diff tcsm2-notes/trco @ 18:b36419bbc2c4

tcsm2-notes/trco: update after peeling off IC stickers
author Mychaela Falconia <falcon@freecalypso.org>
date Wed, 29 May 2024 23:11:28 +0000
parents 776229c5d640
children
line wrap: on
line diff
--- a/tcsm2-notes/trco	Wed May 29 03:00:02 2024 +0000
+++ b/tcsm2-notes/trco	Wed May 29 23:11:28 2024 +0000
@@ -3,17 +3,31 @@
 1x Lattice ispLSI1032E PLD, same as on TR16-S
 1x CPU: Intel TN80C186EB20
 
-1x really big oscillator component, labeled 16.384 MHz, likely OCXO (or TCXO?)
-1x MAS9316N DAC close to it - likely TDM common clock for the system
+Big oscillator, markings on the metal can package:
+
+	DFA 36-TR
+	16.384 MHz
+	A36060 0049
+
+Close to it there is a MAS9316N DAC - thus the big oscillator is likely the
+system-wide TDM clock (the one that can be synced to various inputs), and the
+DAC is for control loop adjustments aiming to lock this oscillator to the
+selected external clock source.  Based on the size of the can, I originally
+thought it was an OCXO, but when I searched for "DFA 36-TR", the hits I got
+(only sellers, no datasheet) say TCXO.  Who knows...
 
 3x Infineon chips, 2x SAB 82525 (2x HDLC) and 1x SAB 82526 (1x HDLC):
-the 5 HDLC channels which the available documentation touches on
+the 5 HDLC channels which the available documentation touches on.
 
 8x PLCC32 chips (9 pins on wide side, 7 pins on narrow side) with stickers
 on them that obscure the markings underneath.  The markings read IC45 through
 IC48 in the top row of 4, IC55 through IC58 in the bottom row of 4.  All 8
 ICs are directly soldered.  Suspected flash memory.
 
+Update with IC55 sticker peeled off: the IC underneath is AM29F010B-120JC,
+just like I suspected.  Interestingly, it also says "(C) 1998 AMD" - the year
+is slightly newer than my original assumptions of 1996 vintage.
+
 Next to these 8 ICs there is a socketed PLCC44 IC (11 pins on each side)
 with a sticker that reads:
 
@@ -24,7 +38,22 @@
 It is clearly a programmed part of some kind, but not clear if it is flash
 (boot code?) or some PLD or whatever.
 
-Further insight can be gained by peeling off the stickers, but I am not going
-there yet: my current plan is to try powering this unit as-is and see if it
-comes to life in a useful way.  If it turns out to lack the needed fw, or lack
-TFO feature, then proceed further with reverse eng.
+Update after peeling off the sticker: the part is ST M27C4002, an OTP ROM!
+It is clearly the boot PROM, whereas the other 8 ICs (all 29F010?) are the
+flash memory array.  The large size of both the flash array and the boot PROM
+is surprising:
+
+* The boot PROM is 4 Mbit (256K x 16) - that's 512 KiB, half of 80186 address
+  space if fully and directly mapped.
+
+* Assuming that the 8 bit wide 29F010 chips are wired in pairs across the
+  16 bit wide data bus, we got 1 MiB in total, in 4 banks of 256 KiB each.
+  More than 80186 can address directly - is it a bank-switched memory system?
+
+RAM chips:
+
+* 2x Samsung K6T1008C2E-GL70.  Each of these ICs is 128K x 8 static RAM;
+  the design probably uses two of them across the 16 bit wide data bus.
+  256 KiB of RAM here.
+
+* 1x CY62256-70SNC - 32K x 8 (256 Kbit) static RAM.  What is it for???