MC9S08GT32A Too Much RAM ... really

キャンセル
次の結果を表示 
表示  限定  | 次の代わりに検索 
もしかして: 

MC9S08GT32A Too Much RAM ... really

2,862件の閲覧回数
admin
Specialist II

OK, I'm confused. I've been working with this 9S08GT32A for months now, and I've noticed random data in the memory region above the top of my stack, at $087F, the top of RAM for a GT32A. So just now I got curios and started poking around up there, and I can read and write to memory all the way up to $107F! So, why has this chip got 4K of RAM in it (like a GT60) when a GT32A is only supposed to have 2K?

Markings on the chip are:
MC9S08
GT32ACFBE
3L11Y
CTAA0650

 

 

 

 

 

Added p/n to subject.



Message Edited by NLFSJ on 2007-10-20 09:33 AM
ラベル(1)
0 件の賞賛
返信
4 返答(返信)

1,016件の閲覧回数
UcTechnoGeek
Contributor II
this is a very common thing that uC vendors do.  It's the same mask, many times the same memory and peripherals are there as in their common full up parts, but beware it's not tested nor guaranteed good.
 
You may get lucky using it this time, but next time it might not work (or it can change in future parts).
And what would you complain about, "I know it's not supposed to be there, but...".
 
my two cents.
 
uCTechnoGeek
0 件の賞賛
返信

1,016件の閲覧回数
PeterHouse
Contributor I
Wingsy,

Are you sure it is really 4K and not the same 2K appearing twice?

Peter House
0 件の賞賛
返信

1,016件の閲覧回数
admin
Specialist II
Yes, it's not that, but it may not be real RAM anyway. I just executed a routine to clear all memory up to $107F and locations above $087F did not clear. However, I'm still able to change any location up to $107F using P&E Multilink debugger. Maybe P&E is just making me think there is RAM there when really it's not. I've tried this on 3 different chips and I get the same results. I don't know what it is that's doing this but I'm pretty much convinced that I really don't have RAM up there.

So, if I could I would remove "really" from my original subject line. :smileyhappy:
0 件の賞賛
返信

1,016件の閲覧回数
peg
Senior Contributor IV
Hi Wingsy,
 
It is very common, especially in the early days of a device to actually give the full memory (flash and ram) in what are supposed to be reduced memory devices such as the GB/GT range. This has occurred before with QY/QT devices.
 
The mask code you have quoted is shared between both GB and GT and both 32A and 60A. So it is quite possible this is the case here.
 
P&E attempt to limit access to what is supposed to be there, so try telling P&E you have got a GT60A and give it a GT32A and see what happens.
 
0 件の賞賛
返信