Frying An Arm Chip

[WizardStan]

It's a bad thing to point out how, if a given method were used to detect tampering, it could be defeated?

It's a bad thing to point out on a public forum how to break something that the developers haven't had a chance to fix, yes.
The polite course of action is to wait until OpenPandora has made their decision on how to handle, then try to figure out a way around it, and then tell the team privately of your concerns so that they have a chance to fix the problem. If you're in the habit of publicly blurting out faults, you increase the risk of telling someone with less than noble intentions how to circumvent the system.
Of course, this was a hypothetical solution from a user with no connection to OpenPandora at all, so really no.

edit: actually on the specific topic and post, the ellipse suggested there was more to the person's thought: "Oh, I've never done this, but..." but what? But someone else maybe? But you would consider it? Ellipses are dangerous tools that are often wielded incorrectly. If the unspoken words are not blatantly obvious, you risk someone misinterpreting your words (or lack thereof). It's just easier to, you know...
(finish your sentences.)

 

[aq21]

There is one thing that has bothered me.
It has been said that you won't be able to change OPP by hand, that probably means that it will ship set at OPP3.
This means that to reach the advertised 600Mhz you would need to overclock. (And this is outside TI's spec, so there is no guarantee it will reach that high, even if it is higly likely.)
But what really bothers me is that there is no way to change to OPP1 or OPP2, in order to save battery life.

The kernel's cpufreq code changes OPP level according to the frequency in use. They are not fixed, and you do not change them manually.

 

[Dimacus]

^ Ah, that's alright then.

 

[Asmo]

By no means I'm saying I'd do this..

 

[mali]

 

[(naw)mcx]

one of the worse days I've ever had, and you made me laugh.

asmo. you are awesome

 

[mindlord]

In my opinion
No one has been able to submit evidence that overclocking has any short term effects that should concern OpenPandora Ltd concerning the Warranty. I submit that the TI spec sheet consider clocking and volting the OMAP to be "normal" within the limits they specify. TI's spec sheet most likely takes into account the PoP architechture of the OMAP since testing had to occur in their labs with fully assembled OMAPs to get the ranges they indicate.

Moreover, clocking too high leads to a crash/reboot rather than permanent damage to recent ARM chips as evidenced by the abuse some of us put our GP2Xs through. We know the the OMAP has sensors in place to shut the chip down if it leaves "healthy" operating temperatures. Which is what has been said all along by nearly everyone, everywhere, infinty++.

One person said they cooked an ARM processor by both overclocking and overvolting. More than likely WAY far out of spec for the device - which was not a Pandora. The chips may not have had in-built protection against damage from overclocking in the same way the OMAP does. I looked at the chips involved (Intel XScale PXA250 and Intel XScale PXA270), but couldn't determine what kinds of fail-safes they possess. Nor do we know of any other extenuating circumstances which may have been involved in their untimely deaths.

As far as I know, and as deeply as I've researched on the interwebs (I put a couple hours into it) frying ARM CPU's seems to be a very rare thing. All of this turnmoil is from one person's experiece. Can anyone else step forward with evidence of permanently killing an ARM chip though overclocking?

 

[darkblu]

The question has never been whether or not the warranty should guarantee functionality at an overclock, obviously it shouldn't. The question is whether or not the warranty should be voided if you do overclock. 600MHz should not be considered an overclock based on Open Pandora's product specifications, so it's on them to accommodate this (and yes, any failure to do so should be covered by RMA). What I was trying to say is that if random crashes are dangerous enough to void the warranty then we have a serious problem, because any program can potentially randomly crash without first being overclocked.

well, i was referring to random crashes that can be attributed to some state of overdrive (as in 'software X crashes at 600@OPP3, but does not at 500@OPP3'). yes, i'm aware that even that can be a software problem, but there are conditions where the source of the issue can be positively attributed to host instability. also, 'based on the Open Pandora's product specifications' means little here - OP don't produce the SoC. the actual chip vendor dubs OPP5 as a state of overdrive, one which is disallowed on the mil-grade parts, and clearly marked as lifespan-detrimental (by a rate of ~2). for all practical purposes OPP5 is a state of maker-sanctioned overclock & overvolt. you can surely see the forming conflict of positions here if the end user ends up not allowed to do what OP should by your view.

If OP has to pick between OPP5 at 600MHz and decreased life span vs OPP3 at 600MHz and instability then I would think they'd have to go with the former. They promised 600MHz, which implies that the thing performs at that. They didn't promise anything about power draw or lifetime outside of "will work for a year", and can feel free to defer to the TI stated usage for that.

my original point, which i did not quite convey in my previous post, was that people tend to have issues with devices that last their 'full-service' warranty span and then die abruptly - this is usually considered bad production quality.

So where do you draw the line exactly? We're talking about an operating range up to 90C; that's what the power on hours are speced over. I have a hard time imagining that ambient temperatures up to 40C at the most would do an awful lot to tip it over the breaking point, and I think it'd be more of a matter of it failing to work at all past this point rather than having a diminished life span. I doubt any kind of feasible environmental conditions will do much to reduce the operating hours of the chip.

actually, an ambient t of 40C can do a awful lot if it could tip the temperature at junction above the 90C mark. don't forget we are talking of a passively-cooled, tight closure device. it does not even have a heat spreader, as far as i'm aware. again, POH says little here - it gives no idea of the utilization of the chips per se, just temperatures (though it does suggest the chip is under load). hypothetically, the chip could reach 90C @ junction while idling, given a sufficiently high ambient t. so if the t@junction could be brought to 75C by an extra heavy crunch at an ambient t of 25C, 40C ambient could be your pandora's private hell.

 

[Aninhumer]

While it's somewhat of a cop-out, how about OpenPandora only have an official warranty for very conservative usage, but allow other replacements at their discretion?
I don't know how that works legally, but I'd trust Craig or EvilDragon to be fair about whether someone just had a broken chip or whether they overvolted (if indeed that can be determined at all? If it can't the argument is mostly meaningless)
Maybe I'm just being too trusting?

 

[mvickers03]

Personally I don't understand why this bothers some people! I know it was advertised as being OC'able because it is, but whats the problem if it voids your warranty FUCK warranty you wanna have as much fun with this as possible push it too its limits make it do things its not supposed to do and don't be held back by a warranty. If your that scared get 3rd party insurance or something

For me, if I broke my Panda of course I'd be gutted but hey ill just beg OP to sell me a MB and fit a new one. pretty much every console I ever bought had the warranty voided in the first week. BTW I'm not made of money either I have to work very hard for my money but it's a risk I would be willing to take and definitely will do..

Of course I do understand people that are bothered about warranty but then you will have to live with not pushing your hardware to the extremes..

 

[darkblu]

i guess i should clarify my position here. i'm not so much bothered by the warranty/overclocking prospects - warranties just need to be reasonable, and people can knock themselves out with overclocking (likely outside of said warranty). and re the warranty, as long as it allows the user to freely go across OPP1-5, ergo the original promise of 600MHz is met, i don't see much of a problem. what i would see a problem with is if:

a ) the device comes out of the box overclocked at, say, something like 600MHz/OPP3
b ) the default (if there's such) OPP - the one reverted to at reboot, ect, is above the 'nominal' OPP3.

the first point is something, i, as a potential sw developer for the platform, don't want to see due to the implications it could have for sw development- software cannot rely on out-of-spec conditions for its normal operation - that's a whole can of worms there.

the second point is merely about the reliability and expected lifespan. pandora is a pocket 'personal computing' device. it will be (ab)used in god knows what unfavorable conditions (at least i know i will). i don't want it to last me anything less than 5 years, and i don't want to suffer random instability just because there's no safety margin left to the chips in there. otherwise, i too, will try to get a spare PCB the moment those are available.


ps: here's something unrelated to the discussion but darn interesting, which i hope makes up for any unrest i may have caused: http://queue.acm.org/detail.cfm?id=1716385