Author Topic: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles  (Read 19735 times)

0 Members and 1 Guest are viewing this topic.

Offline Mister Xiado

  • Newbie
  • *
  • Posts: 49
  • Karma: +0/-0
    • View Profile
https://db-electronics.ca/2017/07/05/the-dangers-of-3-3v-flash-in-retro-consoles/

René has performed a pretty detailed rundown of voltage issues, and while I have enough Genesis systems to handle one going belly up, but not everybody does. Does this put a new revision of most ED models in the design pipeline?

Offline jjennings089

  • Newbie
  • *
  • Posts: 2
  • Karma: +0/-0
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #1 on: July 06, 2017, 06:38 AM »
or better yet a voltage step down adapter for those of us who have multiple versions of Krikzz products.

Offline librajo

  • Newbie
  • *
  • Posts: 2
  • Karma: +0/-0
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #2 on: July 06, 2017, 06:46 AM »
Thanks for posting this, I came to see if anyone else had reported it. Kind of a bummer that using these carts could potentially kill your console.

Offline BuckoA51

  • Newbie
  • *
  • Posts: 7
  • Karma: +0/-0
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #3 on: July 06, 2017, 11:58 AM »
Would be good to get some comments from krikzz on this, I'd be willing to change my affected Everdrives. I don't believe it's his fault even the greatest hardware engineers in the world can make honest mistakes.

Offline cleeg

  • Newbie
  • *
  • Posts: 45
  • Karma: +2/-0
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #4 on: July 06, 2017, 08:57 PM »
I agree, I've just headed over here from Sega-16 after posting the same thing.  I'd be willing to drop on a new MEDx5 which is a shame as I have the original model that plays every ROM I've ever thrown at it.  I also have an early model of the original ED too.
The same cleeg from Sega 16

Offline marionintendo

  • Newbie
  • *
  • Posts: 22
  • Karma: +1/-0
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #5 on: November 20, 2017, 01:58 AM »
https://db-electronics.ca/2017/07/05/the-dangers-of-3-3v-flash-in-retro-consoles/

René has performed a pretty detailed rundown of voltage issues, and while I have enough Genesis systems to handle one going belly up, but not everybody does. Does this put a new revision of most ED models in the design pipeline?

Krikzz replied.

http://krikzz.com/forum/index.php?topic=6614.0

Sorry for the old bump, I came from Google. The post you linked was either deleted our posted in a restricted thread. Do you remember what krikkz said?

Offline FeverDrive

  • Hero Member
  • *****
  • Posts: 1299
  • Karma: +54/-0
  • Blast Processing advocate
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #6 on: November 20, 2017, 06:59 AM »
Sorry for the old bump, I came from Google. The post you linked was either deleted our posted in a restricted thread. Do you remember what krikkz said?

Krikzz said that there is nothing to worry about, the consoles aren't going to be damaged or their lifespan reduced.
That thread got deleted because of trolling if I remember correctly.
WELCO
METOT
HENEX
TLEVEL

Offline KRIKzz

  • Administrator
  • Hero Member
  • *****
  • Posts: 3378
  • Karma: +508/-2
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #7 on: November 20, 2017, 09:21 PM »
Original topic about this question were accidentally removed, but sense of this topic it two words sound like that: Rene's sensational "research" were based on wrong numbers, by some reason he decide that CMOS device current limits is 5.2mA per pin, but JEDEC specification clearly says that CMOS device current rating is 20mA. Old everdrive carts without level translators has around 7mA load per pin, and this is notably less than max acceptable 20mA, as we can see.
I say it to Rene in comments to his article, but he just ignored it. Anyone can read my comment in his article here: https://db-electronics.ca/2017/07/13/ac-analysis-of-clamping-diode-effect-in-5v-to-3-3v-mismatch/

Offline Bratwurst

  • Newbie
  • *
  • Posts: 6
  • Karma: +0/-0
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #8 on: December 12, 2017, 03:06 AM »
I recently acquired a Mega Everdrive X7 and decided to make a detailed measurement log of IC temperatures running an official Sega manufactured game cartridge, then the following day while running a rom of the same title on the Everdrive. With so said voltage shifting through just a resistor array the shunted voltage should be turned into waste heat, if there is going to be damage from that I would expect an appreciable difference, especially cumulative over time.

The Genesis was an NTSC VA4 MK1 with the top removed, regulated power supplied as 10 volts DC, indoor ambient temperatures both days essentially 23C. I chose NFL 95' because the demo/attract mode appears to be performing random matchup calculations on the fly as I never observed a repeat play/result while I let the system run, which is a good simulation of human usage as opposed to an endless loop of title screen and 15 second attract like in Sonic 1 as an example. I ran the official cartridge on the first day starting at noon EST, then the Everdrive on the following day noon EST.

Note that the circuit board in the pictures is PAL but it's only for visual reference of each chip's measurement.

https://imgur.com/a/26oiS

Running standard 2MB Genesis cartridge: NFL '95: 10 volts supplied, consistently 0.74 amps drawn

Running Mega Everdrive 7: NFL '95: 10 volts supplied, initially 0.84 to 0.81 amps drawn, then steady 0.80 amps running the game's attract mode.

I can't claim to be an electrical engineer but just over a 2 hour period for either arrangement (which I would assert is within the average play session for most people these days) there really wasn't much difference, maybe 2C higher for some chips, and in some instances the Everdrive overall ran the Genesis cooler by a few degrees C than the original cartridge. We see the CPU get up to 36C at one point for the one hour mark with the Everdrive but that was it, conceivably a fluke of the IR thermometer since it dropped even lower by the following hour. Should be notable that the Motorola 68000 has a temp ceiling of like, 70C-85C.

Interestingly, with the Everdrive the headphone sound amp did run hotter by 4C over the stock cartridge even though the stereo jack was not used in either case.

Offline Majestic_Lizard

  • Full Member
  • ***
  • Posts: 147
  • Karma: +17/-58
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #9 on: July 25, 2018, 08:24 AM »
I realize this is an old thread and I may get the boot for bumping it, but there were a few things that bothered me about the article. You don't have to be an engineer to see the holes in the claim made by the author, you just have to have a little common sense.

First, the author stated that various carts had damaged some of his friend's consoles. However, he offered no physical evidence of this and no concrete examples. This is deceptive. It is not enough to claim that something is theoretically possible. This type of claim requires physical proof.

Second, he condescendingly implied that Everdrive designers were not knowledgeable of elements of electrical engineering and physics. Somehow, I doubt that is the case.

Third, his "common rebuttals" section  did not contain rebuttals of serious criticisms, but instead contained rebuttals to infantile straw-men arguments. Serious criticisms, like those of Krikzz, were  ignored.

Fourth, the idea that he did not have a selfish motive for creating a misleading article due to the fact that he doesn't sell flash drives is not correct. He doesn't have to be selling a competing product to benefit from increased traffic. Any product at all benefits from increased traffic.

Finally, people are saying that science substantiates the author's claim, simply because it sounds logical. While I don't doubt that there is a grain of truth to what the author is claiming, the fact is that he did not offer a single concrete example of a console damaged in the manner he suggested. That is the very first thing he would need to start with.

My feeling is that these older consoles were designed with greater tolerance than the author is representing and he over looked variables that mitigate the "design fault" he describes. Otherwise, we would have mountains of dead consoles and thread upon thread of people bitching about flash devices wrecking their consoles.

As old as these consoles are, common sense would point to the notion that they are more likely to simply disintegrate with age than blow up because of a discrepancy of slightly more than 1.5 volts.

Offline EverDriver

  • Administrator
  • Hero Member
  • *****
  • Posts: 547
  • Karma: +42/-0
  • support
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #10 on: July 25, 2018, 02:11 PM »
Thanks Majestic_Lizard !

The total absence of even single damaged console after long years can't stop our customers asking about it from time to time...
That's a psychological thing. Some people prefer to trust his lie than just turn their common sense on and check the facts. "Maybe he was right. I don't want to risk. I don't have time and skills to check it." and so on.

The most interesting and ironical is that it seems the author already understood his mistake but something still prevents him to admit it.
He just prefers to ignore the facts. Maybe he thinks he would lose his reputation if he admits the mistake.
If one makes a serious accusation and then it turns out that the accusation was based on his own mistakes it means for him that he loses his own reputation and he doesn't want it to happen. Nobody will read the articles of a liar.
And he found a compromise solution: to ignore the facts.

But it's not a behaviour of a responsible person. It's a behaviour of a weak person.
If you made a mistake then admit it, learn it not to repeat it in the future and move on.
It's sad that he damaged our reputation a bit and doesn't want to be responsible for that.
Fortunately the majority of our customers have common sense and don't care about his lie.


Offline leonquest

  • Sr. Member
  • ****
  • Posts: 456
  • Karma: +22/-0
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #11 on: July 25, 2018, 10:36 PM »
I generally hate internet drama, but some of your points should be addressed anyway, as it's clear you're not an engineer and you don't know the full story.

I'm not going to bother with every single paragraph, so I'm just going to go with the ones where you didn't do your research.

Second, he condescendingly implied that Everdrive designers were not knowledgeable of elements of electrical engineering and physics. Somehow, I doubt that is the case.

I must have missed the condescending part. Regardless, according to Krikzz's BIO, he is not an electrical engineer and thus learned all of this on his own, so I wouldn't be surprised if he oversaw some design flaws along the way. It happens to large manufacturing companies with specialized engineers and hundreds of employees, why wouldn't this happen to Krikzz? nobody is perfect, just accept krikzz has gotten better and better at his craft throughout the years.

I know for example, starting the Everdrive N8, Krikzz added voltage level translators, (although he specified on twitter that he did so for different reasons than DB electronics claimed). This means that everdrives from 2013 and on forward do not have the voltage issue that older everdrives might have.

Aside from that, Kevtris stated that he thinks the voltage issue will affect the everdrives themselves in the long run, and not the consoles they're used in, but he didn't go into technical detail as to why though, so your guess is as good as mine.

Third, his "common rebuttals" section  did not contain rebuttals of serious criticisms, but instead contained rebuttals to infantile straw-men arguments. Serious criticisms, like those of Krikzz, were  ignored.

This is where you could have done your research better. Krikzz and Rene had quite a lengthy technical discussion on twitter, and neither could agree with the other: https://twitter.com/krikzz/status/969011006142894080


Fourth, the idea that he did not have a selfish motive for creating a misleading article due to the fact that he doesn't sell flash drives is not correct. He doesn't have to be selling a competing product to benefit from increased traffic. Any product at all benefits from increased traffic.

*He does not have selfish motives, he is an everdrive user. He is an electrical engineer and what he talked about came from scientific observation and research. The accuracy of his conclusion is another matter though, as kevtris' statement made me think there's more to it than Rene concluded.

*His site does not benefit from ad money, he has a day job, so your site traffic argument is a moot point. His products will not benefit more because he sells niche stuff for a niche audience within a niche audience: Turbografx and PC Engine stuff. Again, you need to prove he did for this reason if you are going as far as to accuse him of such "an elaborate lie" just so he can sell his electronics stuff that only a small niche audience could be interested in. As a matter of fact he recently mentioned he will be leaving his retail business to dedicate to R&D (as he doesn't really make a lot of money from it anyways), which rather proves he is not in it for the profit. To add to this point, he has shared several designs with the community for free, which furthers the point I'm trying to make.

*Your accusation also implies that Krikzz's products are somehow above criticism, and this is a fallacy. His products are great, but criticism is always needed in order to improve.

Finally, people are saying that science substantiates the author's claim, simply because it sounds logical. While I don't doubt that there is a grain of truth to what the author is claiming, the fact is that he did not offer a single concrete example of a console damaged in the manner he suggested. That is the very first thing he would need to start with.

It's hard to give a concrete example when we would need to test several hundred thousand consoles in order to prove his point, who has the money or time to do that? the fact of the matter is that he used science to raise a concern over how we are designing modern solutions for old devices, since we intend to keep them a long time. Also testing would be really complicated as no one can really guess how much of the console lifespan is substracted from using cartridge with the wrong specs. Just know that it is not a good idea to use things out of spec with your old consoles and we should strive to do a better product, which krikzz has since 2013.

My feeling is that these older consoles were designed with greater tolerance than the author is representing and he over looked variables that mitigate the "design fault" he describes.

This was Krikzz' argument on twitter, I'm leaning towards this argument too, however, it's always a good idea to 100% within spec to avoid issues, as krikzz himself discovered when he added voltage level translators (reduce EMI and increase device stability at hi clocks).

Otherwise, we would have mountains of dead consoles and thread upon thread of people bitching about flash devices wrecking their consoles.

Rene never said the devices would die instantly, in fact no one ever said that. Constant use through the years might inevitably wear out the consoles a lot faster than using carts within spec, but again, this is still debatable as both sides have valid points.

As old as these consoles are, common sense would point to the notion that they are more likely to simply disintegrate with age than blow up because of a discrepancy of slightly more than 1.5 volts.

Again, consoles will inevitably die with age, but using things out of spec could mean this age is reduced. Please point me where rene says the consoles will blow up?
Everdrive64 V3 - SD2Snes rev. f - Everdrive N8 fami - MegaEd X3 - PS IO - MODE - MISTer

Offline leonquest

  • Sr. Member
  • ****
  • Posts: 456
  • Karma: +22/-0
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #12 on: July 25, 2018, 10:50 PM »
Thanks Majestic_Lizard !
The most interesting and ironical is that it seems the author already understood his mistake but something still prevents him to admit it.
He just prefers to ignore the facts. Maybe he thinks he would lose his reputation if he admits the mistake.
If one makes a serious accusation and then it turns out that the accusation was based on his own mistakes it means for him that he loses his own reputation and he doesn't want it to happen. Nobody will read the articles of a liar.
And he found a compromise solution: to ignore the facts.

This is false until you prove it, neither Rene nor Igor have acknowledge each other's points. In any case further discussion is pointless as this issue only affects carts designed prior to the everdrive N8, which are being dished out anyways now that we have a new master everdrive, new GB everdrives, new mega everdrives, etc. To add to that, older everdrive designs used flash instead of ram, which sucks, so eventually these older designs will end up being discarded in my opinion.

Thanks Majestic_Lizard !
But it's not a behaviour of a responsible person. It's a behaviour of a weak person.
If you made a mistake then admit it, learn it not to repeat it in the future and move on.
It's sad that he damaged our reputation a bit and doesn't want to be responsible for that.
Fortunately the majority of our customers have common sense and don't care about his lie.

Again this is false, because neither party has admitted to anything. Both have valid points, and the only way to solve this is to have a third opinion, who further investigates this. Iam8bit for example, made an interesting warning label when they designed the Street Fighter 2 re-release cartridges, perhaps they are aware of the voltage issue.
Everdrive64 V3 - SD2Snes rev. f - Everdrive N8 fami - MegaEd X3 - PS IO - MODE - MISTer

Offline EverDriver

  • Administrator
  • Hero Member
  • *****
  • Posts: 547
  • Karma: +42/-0
  • support
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #13 on: July 26, 2018, 12:58 AM »
Hi leonquest!

Just two questions:
1) Where are the damaged consoles (ok at least one. after long years and a lot of sold EverDrives there should be at least one, right?)
2) Why he ignores the facts (the official datasheet information)? If one is sure that he is right he just says: look here in the official documentation it's written that the hardware has such normal working ranges. And that's it. No questions anymore! But what reason to ignore the facts and pretend that you haven't noticed it?

Thanks!
« Last Edit: July 26, 2018, 01:19 AM by EverDriver »

Offline leonquest

  • Sr. Member
  • ****
  • Posts: 456
  • Karma: +22/-0
    • View Profile
Re: [ARTICLE] - The Dangers of 3.3V Flash in Retro Consoles
« Reply #14 on: July 26, 2018, 03:24 AM »
Hi leonquest!

Just two questions:
1) Where are the damaged consoles (ok at least one. after long years and a lot of sold EverDrives there should be at least one, right?)
2) Why he ignores the facts (the official datasheet information)? If one is sure that he is right he just says: look here in the official documentation it's written that the hardware has such normal working ranges. And that's it. No questions anymore! But what reason to ignore the facts and pretend that you haven't noticed it?

Thanks!

Hi Everdriver,

I didn't mean to start an argument with you seeing as you are the forum admin and all, but I felt this needed to be said.

I would also like to add to this friendly conversation that I am a fan of both Krikzz's work and DB Electronic's (just look at my signature, I have 3 everdrives and the SD2Snes), and as I've said before, both sides have valid points: Rene explains the science behind his claims, investigation, observation, measuring, law of physics, etc, while Krikzz points to spec documents that indicate the extra tolerance that old consoles are supposed to have.

First, let's get this out of the way. If you check the article thoroughly, and then the current store, you will come to the realization that there are only 3 everdrives considered by DB as "unsafe":

++++ Everdrives without voltage level translators:
- Mega Everdrive X7
- Super everdrive V2
- EverDrive-GG

+++++Everything else on the store is safe, as all others are post N8 designs or redesigns with VLTs like TED V2 or the master everdrive x7.

Now, the problem with the Twitter discussion that I shared between the 2, is that there was too much animosity and the discussion ended abruptly, so I could not see a reply from Rene as to Krikzz's docs, and such. Neither side wanted acknowledge each other's points and just kept insisting the other was wrong, so this leaves me in the same uncertainty that I've had for the past year or so since the article came out, which is "what is the real danger of using devices out of spec?, and based on Krikzz's argument, are his first everdrive models really out of spec?". This is why I argue we need a third opinion from someone neutral that can really look into both arguments and tell us what we need to expect from the everdrives that work "out of spec". I don't think he ignored the facts, I think he refused to continue the argument as Krikzz was just unwilling to acknowledge his responses. Krikzz must have taken Rene's article it very personal judging from his replies on the forums and on twitter.

On to the topic of damaged consoles, well, that's where it gets messy. If a console fails tomorrow while using an older model of an everdrive, can you prove that it really failed due to the everdrive? Lets be realistic: Rene does not have 100.000 consoles laying around with 100.000 everdrives pre-N8 to test with over 100.000 hours at a time in order to get statistical data. I think that the whole point of the article was more of a consideration to the possibility of "due to the law of physics, using out of spec voltage means that consoles which would normally last X amount of hours of use will now last X-Y hours of use" (X being the theoretical console's lifespan in hours, and Y being the amount of hours that would be substracted from the use of out of spec devices). So in that sense, no proof is needed since the possibility of life-reduction is there, based on his research alone, and without taking into account Krikzz's main point of extra tolerance.

Hopefully I was clear enough with my arguments above, and I would also like to reiterate that I do not wish to provoke any animosity here, my intent is to put things into perspective instead of just demonizing Rene from DB electronics, just because you disagree with him.
Everdrive64 V3 - SD2Snes rev. f - Everdrive N8 fami - MegaEd X3 - PS IO - MODE - MISTer