From: Andreas M. <an...@us...> - 2003-12-20 16:42:35
|
Hi, On Sat, Dec 20, 2003 at 05:28:52PM +0100, Marcel Telka wrote: >=20 > Nap=EDsan=E9 d??a 20.12.2003 17:00, (autor: Andreas Mohr): > >> If there is a pin list for the chip and at least one JTAG ball is =20 > >at > > > >> the chip edge then you could use ohmmeter to find out JTAG pin. Or, > >if > >> the chip is non-BGA, then the situation is trivial :-). > >Wow, you mean it's possible to find out the remaining JTAG pins by > >measuring >=20 > No. I mean that if you have a pin list for the chip and if you have a =20 > non-BGA chip, then you have access to all JTAG pins on the chip =20 > directly. Yes, of course. But with a non-BGA chip it's trivial and thus boring ;-) > >from one JTAG pin to another, and if we get an actual resistance =20 > >value > >back, > >then the one we're measuring against is also a JTAG pin? > > > >Also, how to proceed if I know the list of pins, but don't know which > >ones > >are which? >=20 > Hm. In that case you have "a half pin list" :-). But if you are sure =20 > whether the chip has JTAG then you could (theoretically) try all =20 > possible combinations (crazy idea...). >=20 > >Do you know whether there's any risk involved if I connect them to =20 > >the > >JTAG > >adapter in the wrong order? >=20 > AFAIK no. Because 4 of 5 pins are inputs (TCK, TMS, TDI, TRST) and all = =20 > possible combinations of the values at that pins are possible (valid) =20 > and shouldn't damage the chip (no warranty, of course). >=20 > There is only one risk: TDO, but if you connect a small resistor =20 > between adapter's pins and chip pins (say 470 ohm), then current would = =20 > be too small to damage the chip. OK, in case I'd happen to be desperate (in case I killed my whole flash o= r so), then I'd actually try that. > >And how risky is it to measure against random pins or test points? > >I.e. how likely is it to destroy some stuff via measurement currents? > >Anyone with real-world experience here? >=20 > I've measured some devices in that way without any damage... > The devices were unpowered :-)... Uh, yeah... better DON'T do it with power... ;) > >> >I'm highly interested in that since I'd like to have some fallback > >> >solution > >> >in case I happen to brick my Axim during my Linux development > >> >experiments... > >> > >> If the Axim has a PXA255, then the situation is good for you, > >because > >> PXA255 has some JTAG balls near to the chip edge... > >I know, I've been studying the JTAG pins myself already :-) > >(but it doesn't help much if you still cannot find the test pins or > >connectors > >for that :-\) >=20 > Yes. That is bad :-(. >=20 > Have you a (detail) photo of the board (both sides)? You could place it= =20 > on the web... Yes, it's already up (not mine, though), at: http://www.handhelds.org/z/wiki/DellAximX5 Andreas Mohr |