Dare I say Bob Roberts is the Alpo of parts? 🐕

[andrewb]

Nicely done.

(And for anyone who wants to skip to the punchline, jump to 12:00.)

Regarding your comment about "who knows if these were on the verge of failing, or if they'd last another 45 years", the answer to that would be that they are not on the verge of failing.

How do we know? Now that you've tested the caps on one board, test hundreds of them, across many boards. I have. And they've all been fine.

I've also never seen a problem on an Atari board caused by a bad Nichicon cap.

If either of those things were not true, that would be reason to suggest replacing them. If some had actually measured bad on some boards, then statistically speaking, others on other boards would be on their way. But that isn't the case. That tells us there's margin. I've banked my reputation on that fact, selling game boards and AR's with original caps on them for years here, and warrantying every one with no issues.

So if you have original Atari Nichicons, keep them. Maybe this will change in another ten or twenty years. But until and unless there's data to support otherwise, why replace originality when there's no need? If the original interior of your 1965 Mustang was still ok, would you rip it out and replace it? Of course not.

Honestly, the chances of a random logic chip or RAM failing are far more likely. But you aren't about to go and replace all of those.

 

[andrewb]

Get out of the 80's. Wall game remotes run on RF. I also see other pretty uncommon and undocumented things in my travels.

Can you point to a case where you found a bad inductor in a game, and needed to measure the inductance to find a replacement?

 

[rod90]

Can you point to a case where you found a bad inductor in a game, and needed to measure the inductance to find a replacement?

Will you concede that tuning inductors in an RF environment is within the scope of this hobby?

 

[ArcadeTechGW]

Nicely done.

(And for anyone who wants to skip to the punchline, jump to 12:00.)

Regarding your comment about "who knows if these were on the verge of failing, or if they'd last another 45 years", the answer to that would be that they are not on the verge of failing.

How do we know? Now that you've tested the caps on one board, test hundreds of them, across many boards. I have. And they've all been fine.

I've also never seen a problem on an Atari board caused by a bad Nichicon cap.

If either of those things were not true, that would be reason to suggest replacing them. If some had actually measured bad on some boards, then statistically speaking, others on other boards would be on their way. But that isn't the case. That tells us there's margin. I've banked my reputation on that fact, selling game boards and AR's with original caps on them for years here, and warrantying every one with no issues.

So if you have original Atari Nichicons, keep them. Maybe this will change in another ten or twenty years. But until and unless there's data to support otherwise, why replace originality when there's no need? If the original interior of your 1965 Mustang was still ok, would you rip it out and replace it? Of course not.

Honestly, the chances of a random logic chip or RAM failing are far more likely. But you aren't about to go and replace all of those.

Without data, all you have is an opinion. The late Hans Fauske said that, and he's right.

 

[andrewb]

Will you concede that tuning inductors in an RF environment is within the scope of this hobby?

Can you point to a specific case/game where you've needed to do that, with details?

If you can, sure. Let's hear it.

 

[ArcadeTechGW]

Can you point to a case where you found a bad inductor in a game, and needed to measure the inductance to find a replacement?

Found one smoked at one point. Couldn't find a value, so that was it.

 

[rod90]

Can you point to a specific case/game where you've needed to do that, with details?

If you can, sure. Let's hear it.

You must not follow my postings. I have the remotes for Midway Bullseye and have posted about tuning them.

 

[ArcadeTechGW]

This looks SOOO GOOD.

If he did this for 21 boards, and all of them were good, you could establish a 95% confidence interval that the data represents a valid trend.

But 1 board? Even with multiple caps (unless they all were the same voltage, microfarad, temperature rating), nope. Not statistically significant.

One is a point, two is a line, three is a trend. And that isn't statistically significant. You need a minimum of 21 trials, and then look at the results.

 

[cwilkson]

Looking for any Bronze Age games. Is this you?

Yes, the 80's gave us a lot. But there are many other games not from this era.

Hah! I'm always looking for Bronze Agers!
Sadly, I'm +3 to +5 on capacity right now, so I need 2 Bronze Age projects to move out the door ASAP.
(Wheels II and Monte Carlo)

 

[rod90]

Hah! I'm always looking for Bronze Agers!
Sadly, I'm +3 to +5 on capacity right now, so I need 2 Bronze Age projects to move out the door ASAP.
(Wheels II and Monte Carlo)

So, does this mean that you want to get out of the 70's?

 

[andrewb]

Without data, all you have is an opinion. The late Hans Fauske said that, and he's right.

I have data. Zero Atari game board failures due to Nichicon caps with out of spec capacitance or ESR. Do you want me to put 'zero' in a spreadsheet and post it?

My statements are not based on 'opinion'. They are based on actual experience with real boards, and real business. My activity and reputation here is evidence of the volume of repairs I do. And I've repaired more than 21 Atari game boards.

I've done what Sam did in his video to 100+ boards (testing all caps in-circuit for ESR, because it can be done quickly without removing them), and my observations were consistent with his. I used to check every Atari game board and AR for years, until the data made it clear that I didn't need to keep doing it. And I still spot check the occasional board, to keep the experiment running.

If you aren't willing to accept that, then me going and actually measuring the 50+ Atari boards I have laying around and putting the values in a spreadsheet isn't going to convince you any further. And I'm not going to make a video of myself measuring 50 boards, just for your sake.

If you want to question any of these statements, you are free to do your own experiments and post your own findings. I'm just telling you what I've seen, from years of observation, and the need to perform my own tests to base my business decisions on.

 

[Hoppy2021]

You are all hypocrites! Here ya's are talking about being purists, about restoration projects back to factory. You shun LCD monitors and lights, and praise CRTs. And now you can have Vintage Caps, to keep your game as close to factory as possible, and that's not good enough!! No no, new reproduction parts are ok for the heartbeat of the games we love. but don't install an LCD or sense mod! You people make me sick... A bunch of capacitor ageists!!


( Said purely for fun. It's your game, do what you wish to keep it running reliably! )

 

[ArcadeTechGW]

I have data. Zero Atari game board failures due to Nichicon caps with out of spec capacitance or ESR. Do you want me to put 'zero' in a spreadsheet and post it?

My statements are not based on 'opinion'. They are based on actual experience with real boards, and real business. My activity and reputation here is evidence of the volume of repairs I do. And I've repaired more than 21 Atari game boards.

I've done what Sam did in his video to 100+ boards (testing all caps in-circuit for ESR, because it can be done quickly without removing them), and my observations were consistent with his. I used to check every Atari game board and AR for years, until the data made it clear that I didn't need to keep doing it. And I still spot check the occasional board, to keep the experiment running.

If you aren't willing to accept that, then me going and actually measuring the 50+ Atari boards I have laying around and putting the values in a spreadsheet isn't going to convince you any further. And I'm not going to make a video of myself measuring 50 boards, just for your sake.

If you want to question any of these statements, you are free to do your own experiments and post your own findings. I'm just telling you what I've seen, from years of observation, and the need to perform my own tests to base my business decisions on.

No, that's opinion.

Show us your documentation.

It has been found over time that our views of things change. What was once "I've fixed 20 boards" becomes "I've fixed hundreds of boards."

We round the corners, soften the edges.

So if you have objective evidence, provide it.

The point that you fawned over a SINGLE DATA POINT with a tester and think that provides some "proof" shows you don't understand statistics or probability.

And contrary to what you say, if you can test each capacitor in a lot of 50 (same voltage, microfarads, temperature range, brand) and document it then you will have statistically significant data which is NOT opinion.

After you run a statistical analysis of course.

I have worked on two of these recently for nuclear utilities, in one case, effectively disproving that a vendor's strong recommendation to do some expensive work was in fact, not statistically valid, and a second looking at the preventative maintenance frequencies of a component with a large set of data, and parsing / statistically analyzing the data, determining failure modes from the associated work packages, and using that to determine that the primary cause of the component failures was the testing that was done on them, and that it was acceptable (as there was no time basis to the failures) to extend their PM intervals.

Both of these studies (I signed off on the former after a my comments were incorporated, and created the latter) used large data sets which were analyzed to determine what the data showed us. They had extensive documentation and data sets, all of which was written down, with as found, as left, and failure analysis.

Contrast this with what you are asking people to accept on face value.

1. You have fixed hundreds of boards.
2. In each board (with no business need), you have tested all the capacitors.
3. In this testing, you have found the capacitors were still good.
4. You apparently never found a single bad capacitor.
5. You don't have any records of your readings. We just need to trust you.

Nope. That isn't data. That's opinion.

If you have data, I'd be happy to look at it. I can run a statistical analysis and see what it tells us. I suspect you just looked and didn't write things down. That's fine. Note that until you do, the results are just your opinion.

 

[EVB]

I have a G07 with original caps. Yeah, it could use a kit but barely. If it ain't broke don't fix it.

 

[EVB]

There are many therapeutic outlets in fixing these old games, but I've never in my entire life said "Man, I've had a rough day, I think I should do a cap kit to take the edge off, you know what, let me do this cap kit TWICE!". For $10-15 in caps, I'll save myself the headache lol.

Never done one twice with all the goofy surplus caps I've used. It's a miracle! If I was doing a kit for someone I'd buy a new kit. For myself I don't really care. I'll literally put whatever caps I have laying around it in. Never have I said, wow that ;looks like crap, I should do the caps again. lol!

 

[advans13]

I can say one thing. People have made a killing off "pcb cap kits" lol. Shits ridiculous

 

[Alpha-tron]

Never done one twice with all the goofy surplus caps I've used. It's a miracle! If I was doing a kit for someone I'd buy a new kit. For myself I don't really care. I'll literally put whatever caps I have laying around it in. Never have I said, wow that ;looks like crap, I should do the caps again. lol!

And the klov buyer in me peeks out.. lol

So if I am buying something from your personal collection, I am going to assume you have a goofy surplus caps are in your arcade machines? Does this mean I get a discounted price? Laughs ducks and runs.

 

[EVB]

And the klov buyer in me peeks out.. lol

So if I am buying something from your personal collection, I am going to assume you have a goofy surplus caps are in your arcade machines? Does this mean I get a discounted price? Laughs ducks and runs.

It's possible I suppose, but I'm not selling. I have games that have been running for decades on surplus caps. So what? If I was selling and the buyer looked at it like a cap snob and said "nope, I'd do it but you have all different brands of caps in there from 40 years ago" I'd probably say well it's double the price now anyway.

 

[Alpha-tron]

It's possible I suppose, but I'm not selling. I have games that have been running for decades on surplus caps. So what? If I was selling and the buyer looked at it like a cap snob and said "nope, I'd do it but you have all different brands of caps in there from 40 years ago" I'd probably say well it's double the price now anyway.

Capacitor plague - Wikipedia

en.m.wikipedia.org

If the games in theory can be repaired by just a cap replacement, I don't see an issue.

But.

Games that are not built the same ways as your typical 80's arcade machine.

With security chips and dongles with different interfaces. A working motherboard might be your only option.

Being picky on the repair is important in those situations. Customer are happy when they can drag thier arcade machine home and play them.

There is a reason why I buy some games a state away. Just because the state away person has the technical knowledge to keep those odd game working and alive.

 

[andrewb]

The point that you fawned over a SINGLE DATA POINT with a tester and think that provides some "proof" shows you don't understand statistics or probability.

Oh, for fuck's sake. If you want to start touting professional credentials, I have an EE degree, and have worked in the digital communications field for 25 years, hence my avatar. Yes, I'm familiar with basic probability. If you want, you can read a tutorial I wrote about Bernoulli trials, confidence intervals, and hypothesis testing twenty years ago.

The reason I support the video is because it backs what I *already* know to be the case, because I've *been* studying this for years.

Question 1: Why would it change your mind if provided you a document with numbers in it? If you don't accept what I'm saying now, how would a spreadsheet of numbers make you feel differently?

And contrary to what you say, if you can test each capacitor in a lot of 50 (same voltage, microfarads, temperature range, brand) and document it then you will have statistically significant data which is NOT opinion.

I have a statistically significant data set. That's what makes it NOT opinion.

Opinion is defined as "a view or judgment formed about something, not necessarily based on fact or knowledge." My view IS based on knowledge. Knowledge gained from years of repairing, measuring, and standing behind these boards, which has *included* observation of a large data set of caps.

Contrast this with what you are asking people to accept on face value.

1. You have fixed hundreds of boards.
2. In each board (with no business need), you have tested all the capacitors.
3. In this testing, you have found the capacitors were still good.
4. You apparently never found a single bad capacitor.
5. You don't have any records of your readings. We just need to trust you.

Yes, that's exactly what I'm presenting. (Specifically for Atari game boards and AR's.)

I have also found plenty of bad caps on HV cages, and non-Atari game boards. I can even tell you which caps fail most often on specific boards (because I've seen a significant number of those that HAVE been bad, as well).

I *had* a business need to test all of the caps. Because they were all boards that I was either repairing or selling, with a warranty that I stand behind. Therefore I needed to do some testing to build confidence that the caps could be left original and the board would be reliable. Hence me incorporating a study of this into my process.

I've been publicly stating for years here that it isn't necessary to replace original Atari caps, unless they're physically damaged or missing. I've sold countless game boards and AR's here, and my volume is backed by public data. What do you think my recommendations are based on? Opinion?

Question 2: If my practices were somehow flawed, do you think I would still be in business? Don't you think I'd be having chronic problems with boards failing in the field?

If you have data, I'd be happy to look at it. I can run a statistical analysis and see what it tells us. I suspect you just looked and didn't write things down. That's fine. Note that until you do, the results are just your opinion.

Yes, I didn't record every ESR value. Because that isn't necessary in order to form one type of statistical study.

I checked the ESR values, compared them to the acceptable values you find in typical tables (and/or I would grab a new cap to compare against), and none have shown high ESR. Nor has a single Atari game board or AR failure been due to a bad electrolytic cap. Caps with high ESR are obvious when you see them, because they will measure significantly above the typical value. It's a specific failure mode.

The data comes in the VOLUME of Atari boards I've looked at, and seen none with bad ESR, and no cap-related failures. That *is* data. Specifically it is Bernoulli trial data. The data is represented in binary form. And the volume of it is what gives statistical confidence, particularly when the number of observed failures is zero. See references above regarding Bernoulli trials.

Question 3: If I were to give you a spreadsheet with measured ESR values, what statistical analysis would you perform? Be specific.