Test 2: A bit of Deja Vu and The P2.5 line
Test 2: A bit of Deja Vu and The P2.5 line
The data line from the 80c51 (power mgmt control) labelled P2.5(pin 29), if pulled low (bwo the el-cheapo logic probe I built) while executing power-up (pressing the button!
)
seems to enable a
mostly complete power-up sequence:
** Button is pressed,
** P2.5 goes high (logic probe glows weakly),
1) Keyboard lights come on and clear (numlk,scroll, 'special' lights),
2) System accesses the floppy drive,
3) System accesses the hard drive,
4) System beeps twice.
** P2.5 goes low (logic probe goes out).
At this point the probe can be removed and the system will remain active (by indication of the power/speed light). The system will drop/shut off if the probe is removed BEFORE this point.
It will sit there spinning the hard drive and remaining quiet UNLESS a key (or reset button, obviously) is pressed, presumably causing an interrupt (tbd). I
presume this interrupt causes the PMC to reassess its inputs and recheck things...
There are only 4 ICs (as discovered so far) that are DIRECTLY affected by this line:
1) the 80c51 PMC (originator),
2) an HC273 (pin 14, INPUT from 80c51-29) which is where the probe is usually placed. HIGH on this line is ORed into a NMI,
3) an HC174 (pin 13, INPUT from 80c51-29), whose HIGH output (pin 15) is responsible for enabling the -22v to the LCD panel (LCDSW).
4) the 727459 Battery mgmt control (pin 15) INPUT FROM 80c51-29. The 459 feedbacks various info regarding charging status and as-yet-unknown 'stuff'...
While on the subject of the BMC...
I have done some more poking around and found out that there MAY be a charging problem with the 2810... which may be related to the power-up issue... and maybe not.
Specifically, It shows NO attempt to produce voltage (of any significance) at the battery-charge terminals when 'tricked' into believing a battery is there. Or so I think... the behavior of 2810 SEEMS similar to the 1800 in design.
MORE interestingly, when there is NO battery at all, the LOWBAT light (red) BLINKS... and that is the 1st time (I know of) that Ive seen that.
EDIT: was caused by a 'short' rather than a 'resistance' like the probe
Now that the 'weekend' is showing up I may learn more!
A previous entry...
very early on.
I did some more checking of the power/charging circuits due to the feedback mentioned above, and found:
The 2810 (& CF270) show 0.19v on the battery pins [BATv] in a quiescent state, while showing 2.53v on the 'overheat sense[tsens]' line at the same time. If tsens is brought down
above 0 (the probe drops it by about 1/2 to 1.77v) the charge sequence is begun and BATv goes to 17.6v (approx). Once tsens is released BATv returns to 0.19v.
Edit: I forgot to mention that I changed out the J220 transistor that switches the voltage to the charge.
The BATv now rests at .25v instead of .19v...
Edit2: The BATv has settled back down to 0.19v. This MAY be normal after all...
However, the 1800 and 3800 show different values, specifically a tsens of 5.12v and BATv of 3.65v.
I believe (edit: hope) the powerup issue may be caused by the discrepancy between the values shown by the 2810 and the 3800.
(IF tsens is pulled to 0 ALL units' charge lights blink RED. I believe this is normal.)
I am well aware that:
1) These are mechanically different systems, and
2) I do not have schematics available for the 3800 to do any verifiable comparisons.
3) I do not have a WORKING 2810 system to check values.
At the same time, I have little reason to believe that charging circuits for the SAME MAKE/MODEL of battery would have such disparate operating values. I only hope I am correct in this assumption.
IF ANYBODY has a Tandy 2810hd or Panasonic CF270 who would be willing to check these voltages I would be ETERNALLY GRATEFUL!!!
