So, Stern has recently shipped a fair number of machines across many titles (recent Maiden, Godzilla, Foo, Venom, and Bond 60th are confirmed to have machines with this issue in the wild) that excessively click before they start booting. Here's one (extreme at around 40 clicks) example of what that is like at boot on the affected machines:

The clicking is coming from the power supply, but after extensive investigation (documented in the Foo Fighters owner's thread where this started), the real culprit turned out to be the capacitors on the Power Distribution Board (or PDB). I've numbered the problematic caps on the PDB here:

PDB numbering (resized).jpg

The caps are made by Cornell Dubilier Electronics (CDE) and are rated at 10,000uf with 20% tolerance. The reason they're suddenly starting to click when they never did before is somewhere in late 2021, CDE changed the makeup of the caps inside. Early (maybe mid) 2021 and before versions of the same caps are 1.4oz. Late 2021, and all 2022 and forward are 1.7oz. The heavier caps have crappier performance, with at least 50% of the ones I tested hanging out at the bottom end of the tolerance with 8500uf or less (MANY in the 8200-8300uf range).

The problem with low-spec caps below 8500uf is the meanwell power supply apparently thinks there's a short when they charge up too quickly, pulling more power in a burst at boot and it kicks in their protection circuit. That's why you hear that relay clicking on and off in the power supply (officially called "hiccup mode" by Meanwell) when the game is turned on. Since that's where the sound is coming from, people think the power supply is the problem, but the power supply is working as designed. The newer CAPS on the PDB are underperforming relative to Stern's engineering assumptions when the PDB was designed, causing the power supply to make the clicking until the caps finally charge up. Inrush (the surge of power at boot) from the massive capacitance Stern built on the PDB exceeds the max temporary draw for the power supply. With the older caps the inrush was low enough that it didn't pull too hard and trigger the power supply safety circuit, but the new material inside the same CDE caps changed the inrush peak so now it's over in many cases, causing clicking while the power supply tries to deal with the boot surge.

Figuring out what caps you have is key. On the top of each cap there's a 4 digit code pressed into the plastic disc. The first two numbers are the year, the second two are the week of the year that cap was made, and a "P" to denote that this is a polarized cap, so install orientation matters. 22xxP and some 21xxP with numbers for the xx over 40 are ALL suspect with about 50% of the caps at or below 8500uf tested on a 10,000uf cap. Given that you have to have THREE >8500uf caps on the board for it not to click, your odds of hitting on the right side of that 2022 lot 50% three times are 12.5% - not good. 2236P is more likely to have a better than 50% pool of good caps, but 2241P and 2245P are pretty uniformly awful. ALL the early 21xx caps I tested were well over 9,000uf, most 9500uf or above, so GREAT performance, no power supply clicking. 2119P is the lot I have most often, all great, zero duds like 2022 has lots of. EDIT: I received some 2141P caps, so late 2021, and they are also heavier at 1.7oz and have lower spec performance at about 8600-8700uf for the ones I tested, so the material change seems to have happened late in 2021, not 2022. Early 2021 lot caps are still the ones to get.

If you have excessive clicking at boot (I would call that 5 or more clicks - some people have had 50 or more, and some so many the game NEVER booted), I would pursue a warranty claim with Stern for another PDB. It doesn't appear that they are pre-testing the replacements they send out, so you may get another click-causing board, but persist. For some reason Stern is ALSO sending a new power supply in some cases. This is COMPLETELY UNNECESSARY since the problem IS NOT on the power supply, despite the fact that the power supply is clicking. If this happens to you, only replace the PDB and that should do the trick (unless you're unlucky enough to get ANOTHER click-causing PDB as a replacement - it's happened).

Once you get the replacement PDB installed, give it a few days of power cycles. It's not uncommon for the first one or two or maybe more to not click and it seems like problem fixed, but then it starts clicking again.

If you cannot get satisfaction from Stern, or you have the skills and just don't want to hassle with them, you can 100% fix this yourself by replacing the caps with the known good early 2021 lot caps. DO NOT get them from Mouser or Digikey - I ordered from both in the investigation phase and got late 2021 and 2022 lot caps from both. I found a distributor that has a few hundred early 2021 lot caps (2119P). I've ordered a few dozen from them and all are early 2021 (EDIT: As of 10/24/23 they're now shipping later 2021 caps). Best part is, it's $10 plus $8 shipping to get all three caps you need - cheap. Here's the link:

https://us.rs-online.com/product/cornell-dubilier/slpx103m063e9p3/70189937/?m=70189937

These Caps are all ok for the PDB, according to Stern:
Samyoung RDC63VN10000M (Samyoung is a garbage cap company, so not sure I trust these)
Cornell Dubilier SLPX103M063H7P3 (shorter, fatter version of one Stern uses)
Cornell Dubilier SLPX103M063E9P3 (the one Stern uses)
Surge / Lelon LS-103M1J- - -3050S (haven't been able to lay hands on one of these)

Note that the caps are pretty hard to get desoldered and off the board. I had to crank up the power on my weller to almost max temp and use donor solder to get them off. Definitely a danger zone for accidentally lifting pads if you leave a super hot iron on too long, which would multiply the problem, so be careful. This is probably a good time to break out that hakko desoldering gun if you have it (equipped with a larger than stock tip).

Also note that the right side of the PDB has live 120v power, so UNPLUG your pin before attempting to remove the PDB board. Switch off isn't enough.

Whichever route you go, good luck. The good news is, the clicking CAN definitely be fixed, one way or another.

I think you'll probably be fine. If it were extremely widespread I think you'd see a lot more complaining. Right now it's at what I would call a simmer. Stern engineering is definitely aware of it, so hopefully they'll have their board vendor pre-screen the caps to make sure they're >8500uf before they put them on boards.

I did OK with the NIB new run Maiden Prem I just got (Edit: Not so fast - the person I sold it to started in with the clicks, up to 30. I am fixing their PDB), and my FF LE didn't have this issue, either. I had to have someone SEND me their click-causing board to do the original investigation. I also have a Venom Pro and Premium coming.

Buy your pin.

Note that Stern doesn't seem to be pre-screening replacement boards, so you might get a PDB that still causes clicks in exchange.

Note that PBL stock also potentially causes clicks. I got one from them with the same low-spec cap lot that made the power supply click (until I replaced the caps).

Edited the post. Received some late 2021 caps (2141P) and they are the heavier (1.7oz), crappier performance ones. So the change of materials that made the caps perform worse happened sometime in late 2021. Early 2021 caps are all fine.

I've done 3 boards so far and the 2119P 2021 caps from the link I provided solved the clicking on the power supply for all three. All of the replacement caps tested out at over 9000uf before install, so nowhere near the danger zone. None of the boards have slipped back into causing the clicking again.

The cap I linked is the exact part from the same manufacturer as the ones on Stern's PDB. Same capacitance, same voltage, same temperature rating, etc. Yeah, the prices at RS are markedly better than anyone else, which is a nice surprise given that they have just the caps needed for this fix. Makes this repair a very reasonable $10 plus $8 shipping (if you only get 3, I bought many so I have a stockpile).

Yeah, I wasn't doubting my results, but when that one pinsider reported only getting a power supply to replace from Stern warranty support, I was thinking the solution may get more complicated if it could be the power supply OR the PDB OR Both. But in the end, it's the PDB. Keeping it simple.

No idea why Stern warranty support is sending out just power supplies when I KNOW Stern Engineering knows the problem is with the PDB.

I have no idea. It may be that meanwell's power supply operates differently in 230v/50hz mode. It will be an interesting data point if the clicking isn't happening outside of North America.

It's not impossible to get a PDB with 22xxP series caps that doesn't cause clicking, it just is more likely to have issues than a 21xxP cap because of whatever manufacturer change to the cap's composition was that made them heavier and crappier. But I stocked up on 2119P caps so I have enough on hand to handle any ones I may end up with that cause the power supply to click.

I'm sure Stern will get this worked out, either by changing cap suppliers or by forcing the supplier to screen their caps for ones that fall above 8500uf. It will just take time to filter out.

Power supply has proven to be reliable. This is an issue of a manufacturer changing the cap construction (likely without telling Stern) for the caps used on the PDB. Fix the caps, fix the problem. No need to jump to an unproven new power supply.

Check out the video in the OP. That's an extreme case of like 40 clicks. If you eventually have less than 5 clicks, I wouldn't be worried about it. A couple clicks is normal.

Glad to help. I know i'd be pissed if Stern sent me a NIB that was clicking excessively and support told me it was "normal" with no way offered to fix it back to ACTUAL normal except maybe another PDB that might have the same issue.

Happy to provide some options for people in this situation (and knowing I will eventually lose the Stern lottery and end up with a clicker myself that I now can fix).

I have a Bond 60th PDB coming in that is causing the clicks. Interested to know what the caps read on a $20k machine. If Stern isn't pre-screening PDBs for their highest-end machines, they just don't care.

Capacitance. It gave me the info I needed to sort caps by which would cause clicking and which wouldn't.

Since it's not clear what the core issue is, just symptoms and the low spec caps somehow affecting that, it's hard to say. All I can guess is electrical conductivity changes with temperature, so cold vs warm probably has something to do with it.

Doesn't look like it broke during assembly, it looks like that resistor blew from a short or overload. Hopefully the problem isn't somewhere else and it blows the resistor on the new PDB again (or you don't get a click-causing PDB in exchange).

I don't have a schematic, so I'm not sure where R5 goes on the PDB. Someone posted a partial schematic somewhere, but I didn't save it and now I don't know where I saw it. If the shaker motor is constantly running, something is shorted open. Question is what. I would guess a transistor (there's an IFR540Z surface mount one RIGHT next to the shaker motor connector, which would be my prime suspect in this case), but...

If you have a multimeter you could test that transistor to see if it's shorted open. Alternately, do you have another Spike2 machine? You could swap the coin door interface board and see if that fixes your shaker problem. If it does, the problem is on that board and the cracked PDB resistor is unrelated.

It's pretty common that when a resistor or transistor blows it cracks the casing (sometimes blows it completely off). That looks like a blown resistor to me. It's surface mount, so unlikely to have enough force at install to crack the casing that way.

Something before that transistor could be telling it to open, then. Without a schematic, can't really guess. Will be interesting if that PDB swap fixes it.

I processed one of his pics. Is this enough?

Spike2-PDB.jpg

Unfortunately the cracked resistor means this one will never be "better" since the crack blew out the part number so all numbers will never be visible as-is If you just need a clean PDB pic, I can take one for you.

If it's going to click usually it's within 2-3 days of a PDB swap, so you're almost home free.

My 50% of the late 21xx and all of 22xx caps being below the 8500uf threshold is based on a small sample of caps I've personally tested (probably 30-ish). So the percentage that are ok, may be 60% or even 70% across all the ones made, increasing your odds of getting a good PDB with 22xx caps. Given that I'm working from the outside, I don't have all the information. And, maybe Stern IS finally pre-screening PDB boards. That'd be a step in the right direction!

Fingers crossed your swap holds!

Boo. Hopefully Stern warranty eventually starts pre-testing these PDBs before sending them out. They know what the issue is, they just need the will to do the screening.

Board worked fine for like 6 years (I don't remember if Spike1 used the same board) with no problems. When the caps changed, the problems started.

But, Spike3 is imminent, so maybe there will be a new setup. It's just not really necessary to throw the board out when it's been ok for years. Just solve the cap problem and pre-screen boards until that is fixed.

The problematic caps are through hole.

But surface mount is clearly the way everything will be, so I'm going to have to break down and get a station to work on them, too. Been putting it off...

Haha, I mean for my machines. There are already pinsiders that can do board level repair much better than me, like ChrisHibler and @borygard. Both of them already offer node board repair service.

2245 (45th week of 2022 build) capacitor date code, so still possible to click because they're the newer, heavier, crappier material inside the cap. Does it click excessively?

I would say 3-5 clicks is about normal. Problem is these 2022 caps can turn from 5 clicks to 15 clicks in weeks for some reason.

The RS company I linked in the OP has only ever sent me 2119P, so I think that's all they have until their original 500 run out. They've around 350 left now. I stocked up on a few dozen so I can fix NIB machines I get with the issue and don't have to bother with the Stern warranty lottery.

I don't know if ChrisHibler or Borygard will be at the Cleveland show, but if they are, I know Rob often does board repair at shows, so that would be ideal.

Wow, that's crazy about all the missing components.

Can you post some pics from the side of the new/different caps so I can try to figure out what they are?

Samyoung is a low-tier Korean cap that got a really bad reputation for swelled/leaking caps in the capacitor stolen formulation plague in the mid 2000s. I hadn't heard of them in more than a decade since then as they were to be avoided at all costs. It'll be interesting to see if Stern made a wise or foolish choice giving them a shot.

I'll try to get some in to test, but after a cursory look, they seem to be hard to get in the US.

Anything over 5 clicks and I'd ask for a replacement. The problem is, the clicking is the power supply failsafe kicking in - 10x at boot is NOT normal. And we have no idea why the materials in the late 2021 and later caps is performing so badly and if they'll get worse. I'd get yours performing correctly while it's still under warranty. We ALSO don't know the long-term consequences on the power supply this is stressing. Will this PDB with bad caps cause it to fail earlier? We don't know.

With 5-6 clicks you were right on the edge as far as where you decide to get a warranty replacement or not. Glad the 2021 caps worked for you. 38 is late in the year, so might be close to when they changed the material to make them that made them crappier, but sounds like you ended up on the right side of that. All the caps I got from the link I posted were 2119 dates, which is why I hoarded a bunch (fortunately they were cheap) to make sure I didn't have to deal with clicking if I got a machine with the issue.

Sounds like Stern is trying other cap suppliers, but the one they're shipping with right now doesn't inspire great confidence since they had the bulging and leaky caps with a bunch of lower-tier manufacturers working from the stolen cap formula about 15 years ago. We'll see...

But the parts that are missing are for the door switch, according to the schematic...??

I marked the stuff from that section of the schematic with green dots and they're all in the door switch support area...

revised pdb (resized).jpgannotated-revised PDB (resized).jpg
So are they not checking for door switch state anymore?

They ceased checking the door state before, but last time they just didn't ship with interlock switches until the hatemail got too deep and they put it back.

Total BS. You can put a new bad PDB in an older Spike 2 machine and THAT power supply starts clicking. It's been done. No idea why Stern engineering can't do this test themselves with the breakroom pins.

It's the PDB.

They're aware of it all. The hard part that I don't have a specific answer for is "why." I know the cap manufacturer changed the materials in mid 2021 and all the caps after that perform worse, but I don't know why that is. But jumping to Korean Samyoung caps as a replacement feels like out of the frying pan into the fire given that company's history cutting cap corners.

Pinball IS hard.

Yeah, my LCR meter didn't really do ESR, but I'll be able to borrow a dedicated ESR tester shortly and will check the heavier and lighter caps to see if there's a difference.

The unit seems to have a self-calibration mode, so that will be ok.

Okay tested a bunch of crappy 2022 CDE caps and the benchmark 2119P ones that always test good on a dedicated ESR tester.

SURPRISE! ESR is the same on both - .02 Ohm, which is apparently good since acceptable ESR for 10000uF 63v is .04 Ohm. The tester did state that it gives an absolute ESR and does not judge the REACTANCE of any specific cap. So as of now, the only two things we know are different after mid 2021 caps is:

Weight (1.4oz = good, 1.7oz = bad)
Capacitance (>8500=good, <=8500=bad)

I don't know how much the reactance affects this issue, but it's predicted by the low capacitance in the tests of the bad caps. For example, an 8500uF cap at 100Hz would apparently be .187 Ohm reactance. The same model cap with 9500uF at 100Hz would be .167 Ohm reactance.

Not sure what else to measure to find a difference. I know how to spot caps that will cause clicking (weight is a maybe, Capacitance below 8500uF is a definitely), but I still don't have the "why" in all this. I just suspect is poorer materials in the heavier caps from mid 2021 onward, but it's only a guess since I haven't come up with anything else.

They had around 500 caps when I first posted that link, but are now close to only 300 left, so they must have run through their early 2021 stock and are now into late 2021. I know that I got 2136P elsewhere and they were the heavier, crappier caps, so the changeover happened somewhere between the 19th week (2119) and the 36th week (2136) of 2021. I just don't know exactly where.

That said, if you're testing at 8900-9000uF, you got perfectly workable caps that are well over the 8500 cutoff zone, so you should be fine.

It's a dedicated ESR tester - ESR70Gold. Calibrated it before testing and did it on a bunch of the 1.4oz and 1.7oz caps. The result was consistent.

Yeah, two digits are all you get. Should be enough though. I don't think ESR is the problem if the difference between them is down to 3 digits past the decimal.

He got a free spare $75 PDB. I'd be happy to have them send me one. Yeah, the situation sucks, but a free board's a free board. The caps to fix them are cheap.

I don't think it would mess anything up on the bootup side. It just will probably bleed the capacitor reserves too low for proper functioning in power-intensive modes like multiball once you're playing a game. Houdini had this issue until they added a capacitor board to buffer the power. Everything got very weak when a lot was going on.

Trying this is on my to-do list, but I need a board that causes clicks enough to see the difference once I've done it. I have a couple new ones I haven't tried yet. If one of them causes enough clicking, I'll try it.

A> We now have a name for the clicking - "Hiccup mode"
B> No way Stern is coughing up 30-40% more money for the higher-spec HRP-600N3-48 power supply.
C> However it WOULD give them 124W more power output for more BoM toys.

Nice that Meanwell got back to you with some clarification about the issue from their end.

Seems like way overkill though, especially since the machine can't take advantage of the extra power budget AND the power supply is a LOT bigger. Putting in good caps seems like a much more reasonable fix.

Basically send it back and try again hoping you hit the estimated 12.5% odds of getting a 100% good one, or send one out, paying someone to replace the caps with the known good ones. I've done a few PDBs of mine now and it 100% fixes the issue.

Essentially you have to choose time (repeated free warranty replacements until your luck getting a good one hits) or money (you pay to have the caps replaced). Thanks, Stern.

I don't know when in 2021 they switched over, because I have 36th week caps that are the heavier, crappier construction. I will say that the 2136's are still usually rating higher than anything in the 2022 run, though. 2119 are the best ones with ratings usually 9500 or above readings.

That sucks. How many clicks are you getting at boot?

I wouldn't add more potential failure points to the problem. Pre-screening the caps to try to get them from the "good" batches before mid 2021 (or worst case, testing the heavier crappier caps so you only use ones above 8500uf) will absolutely solve the problem. 100% success to my knowledge. I'd just do that and call it good.

Yeah, that's way too many. Time to sign up for the Stern warranty lottery or just have the caps replaced directly with known good ones.

Updated a clicking PDB board this week and tested some 2138P (38th week of 2021) caps that I expected to be the 1.7oz crappy ones, and they were actually the 1.4oz good ones that all tested above 9200uF. So I'm still not sure *exactly* when in 2021 the changeover happened. I only know that 2119Ps are all 1.4oz "good" ones, and 2138P has been the 1.4oz and the crappy 1.7oz, so it may be the time when the production changed over at CDE.

Swapping the PDB is straightforward. Just make sure the machine is UNPLUGGED as there is live 120V on the right side of the board, even with the power switch off. You need to remove two nuts first, one on the service plug cover next to the fuse, and one on the upper right of the PDB. Then remove the service cover and remove the remaining 3 PDB nuts, unplug everything and make the swap, then reverse the process.

If you end up doing a cap replacement, if you order from the link I provided, they still seem to be sending out 2021 caps, but later in the year. 2119P are the best, in my experience (maybe send a request with your order for that lot) as I've had zero of those test below 9200uF, where some of the 2138Ps I've received have tested 8800uF range (which is still good, but may still cause a handful of clicks).

Cap replacement is straightforward. I've been using a Hakko with a larger tip (the stock tip is too small) and it takes seconds to get all the caps out. With a weller temp controlled soldering station I was using initially it was harder as the solder was stubborn. Just make sure you don't linger on the pads too long and lift the pads off the circuit board.

What brand? If they were CDE, what lot number(s)?

I don't know if I'd trust those caps. The amazon listing says the brand is both "Eboxer" and "ASHATA" and neither are brand-name caps, so likely generic chinese junk cap parts of the kind you see resold all over alibaba with no traceable source. The listing is also riddled with errors. AND the price isn't even better than the legit CDE caps I linked in the OP that are $12.52/4. The price diff is still only $4 more if you factor in shipping (assuming the Eboxer/ASHATA was free ship with prime).

I was making the comment for others that may be interested in trying those as well. I know you're already into the beta test for them. Generally, in my experience with no-name Chinese bulk caps sold under a million names, they don't fail up front, but have dramatically shorter lives than more recognizable, traceable, brand name caps. So I would look for issues in 1-3 years. Fortunately, if they leak, the caps are on the bottom of the PDB so they're not leaking all over other components, just the inside of the backbox.

EDIT: I could not find a spec sheet for ASHATA or EBOXER, both of which those caps are sold as. I was looking for the ripple rating, specifically, and came up empty. However, one of the reviews was verifying the concern about the wide variance in quality of those "no-name" caps out of the box:

"Just got all 4 and measured.

C1-9,870uf -1.3%
C2-10,040uf +0.4%
C3-10,350uf +3.5%
C4-11,910uf +19.1%

Total-42,170uf +5.425%"

The problem is 100% NOT the power supply. Stern knows this. No idea why their warranty dept keeps sending out power supplies.

In mid-late 2021 CDE was switching over to the crappier cap materials. The only way to be sure which ones you have in mid-late 2021 (all 2022 are the crappy materials) is to weigh them. 1.4oz is the older materials that are rock solid. 1.7oz are the newer materials that are inconsistent. You can still have 1.7oz ones that do ok, but it's only a small percentage of them.

Yeah, I'd say anything under 5 is ok with this new normal. Problem is, if you're getting any clicking it can progress over time to more.

Stern QA for the win. I think the factory changeover may have introduced more chaos than expected. Hopefully it smooths out.

Sounds like your scale needs to be recalibrated. It's weighing heavy.

Thanks for posting the update! And yes, I wasn't joking about the risk of lifting pads due to the heat required to remove the caps. If anyone is not sure about their (de)soldering ability, let someone else with more experience do it to save those solder pads.

I think it's great that people are able to diagnose their clicks and fix their PDBs with the instructions provided, it just sucks that Stern is shipping pins out every day of the week now that they know are prone to this problem with no fix seen for months to address it at the source.

Did you request a specific lot of 2021 CDE caps? It may be a service fee to dig through and find the lowest date code you requested...that's all I can think of.

I'm still working to find other caps from reputable manufacturers that just work without specific lot codes - it IS ridiculous that we are scrounging to get back to where we were in 2022 machines.

On the "approved" list, Samyoung has a very bad rep historically, and the Surge / Lelon ones only seem to be available overseas in Asia. I tried the other CDE ones, but they were likely the same questionable material as the taller ones with middling performance, so those are out.

I don't see anything special about the pads. Here's a pic of one after the caps were removed. Also, this might be a PDB pic for your file, ChrisHibler as the part numbers are all pretty legible.

Stern_Spike2_PDB.jpg

Welcome to the club. I don't think anyone knows the root issue at the moment. Just the symptoms and how to fix it with better performing caps.

The power supply didn't change. The only thing that changed was the capacitor performance, and I don't think Stern knew about the change of materials inside the caps (making them clearly heavier) that made them perform worse. CDE just DID it. Something about the material change in the caps lowered the capacitance to the very bottom end of the tolerance compared to prior version of the same caps that performed near the top of the tolerance range. That is the only differences I've noted - weight (signifying a change of materials in the cap) and very low performance at the bottom of the tolerance range for the newer run caps from CDE. I believe there's some other parameter not yet sussed out that's causing the hiccup mode in the power supply. Still working on it, but I am not an EE, so it's slow going. Stern PAYS EEs, so this should never have made it out of the factory to become an end user problem with an unacceptable Stern "as designed" cover story.

I don't think anyone wants to degrade the machine's coil performance to compensate for lower performance caps. The goal is to understand exactly what's happening and why the caps with the newer lower actual capacitance rating seem to not charge as well as the older caps with the much higher real capacitance rating that don't have the clicking. It seems backwards from what you would expect.

It's the same power supply that has been perfectly adequate in spike prem/le machines for about 7 years, so thats not the issue. The material change in caps on the power distribution board external to the power supply that happened this year have caused the problems. The question is why? Still working on that one.

That HRP-600N3-48 is 1.5x as expensive as the RSP-500-48 Stern's using now, so there's no way Stern is using it, even if it is "right sized" for the initial load they're pulling. I AM thinking about trying to mount one in a Stern that currently has a PDB causing clicking to see if that surge ability it has solves the hiccups. Unfortunately the HRP-600N3-48 that has that 350% surge demand feature is more expensive than just the HRP-600N-48. Because, of course it is.

It may all be moot if Spike3 debuts with Jaws next as that will have more cost reductions and redesigns, including possibly the new power supply George promised over two years ago.

No, but he had specific areas that he was talking about it improving, so it's pretty clear it was underway. It seems like a new Elwin at the beginning of a new year would be as good a time as any to debut it. Cost savings are like crack to Stern, and Spike3 is all about squeezing more dollars out of the cost of making a Stern.

The SPIKE2 CPU also puts out more heat than that crappy heatsink can dissipate. It really needs a fan as well, but that's another few bucks. Probably getting another simple heatsink with Spike3.

Long hard road, my friend. You might get lucky, or you might be stuck swapping un-tested warranty replacement PDBs that also click a bunch for the next 6 months, every 3-4 weeks until Stern sends you one that won't click more than a few times. Stern has Game of Throne'd this problem moving the clicking to "as designed" and calling it a day.

Yeah, it's unacceptable, but, well, Stern.

I haven't seen it reported, but since the theory is that the problem is caused by the current inrush being too much for the new crappier caps to handle, which is after the ac/dc conversion in the power supply, so it seems like the AC power shouldn't matter. But it will be interesting to see some overseas reports.

Stern knows. I've shared everything with them. No idea why the warranty department is still occasionally sending out power supplies and hoping that helps. They KNOW it won't.

Yes. There have been reports of machines that won't boot. Seems to be the ones that get over 50 clicks that fail to boot.

Beyond the PDB? I doubt it.

But the the PDB can and has failed over time because of this. Just doesn't boot. That's a $75 cost to the user every time past warranty. AND you're working relays in the power supply that almost never get triggered beyond power up, so eventually it could weaken or kill them, too. That's $130 to the user. But those will be down the road and not Stern's problem, so I'm sure that's their thinking. Call it "as designed" and hope their customers buy it.

RS is shipping caps later in 2021. Mouser and Digikey have always had the newer (wrong) ones. 2023 run is not safe, someone just posted a machine with 2314 caps that clicks. You may find some good ones in there, but it will probably be a small percentage like late '21 and all of '22.

It's one reason I'm trying to find a more reliable overall cap that doesn't rely on lot codes. One thing's for sure, it will be more expensive, though. All the ones I'm seeking out are >$5 each. Some as high as $16 each.

35mm will have them touching, and maybe overlapping. It's *right* on the circle around the cap on the board. CDE has a fatty that's shorter, too.

EDIT: Tried three of the fatty 35mm version of the 10,000uF 63V cap and they do fit, but just barely. The caps touch at their bases. Makes for a cuter, more compact PDB, but with the same CDE mid-2021 and forward performance issues. So 30mm diameter caps from other manufacturers are preferred, but 35mm diameter will fit, if just barely.
CDE_Stubby_Caps (resized).jpg

The ripple current on that Vishay is 4.23 @120Hz. The CDE is 5.25 @100Hz. There's no straight comparison, but the ripple current rating on the CDE is definitely going to be higher. Nichicon (a more trustworthy name than Vishay, IMO) has a cap where all the specs work out except, again, the ripple. But it could be a case of the more reliable cap (Nichicon) giving a more truthful rating and CDE fudging. Kind of like RMS power for amps.

Nope. Zero interest in doing that. I'm happy to point people to a solution that will work, but I don't want to be in the board replacement (or rebranded cap) business.

Ah, right. Great point. I forgot that the Meanwell is smoothing out the DC anyway before it goes to the PDB. I won't get so hung up on ripple rating for the PDB caps.

I would definitely put Vishay above CDE, but for me, Nichicon is the gold standard for good specs and reliable operation. So I was happy to see they have a 63V 10,000 uF cap. However, that awesomeness does come at a cost. They are NOT cheap.

Yeah, Nichicon is into math. They make you work for the specs that everyone else just gives you. I have no idea why, other than they're Japanese and hardcore.

There's a 105 degree version of that Nichicon (GU series), but it is also 35mm instead of 30 like the LS series/stock CDE caps and is probably overkill, so I'm going to give the 85 degree one a shot now that ripple isn't so much of a concern. Nichicon has never let me down, so I have high hopes for these. The price is like 70% more than the bargains at the RS site for the CDEs, but if the correct lots of those are drying up, that doesn't matter. This'd still be a pretty reasonable ~$25 with shipping for 3 caps.

That date code could go either way on the internal build. I've had 2136 that have been ok and others that were the heavier, crappier ones.

I have the Nichicon caps on the way, and if they work out, that will be a more reliable way to go since people won't have to worry about date codes.

The N3 power supply also arrived today, so in the next week or so I'll be trying that out to see if that solves the problem a different way by being able to handle 350% of spec inrush. It's more expensive, but it's easier because you just swap out the power supply and continue on as normal.

Sounds like Stern Engineering is pretty set on not addressing the core problem for whatever reason. Weird.

I will be interested to know if a new power supply solves the issue on @skyking2301's machine. Everyone that did a power supply only swap still had clicking, though it worked okay for a few days before returning to clicking in one case.

Especially since ~30% of those are going in route machines and the operators may not care as long as they eventually boot.

I do think that Spike3 is imminent, though, which could account for the lack of urgency. If it's backwards compatible, they could just move all re-runs to Spike3 and leave Spike2 behind. If there's a new power supply (likely) and new electrical engineering, problem solved in their eyes. The "as designed" Game of Thrones "fix."

It may or may not get worse. But 20-25 clicks is pretty annoying. I'd fix it if I were keeping it.

Stern can do a Half-Life pin and the clicking would be on brand.

They're off in the weeds now saying it's the power supply. It's not. But they have papered engineers (from DeVry?) saying it's meanwell's power supplies, not their cost-reduced PDB, and have chosen to go with that. <sigh>

Replacing the caps with ones from a lot that is not the crappier, heavier build fixes the problem 100%. Proven.

They have, they do. I was sharing the information as I worked it out. They took a hard right last month to blaming the power supply, which makes no sense at all as replacing the caps with correct ones from non-crappy lots always fixes the issue (in my experience), whereas replacing the power supply without addressing the PDB rarely or ever fixes the issue.

Weird. Dunno what's up with that, but maybe they'll eventually come back to reality. I'm totally open to blaming the power supply, but based on all my testing so far, it makes no sense to do so, EXCEPT that the PDB seems to be pulling above the power supply's spec at initial power on, but that's not the power supply's fault, and it's been that way since ~2016 and the clicking just started now with the cap change on the PDB.

So, essentially you ended up with a new PDB that doesn't click. It's great news, but don't count your chickens before they're hatched. Sometimes the replacement PDB can start causing clicks again within a week. So schedule the celebration for two weeks from now.

It may have to do with temperature or how long it's been since the last time it was turned on. Even Stern obviously does not know.

The problems started when CDE changed the construction of their caps. It may have shown up a problem with Stern's engineering (seems most likely) of the PDB board, which apparently exceeds the rated capacitance of the power supply, so at inrush you see the hiccup mode as the power supply tries to deal with it. I still have my 350% "surge capable" power supply I haven't tried yet, but I bet that fixes a "click-causing" PDB because it addresses what seems to be coming into focus as the actual problem - inrush demand overwhelming the power supply at boot, triggered by these lower-quality build CDE caps that happened in mid-late 2021 and showed up in Sterns late in 2022.

They are off in the weeds. We have a difference of opinion, based on the testing and trial and error I've done here with support from other pinsiders. "It's the caps, stupid" comes to mind. But it's not TOTALLY the caps. Seems like the engineering of the PDB was weak from the beginning, and the cap change just showed that up.

Um, that part has blown. It's supposed to be a square. All the casing has blown off.

Here's another one where R5 blew, but unlike yours, the casing just cracked...

Spike2-PDB (resized).jpg

I think it is, it seems like the black square casing just got blown off.

I have some of these here on my "to do" list to try. I think they will work based on the specs, but I haven't tried them yet. I'll try to get to them this weekend.

I think it's just a matter of perspective of the side shot of the blown one vs the overhead shot. If you use the 4 in "48" as a guide for the left end and the 5 in "R5" as a guide for the right end, both the casing version and the blown off exposed resistor version are about the same length.

Give it a week before you celebrate. It can take a bit before clicking is verified gone.

Interesting. 2311 is a BIG jump in stock lots to March of this year. They jumped over almost a half a YEAR of questionable/bad cap stock from CDE. I wonder if those caps are back to the other materials inside that weighed 1.4oz? I guess I'll see if Stranger Things or Jaws Pro/Prem are using these new PDBs. I'll unsolder one and weigh it.

Probably some variance (accepted tolerance) in the power supply surge inrush current at power on. The caps are so crappy that small variances in the power supply that normally would be invisible make a big difference. That may be why Stern is off in the weeds thinking it's the power supply when it's very clear it's the crappy caps.

Since Stern now has at least SOME PDB boards with caps dated 2311, I'd request that the PDB they send you is from this lot. It's dead simple for them to check the date code on the caps before sending it out. We have one success story with 2311, maybe you'll be the second. Here's hoping CDE went back to the more reliable construction and 2311 onward are rock solid again.

Fingers crossed that it doesn't cause clicking and 2023 lot caps may be the answer...

Thanks for confirming your 2023 caps were still the heavier (crappier) builds. I guess that means 2023 CDE caps Stern is still using remain a lottery, hoping you get ones with high enough capacitance to not cause the power supply to click.

I think temperature certainly plays into how many clicks you get on a "click amenable" PDB.