse. A client asked me to repair her SPIKE "0" MPU for a "Transformers, The Pin". After doing the research, the buck converter they used at the time is obsolete, either pricey to obtain or unobtainable, and there is no pin for pin suitable replacement. The part is an A8498.
--
Chris Hibler - CARGPB #31
http://www.ChrisHiblerPinball.com/Contact
https://www.youtube.com/c/ChrisHiblerPinball
http://www.PinWiki.com - The Place to go for Pinball Repair Info

Not purposefully avoiding your main point Chris... but this part came up as a still in production 8498. Of course that doesn't mean you can find stock anywhere, but it is at least a distinction from obsolete.
https://www.allegromicro.com/en/products/regulate/regulators/single-output-regulators/a8498
https://www.digikey.com/en/products/detail/allegro-microsystems/A8498SLJTR-T/1121716

No worries. You are right. I should have said "out of stock everywhere" instead.
--
Chris Hibler - CARGPB #31
http://www.ChrisHiblerPinball.com/Contact
https://www.youtube.com/c/ChrisHiblerPinball
http://www.PinWiki.com - The Place to go for Pinball Repair Info

What is Stern official statement to this customer request for a replacement board ? Do they simply ignore it ?

The customer didn't go into the details.
She mentioned only that she'd been trying for two years to get it fixed.
Perhaps I'll call Stern myself.
--
Chris Hibler - CARGPB #31
http://www.ChrisHiblerPinball.com/Contact
https://www.youtube.com/c/ChrisHiblerPinball
http://www.PinWiki.com - The Place to go for Pinball Repair Info

Just curious, what are the warranty legal terms that would apply in the US ? It can't be considered as a consumer product with one year warranty only, right ?

They are generally printed in every manual. Here is the warranty from James Bond 007 (60th LE)
"STERN PINBALL INC LIMITED WARRANTY
Stern Pinball Inc (‘SELLER’) warrants only to the initial purchaser of its
products that the items listed below are free from defects in material
and workmanship under normal use and service for the warranty period
specified:
• Printed circuit boards (game logic): 2 months
• Dot Matrix / LCD Display: 9 months
No other parts of seller’s product are warranted.
Warranty periods are effective from the initial date of shipment from
seller to its authorized distributors.
Seller’s sole liability shall be, at its option, to repair or replace products
which are returned to seller during the warranty periods specified,
provided:
1. Seller is notified promptly upon discovery by purchaser that
stated products are defective.
2. Such products are properly packaged and then returned
freight prepaid, to seller’s plant."

It's near the back, right before the licensing agreement that says you can't modify the software on your own machine and if you do they can brick it if they feel like it.

but how long until they have no requirement to keep replacement parts for owners?
Or once they run out and make no more ,you have a paperweight or sell for parts .

And likely not the parts another person needs as it will most likely be a NODE BOARD issue and since you can not repair ones that go bad,if the case ,these are just disposable Games like has been hashed by so many stating these SPIKE Games are not a logical buy for people who own real pinball machines that have been working and able to repair the boards due to being very fixable and parts at the ready and think your Spike game will be working forever but may not .

So how long must they be able to repair said product even out of warranty or is their no support required to be given out on an older product ?

I've been going through some of the manuals and schematics, and I found something that I can't seem to match up with anything. There is a schematic file for a 520-7014-00A MPU board, but I don't see this part number listed anywhere other than the schematics itself.

https://sternpinball.com/wp-content/uploads/2019/02/520-7014-00A-SPIKE2-MPU-Module-brd.pdf

Has anyone encountered this part number?

If I had to guess, I would say this part number was an initial placeholder part number the Spike2 MPU that was never released. It probably was just not updated in the schematics. This seems logical since no schematic exists for the actual Spike2 MPU part numbers; 509-1000-XX and 509-1003-XX.

Here's an odd thing tho...I don't see any of the backbox LED's in that schematic. I spotted some LED's in the schematic, but they don't appear to be for the backbox.