Author Topic: Motorcycle Alarm Teardown (or Why you don't want to buy a cheap alarm)  (Read 10916 times)

0 Members and 1 Guest are viewing this topic.

Offline phil_jp1Topic starter

  • Regular Contributor
  • *
  • Posts: 103
  • Country: ua
Re: Motorcycle Alarm Teardown (or Why you don't want to buy a cheap alarm)
« Reply #25 on: September 01, 2012, 07:11:48 pm »
Which is pretty much overclocking  ;)

Thanks for sharing,I will try 8)

Don't do it, it's a bad thing! You will potentially steal money from CPU manufacturer, by not buying a new more powerful processor! And if you have PC made by Apple - they'll sue you, like they did with Samsung, but on smaller scale!  ;D
http://JumperOne.com - Electronic projects, tutorials, hacks, etc.
 

Online AndyC_772

  • Super Contributor
  • ***
  • Posts: 4278
  • Country: gb
  • Professional design engineer
    • Cawte Engineering | Reliable Electronics
Re: Motorcycle Alarm Teardown (or Why you don't want to buy a cheap alarm)
« Reply #26 on: September 02, 2012, 09:08:52 pm »
Probably depends on the exact manufacturer. But it does make sense from an economic standpoint if a part isn't 100% functional to make a profit selling the parts of it that may still be OK. Similar to how AMD produces multicore CPU dies but sells them cheaper with one or more cores disabled because they didn't pass testing. Likewise an MCU with broken subsystems would still be usable in an application that didn't make use of them.

Microchip also has MCV08A/14A/28A and the similarly obscure HAxxxx series which are apparently for "home appliance" use.
That's fair enough, though the impression I got was that these PICs are being sold not just as 'guaranteed to work in your particular application', but as 'not tested at all, and may therefore be scrap'. That would make some sense if the testing were actually the most expensive step in the manufacturing process, and I wouldn't be at all surprised if it is.

Selling devices with faulty cores disabled and the like is common, familiar stuff, and of course it's perfectly OK if the rest of the chip has been tested thoroughly, and so each die's failings are identified.

The most extreme example of this I've come across is what some FPGA vendors will do, whereby they'll sell a device (at lower cost, of course) which is a known test failure, but which is nevertheless guaranteed to work with a particular configuration. That's a pretty good way of salvaging some revenue from the scrap bin, but I can't see it being done on a PIC somehow. Not at <$1 each.


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf