[time-nuts] Austron 2100F
cfharris at erols.com
Wed Apr 1 11:47:19 UTC 2009
You're asking me if I can say more? Most wish they could get
me to shut up;-)
The design for the receiver of the 2100T/F is a mishmash of synchronous and
asynchronous logic. They mixed logic families in an apparent attempt to
battle a few timing problems. On one board, there is 7400, 74LS, 74L, 74C,
25LS, and 74HC all at the same time, and all in the same signal path!
Following the trend of the time, they used signature analysis to aid in
troubleshooting... only they didn't try very hard. The signatures are
fine for working out problems with the CPU, and IO section, but they do
nothing for most of the logic in the receiver. And signature is unable
to do anything for timing problems.
So, how do you find a thermally related timing problem?
I had two 2100F's one was rock solid, and one was rock solid, as long as
my rack stayed below 75F. Summertime would hit, and the receiver would
lock, run for about 2 hours, and then it would be unlocked. I swapped
boards between the two receivers until I determined that it was without
doubt the Acquire/Track board. Wherever that board was, the receiver in
question would fail when it got warm.
The Acquire/Track board is where they divide down the reference oscillator
to form a mask of the loran signal. The mask is matched to the signal,
and the divider is adjusted by adding, or subtracting counts, to track the
I checked every trace on that board, tested every chip for logical function,
and ultimately replaced every part on that board with new, and it still had
a thermal problem. I tried to concoct a method of even sensing the problem,
but the very slow nature of the problem (due to the slow changes between the
reference and the loran signal when you were within the range that the
receiver could tolerate) defied my attempts.
If your design needs 7400, 74LS, 74L, 74HC, 74C, and 25LS all in the same
counter/comparator to work, you have botched the design! Although the
manual is silent on the fact, I am pretty sure that they had to spec the
parts on this circuit down to the manufacturer and lot number to come up
with a set of parts that had all of the timing and threshold problems
settled so that the board could work.
I ultimately concluded that the use of 7400, 74L, and 74LS with its TTL
(1.3V) threshold levels was only marginally compatible with the 74C and
74HC with their CMOS (2.5V) threshold levels, and could only work at
christopher hoover wrote:
> Chuck Harris <cfharris at erols.com> wrote:
>> They can be very hard to fix, because there is some really poorly designed
>> logic circuitry, and the signature analysis functions really can't find
>> very many problems.
> Interesting. Can you say more?
> I tried to repair mine -- I have a signature analyzer -- but did not ever get it to work. The initial indication was a bad memory chip. I carefully replaced it, but it never came back to life. Maybe that was a red herring.
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts