On Mon, 8 Jan 2018 10:03:50 -0500, bitrex <bitrex@de.lete.earthlink.net> wrote:>On some of the inexpensive Dell laptops they implement a >hardware-lockout system intended to prevent you from using aftermarket >charger bricks to charge the battery. The way they do this is there's a >chip in the Dell charger that contains a unique identifier which a chip >on the laptop mobo requests from it on startup over a one-wire interface. > >If on boot the BIOS can't confirm that the lockout chip on the mobo has >asserted the "OEM OK" pin or whatever the BIOS then proceeds to muck >around with the processor MSRs (model-specific registers) and alter >their state, specifically the BD_PROCHOT register, which is a flag bit >that is usually controlled by the motherboard temperature sensor. > >Flipping that makes the processor think the motherboard is overheating, >which disables battery charging and has the additional nice "feature" of >throttling the processor down to around 400MHz from 2.8 GHz or whatever. >So not only can you not charge the battery you can't really use the >laptop on an aftermarket brick even when plugged into the wall. Also >there's then no way to tell if the laptop mobo is _actually_ overheating >in that state. > >Good news is that the MSRs are accessible via software, so it's not hard >to write a script that runs after the OS start to flip the bit back, and >everything is back to normal.Excellent and interesting info on the flag. I knew there was something that told the laptop to charge or not quite a while ago. I wasn't sure if there was a "signal" in that 3rd line into the charging jack OR just a simple voltage that the motherboard reads. I looked into the broken charger and did not right away see anything that obviously send a signal or squarewave or anything. Next, I tried a cut charging cable to monitor the line but of course, the charger I used was broken anyway so no luck there. I have so many extra chargers now, I just haven't cared that much to dig into it again. boB
Shady Dell laptop practices
Started by ●January 8, 2018
Reply by ●January 8, 20182018-01-08
Reply by ●January 8, 20182018-01-08
"bitrex" <bitrex@de.lete.earthlink.net> wrote in message news:auP4C.213865$b_.118286@fx23.iad...> It's superfluous from a battery-charging safety perspective; the laptop's > charge management system/IC is perfectly capable of handling an > under-spec aftermarket charger being plugged in. It'll have undervoltage > lock-out and likely can sense the brick's output impedance well enough to > crowbar the charging current to a safe value if the input rail starts > sagging. It won't be the laptop itself that catches fire, in any case.What about overvoltage and reverse? The input circuits I've seen, it's more or less straight into the MOSFETs. Maybe a ferrite bead, but that doesn't do jack, it saturates quickly during input transient (and maybe the ceramic caps too, if allowed to overshoot) and in operation. No TVS. It's a very real possibility. I had one mobo get nuked because of a frayed power supply. What happened is the coaxial cable (they're constructed with two braid layers, carrying current, with a data line in the center) got frayed at the PSU's strain relief, and was shorting out intermittently. Somehow or another, this caused enough stress to blow up a couple parts (presumably there was a general power surge that took out whichever chips broke down soonest). The bizarre thing is, the most burnt chip was an LM393 near the battery controller. How, and why, and also why?... Possibly, if there had been a TVS diode on the input, that wouldn't have happened. Tim -- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: https://www.seventransistorlabs.com/
Reply by ●January 8, 20182018-01-08
Den tirsdag den 9. januar 2018 kl. 02.05.25 UTC+1 skrev Tim Williams:> "bitrex" <bitrex@de.lete.earthlink.net> wrote in message > news:auP4C.213865$b_.118286@fx23.iad... > > It's superfluous from a battery-charging safety perspective; the laptop's > > charge management system/IC is perfectly capable of handling an > > under-spec aftermarket charger being plugged in. It'll have undervoltage > > lock-out and likely can sense the brick's output impedance well enough to > > crowbar the charging current to a safe value if the input rail starts > > sagging. It won't be the laptop itself that catches fire, in any case. > > What about overvoltage and reverse? > > The input circuits I've seen, it's more or less straight into the MOSFETs. > Maybe a ferrite bead, but that doesn't do jack, it saturates quickly during > input transient (and maybe the ceramic caps too, if allowed to overshoot) > and in operation. No TVS. > > It's a very real possibility. I had one mobo get nuked because of a frayed > power supply. What happened is the coaxial cable (they're constructed with > two braid layers, carrying current, with a data line in the center) got > frayed at the PSU's strain relief, and was shorting out intermittently. > Somehow or another, this caused enough stress to blow up a couple parts > (presumably there was a general power surge that took out whichever chips > broke down soonest). > > The bizarre thing is, the most burnt chip was an LM393 near the battery > controller. How, and why, and also why?... > > Possibly, if there had been a TVS diode on the input, that wouldn't have > happened. >don't think it was ~20V into the data line that killed it?
Reply by ●January 8, 20182018-01-08
On 01/08/2018 08:05 PM, Tim Williams wrote:> "bitrex" <bitrex@de.lete.earthlink.net> wrote in message > news:auP4C.213865$b_.118286@fx23.iad... >> It's superfluous from a battery-charging safety perspective; the >> laptop's charge management system/IC is perfectly capable of handling an >> under-spec aftermarket charger being plugged in. It'll have undervoltage >> lock-out and likely can sense the brick's output impedance well enough >> to crowbar the charging current to a safe value if the input rail starts >> sagging. It won't be the laptop itself that catches fire, in any case. > > What about overvoltage and reverse? > > The input circuits I've seen, it's more or less straight into the > MOSFETs. Maybe a ferrite bead, but that doesn't do jack, it saturates > quickly during input transient (and maybe the ceramic caps too, if > allowed to overshoot) and in operation.� No TVS. > > It's a very real possibility. I had one mobo get nuked because of a > frayed power supply.� What happened is the coaxial cable (they're > constructed with two braid layers, carrying current, with a data line in > the center) got frayed at the PSU's strain relief, and was shorting out > intermittently. Somehow or another, this caused enough stress to blow up > a couple parts (presumably there was a general power surge that took out > whichever chips broke down soonest). > > The bizarre thing is, the most burnt chip was an LM393 near the battery > controller.� How, and why, and also why?... > > Possibly, if there had been a TVS diode on the input, that wouldn't have > happened. > > Tim >Shifting the cost of a fault-tolerant power input/charging circuit onto a vendor-locked-in consumable item on the low-end models also seems a reasonable possibility, yeah. Who _would_ want to spend money on extra circuitry to protect their product from someone else's bodge-job? But that it's a "safety feature" seems disingenuous; other major manufacturers are apparently confident enough in their designs to not be so concerned about aftermarket chargers being used. "Safety" from their own corner-cutting, more like.
Reply by ●January 9, 20182018-01-09
"bitrex" <bitrex@de.lete.earthlink.net> wrote in message news:DNU4C.80908$mG5.53546@fx10.iad...> Shifting the cost of a fault-tolerant power input/charging circuit onto a > vendor-locked-in consumable item on the low-end models also seems a > reasonable possibility, yeah. Who _would_ want to spend money on extra > circuitry to protect their product from someone else's bodge-job? > > But that it's a "safety feature" seems disingenuous; other major > manufacturers are apparently confident enough in their designs to not be > so concerned about aftermarket chargers being used. "Safety" from their > own corner-cutting, more like.Yep, point in case -- the authentic (HP in this case) adapter damaged my motherboard. Wouldn't have mattered if it was another brand or off-name. It's an excuse that can work, but you have to actually do the work to make it true. Power adapters are way, way closer to mutual compatibility than, say, game console controllers are. The average consumer has no clue what that means, or how to test it (besides plugging it in and seeing it not work, or asking someone more experienced about why). So it's just hot air: marketing wank, unnecessary vendor lock-in. Tim -- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: https://www.seventransistorlabs.com/
Reply by ●January 9, 20182018-01-09
> > The input circuits I've seen, it's more or less straight into the MOSFETs. > Maybe a ferrite bead, but that doesn't do jack, it saturates quickly during > input transient (and maybe the ceramic caps too, if allowed to overshoot) > and in operation. No TVS. > > It's a very real possibility. I had one mobo get nuked because of a frayed > power supply. What happened is the coaxial cable (they're constructed with > two braid layers, carrying current, with a data line in the center) got > frayed at the PSU's strain relief, and was shorting out intermittently. > Somehow or another, this caused enough stress to blow up a couple parts > (presumably there was a general power surge that took out whichever chips > broke down soonest). > > The bizarre thing is, the most burnt chip was an LM393 near the battery > controller. How, and why, and also why?... > > Possibly, if there had been a TVS diode on the input, that wouldn't have > happened. > > Tim >I blew up a GPS antenna the same way. If there is an intermittent short at the end of the cable, when the short is on, current flows through the cable which at DC appears as a good size inductor. When the short opens, the current wants to keep flowing so the voltage raises to whatever it takes to keep the current flowing. m
Reply by ●January 9, 20182018-01-09
On Mon, 08 Jan 2018 17:09:52 -0600, Jon Elson <jmelson@wustl.edu> wrote:>oldschool@tubes.com wrote: > > >> I woundn't own a Dell if it was free. They are the worst brand sold. >> They use parts made only for their systems. Even their case fans are >> made just for Dell, and they cost a fortune. I have my preferred brands, >> even though I build my own desktop machines, but Dell is not even on my >> list. I'll leave them in the store so someone else can waste their money >> on them and then dispose of them. >> >> Dell is Garbage !!! >I love Dells! I have found the commercial-grade models to be VERY good. >I have had several Dell destop (Optiplex) machines run over 12 years in 24/7 >operation. Yes, a lot of stuff in them is special, for instance the fans >are made to be extra low-noise. Who cares, as you can put new bearings in >them if they get noisy, or buy replacements on eBay. > >I have also had really good luck with their laptops, but again, there are >some really great models and some that are a bit mediocre. I buy all my >Dell boxes used for, usually, under $100 delivered. > >I did just have a motherboard go bad. First, it was bad capacitors, but >then it just got sick. Possibly it is a bad capacitor with no visible >signs. I only replaced the obvious ones. > >JonAll brands go bad eventually, but I have not been impressed at all by Dell. My preference is Lenovo. I am posting this from an 18 year old Lenovo computer running Win98. I did have to replace the original power supply (which was under rated), but thats probably because I added a bunch of drives and stuff. Otherwise it's still the original motherboard which has never needed any repair aside from changing the CMOS battery. I run this system to the limit too, as far as RAM and drives. When I changed the power supply, I went from 100W to 350W. More than enough...
Reply by ●January 9, 20182018-01-09
On Mon, 08 Jan 2018 17:09:52 -0600, Jon Elson <jmelson@wustl.edu> wrote:>I love Dells! I have found the commercial-grade models to be VERY good. >I have had several Dell destop (Optiplex) machines run over 12 years in 24/7The commercial Dells might be a lot better. I never saw one of them...
Reply by ●January 9, 20182018-01-09
On 01/09/2018 01:24 PM, oldschool@tubes.com wrote:> On Mon, 08 Jan 2018 17:09:52 -0600, Jon Elson <jmelson@wustl.edu> wrote: > >> oldschool@tubes.com wrote: >> >> >>> I woundn't own a Dell if it was free. They are the worst brand sold. >>> They use parts made only for their systems. Even their case fans are >>> made just for Dell, and they cost a fortune. I have my preferred brands, >>> even though I build my own desktop machines, but Dell is not even on my >>> list. I'll leave them in the store so someone else can waste their money >>> on them and then dispose of them. >>> >>> Dell is Garbage !!! >> I love Dells! I have found the commercial-grade models to be VERY good. >> I have had several Dell destop (Optiplex) machines run over 12 years in 24/7 >> operation. Yes, a lot of stuff in them is special, for instance the fans >> are made to be extra low-noise. Who cares, as you can put new bearings in >> them if they get noisy, or buy replacements on eBay. >> >> I have also had really good luck with their laptops, but again, there are >> some really great models and some that are a bit mediocre. I buy all my >> Dell boxes used for, usually, under $100 delivered. >> >> I did just have a motherboard go bad. First, it was bad capacitors, but >> then it just got sick. Possibly it is a bad capacitor with no visible >> signs. I only replaced the obvious ones. >> >> Jon > > All brands go bad eventually, but I have not been impressed at all by > Dell. My preference is Lenovo. I am posting this from an 18 year old > Lenovo computer running Win98.If it's an 18 years old Thinkpad, it's an IBM and not a Lenovo! I did have to replace the original power> supply (which was under rated), but thats probably because I added a > bunch of drives and stuff. Otherwise it's still the original motherboard > which has never needed any repair aside from changing the CMOS battery. > I run this system to the limit too, as far as RAM and drives. > > When I changed the power supply, I went from 100W to 350W. More than > enough...I have probably 20 Lenovos of various vintages stashed away--T42s for DOSish things and T410-T430 for daily use. Cheers Phil Hobbs (Posting this from a 24-core AMD server that doesn't have a management engine and doesn't suffer from the recently-disclosed hardware bugs.) :) -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net https://hobbs-eo.com
Reply by ●January 9, 20182018-01-09
On Tuesday, 9 January 2018 19:21:06 UTC, Phil Hobbs wrote:> On 01/09/2018 01:24 PM, oldschool@tubes.com wrote: > > On Mon, 08 Jan 2018 17:09:52 -0600, Jon Elson <jmelson@wustl.edu> wrote: > >> oldschool@tubes.com wrote: > >> > >> > >>> I woundn't own a Dell if it was free. They are the worst brand sold. > >>> They use parts made only for their systems. Even their case fans are > >>> made just for Dell, and they cost a fortune. I have my preferred brands, > >>> even though I build my own desktop machines, but Dell is not even on my > >>> list. I'll leave them in the store so someone else can waste their money > >>> on them and then dispose of them. > >>> > >>> Dell is Garbage !!! > >> I love Dells! I have found the commercial-grade models to be VERY good. > >> I have had several Dell destop (Optiplex) machines run over 12 years in 24/7 > >> operation. Yes, a lot of stuff in them is special, for instance the fans > >> are made to be extra low-noise. Who cares, as you can put new bearings in > >> them if they get noisy, or buy replacements on eBay. > >> > >> I have also had really good luck with their laptops, but again, there are > >> some really great models and some that are a bit mediocre. I buy all my > >> Dell boxes used for, usually, under $100 delivered. > >> > >> I did just have a motherboard go bad. First, it was bad capacitors, but > >> then it just got sick. Possibly it is a bad capacitor with no visible > >> signs. I only replaced the obvious ones. > >> > >> Jon > > > > All brands go bad eventually, but I have not been impressed at all by > > Dell. My preference is Lenovo. I am posting this from an 18 year old > > Lenovo computer running Win98. > > If it's an 18 years old Thinkpad, it's an IBM and not a Lenovo! > > I did have to replace the original power > > supply (which was under rated), but thats probably because I added a > > bunch of drives and stuff. Otherwise it's still the original motherboard > > which has never needed any repair aside from changing the CMOS battery. > > I run this system to the limit too, as far as RAM and drives. > > > > When I changed the power supply, I went from 100W to 350W. More than > > enough... > > I have probably 20 Lenovos of various vintages stashed away--T42s for > DOSish things and T410-T430 for daily use. > > Cheers > > Phil Hobbs > > (Posting this from a 24-core AMD server that doesn't have a management > engine and doesn't suffer from the recently-disclosed hardware bugs.) :)Some of the old IBM Thinkpads were pretty nice. Some less so. Too ancient for much use now though. NT
