Finally! Some actual, novel tech: Apple patent to revive geriatric gear Apple has received a patent for a method of detecting if an electronic device's operating parameters are being adversely affected by aging circuitry, and if so changing those parameters remotely through adjustments to its firmware. US Patent 8,671,170, "Modifying operating parameters of a device based on aging information," …
We used to undo two Dzus fasteners, remove the 'handbag amplifier', lick fingers and remove a valve(potentially hot) or two, replace with new stock and replace.
Add a tag to handbag showing date the valves were replaced.
Rinse and repeat in 'n' months.
No need for a stinking patent.
You see, it compares its output against a nice steady reference (zener diodes are popular, but not the only references out there) and adjusts its feedback loop accordingly. The power supply in your computer probably already does this, as it wants to keep its voltages constant. Now if you say the reference drifts, and you need to measure that with an nice A-D converter, everyone should note that the A-D converter ALSO needs a reference or its readings will be out of kilter. The assumption here is that somewhere there is a reference that isn't subject to aging that can measure things.
Sorry, but they really don't exist. This is why we have national standards bodies (NIST here in the USA) where you can send things to be "calibrated" against nice "traceable" things that are kept under (many) locks & keys.
What we settle for is a "pretty good" reference in a nicely controlled feedback loop that isn't subject to drift (it has things designed to null out bad things).
Sure things "wear out", but if the device was designed correctly (and not for planned obsolescence) the design will take care of it and adapt. So, while this may be a nice thing to show your friends as something "special", it doesn't amount to much.
So, Apple, show me how this is used!
It could be used to put a touch of crackle into Siri's voice along with dropping the gain on the mic so you have to talk louder for Siri to "hear" you. All of this acts as a reminder that the current shiny is getting old and dull and you should drop another paycheck or two on the latest shiny.
Used to keep the life of your shiny extended for an extra year or two so you don't need to rush out and buy a new one... Hang on... why does it make any sense for a tech company to prolong the life of shiny and extend the product refresh cycle... unless... by owning the patent they can stop the technology being used and thus keep the wheels turning. Or they can use cheaper components with poor ageing profiles and compensate for it...
>Hang on... why does it make any sense for a tech company to prolong the life of shiny and extend the product refresh cycle.. [?]
Because some products are sold as being more reliable than rivals. There have been several studies, using different methodologies, that suggest that Toshiba, Lenovo and Apple laptops are more reliable than some other brands. (My Dell is soldiering on handsomely. Sample size = 1, though!). Reliability is a factor that some people take into account when they purchase equipment, especially if they intend to resell it a few years down the line.
In addition, it costs a company money if a device fails within its guarantee period.
OK first off II admit it I haven't read the patent but it doesn't normally help as I can't quite work out what exactly they are patenting wrapped up in all the obfuscation,banality and referenced prior art.
Anyway, good smoke detectors can detect any aging effects in its LED by increasing the current supplied. This can also be used to overcome contaminant build up. The detector can then signal it needs cleaning or reach end of life back to the panel, the panel could have remote management to inform the maintenance people as well.
Rf circuitry requiring tight tolerance crystals may also apply compensation based on the aging of crystals used in the system.
I should think a lot of what NASA do with their rovers also has similar capabilities.
Of course they seem to be adding that the changes applied are orchestrated by a remote computer in some way but that just doesn't seem novel enough to me, but then I'm not a patent layer. It just seems like feedback to me.
At my last place of work I prototyped high-resolution, high-frequency load monitoring on DC Brushless motors for exactly this reason- you could determine worn bearings or wearing out electronics and react accordingly.
It worked but was decided to not be worth the complexity and expense.
That's not the only problem. I can't install any apps on my iPod touch because iOS is out of date. Apple won't let me install a newer one, and made the devs remove all apps that worked with the old one. Worse; the last restore I did wouldn't re-install some apps because that version was no longer on the app store, so I'm hoping that I never have to do another one. So try to convince me what Apple would really do it they could automatically detect that my kit it old?
Plasma and CRT televisions have been doing this since at least the 1990s. Digital scanners have always compensated for aging lamps, sensors, and optics by scanning a blank area outside the page. Digital cameras take dark frame photos (with or without integration time) to compensate for an aging sensor.
Er, that's the case with most most ECUs. For that reason, the ECU firmware is specific to the vehicle, not the model.There, making a modification to it requires the mechanic clone the firmware, alter it, and then re-flash the ECU.
But prior art? Read the penultimate paragraph of the article.
The concept of giving aging circuitry a helping hand dates back to WW2 and possibly beyond. Lots of valve based equipment (including many TV receivers) had the mains supply go through a wirewound resistor tapped at various points to allow for connection to 240/220/200/etc. volt supplies. When one of these was getting a bit long in the tooth and giving a dim, washed out picture, the cure was to move the tap down a notch. Yes, it shortened the life expectancy, but at least the TV was useable for a few more months. Important when a TV set was several weeks if not months income for most.
As it's an Apple patent mentioned on the Reg, it must be bogus.Cue Daily Express-style rant about how we used to do exactly the same thing, when I were a lad back in the Bronze Age, with a lump of amber and a sharpened stick. Now if I could just remember where I've put me teeth and me Werther's Originals...
it *may* have been Sinclair, but didn't someone in the early 80s talk about fabbing chips with logic gates at the junctions, so you wouldn't need to test each individual IC (which in those days had a fail rate of 80%) but instead ran some code which inspected the ICs and simply blew the links to the bad ones ? The idea being you could stack loads of wafers to save on costs somewhere.
Always sounded neat, but don't recall it being used ?
"it *may* have been Sinclair, but didn't someone in the early 80s talk about fabbing chips with logic gates at the junctions,"
I think that the idea of blowing fuses to bypass/fix busted sub-systems was already common place in the early 90s at chip level.
Clive Sinclair was pushing the idea of packaging up the entire wafer instead of chopping it into bits, he would have needed to apply that IC level fuse trick to wafer scale to have any hope of producing a working product.
Since you can't replace your batteries... you want them to last longer. I'm looking at an Iphony right now.
But I see so many ways it can go wrong(tm). Assume a circuit is increasing V to a transistor, because it aged... then you replace this transistor. New transistor, with higher V fries instantly.
Actually it goes well with their walled-garden policy, because when THEY repair the aging transistor, they recalibrate the compensation circuit to fresh-new, lower V, or it would otherwise blow up in your unauthorized face... Unless you try the lower value of V every time you want to fire that transistor... (or after every boot).
On the other hand, every car that runs on any mix of gasoline and etanol already does that... it samples the fumes of the first few strokes and corrects the timing of the spark plugs on the fly, every motherboard has some sort of fine voltage regulation for the overclockers out there, and every PSU has to sample the mains current from 110V to 220V to deliver clean 12 Vcc to the rails.
It is called closed loop control, and there is an entire Engineering career based on it. Google Katsuhiko Ogata "Modern Control Engineering".
There is NOTHING NEW done by Apple. I cry previous artwork.
They COULD release a paper called "Closed Loop Control Applied To Microcircuits To Improve Reliabilty Against Aging" though. It would be a nice Graduation Thesis theme.
OK, I have not read the entire patent but some of this is already done with Smart Reflex in OMAP processors, which track and compensate for PVT variations (PVT = process, voltage & temperature). Inherent in some of the classes is accounting for variations over time due to changes in performance.
As brought up by others how is the watchman chip/controller going to be rendered free of similar aging? Many of the fault tolerant chip Ideas go back to the late 80`s and ( takes deep breath and ducks) Ivor Catts wafer scale integration stuff, featured in Wireless World articles and patented.
Without some real innovation, such as new never used before detection mechanism this is very shaky indeed because of all the prior examples of principle, some already aired here.
The low standard of rigour that seems now to be applied to patents only makes money for shiny suited legals who haven`t a scoobie about the tech aspects.
One of the best flags is power supply ripple and in digital, or mixed circuity ripple signature change, A simple dedicated controller with an ADC <teehee> and NVRAM looks at the output of filters on the supply rails and compares them to the system commissioning values. This is not patentable as it was contained in in an appnote for a controller 25 years ago when ADC and NVRAM onboard had wow factor.
What they need to devise and patent is a duff joint detector. Since the environment has been befriended by unleaded solder, this is the leading cause of failure IME. Temp cycled boards eventually give up. I have actually seen SMD components drop off when even the assembly glue has failed. close examination of the component underside shows a sea of flux with a a few tiny fronds of metal that made contact on factory test. This is in your phone, your laptop, car`s ECU.....
This will happen before any of the rest IME.
A fluxpen and an Iron with the correct tip on it will reconnect components sitting on the assembly glue with hairline cracks in the alloy beneath the solder pads and no contact.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
