29 Jan
2006
29 Jan
'06
1:46 p.m.
signal or the signal is not DC, but some pulse. So
then I pulled the D100 o=
ut.
The inputs are all set to TTL, and the 'ground' I used was pin 7 of E46.
Pardon my ignorance, but what is the D100? I assume some kind of logic
analyser.
I checked the 4-input NAND E5 ... although all 4
inputs (xxx COMBINED
ERROR L)> are continuous "1", I see lots of pulses on the output pin 8 ?=
Again, impossible unless eirhter E5 is
malfunctioning, or you're not
actually seeing what's at that point. I hate to say this, but I suspect
the instrument and/or its use.
=
And again, I was amazed too. But the output of E5 did not look like a
spurious signal.
The D100 gets correct through the initialisation at power up ... I know what you mean :-) The RK07 and its controller
is definately complex=
There's certainly a lot of circuitry in the RK07/RK611 (probably around
1000 chips total), but the DEC prints are pretty clear and divide the
circuitry up into little bits with named signals (and helpful names at
that), so it's not that hard to work out what's going on.
The BC06R is a "-6" version, so a lot
shorter (and shielded).
Regarding noise, I suspect that the print set is not 100% up-to-date, but
given the quality of the drawings from DEC, I assume that I am wrong. But,
pin 2 of the E46 S/R flipflop is not connected (according to the diagram).
I am pretty sure that pin is floating. HP did the same thing on some of
their boards too (and I have verified that the D input then goes
nowhere). If the clock input is tied low, there's no way the D input can
affect anything.
This is the D input of the flipflop, and since the
clock input (pin 3) is
connected to Gnd, the state of the D input is irrelevant according the data=
book. However, with the Fluke 77 I measure 1.55 V on pin 2. I consider not-=
Ah.... Do you read Elektor? Sometime last year there was a puzzle based
on just this effect. Basically, a TTL input will source current, and the
apparannet voltage will depend on the state of other inputs to that gate
(which might well be internally connected in the IC).
connected input pins a bad design, but I will not tie pin 2 to Gnd or with =
a
1k pull-up resistor to +5V, because that pin changes voltage during reset!!=
I will assure you it would be quite in order to tie pin 2 -- an input --
to ground.
If you fold
out the logic cage in the RK07, you should see a 00 01 10
pattern on the head select LEDs and a binary count on the cylinder LEDs.
=
Now, why did I (again) not think of that!? Looking at the head movement Ah well, I think it is just a matter of how you look
at things. Without goi=
ng
into that subject, I'd say that swapping is for some of us the only option.=
It's not just that. One of the replies I got to my comment that the
security on PALs was not very useful (to put it mildly) seemed like a
flame to me.
Not
everybody understands electronics that well as you do Tony. An educated
I don't claim to be a programmer, but I sure as heck don't debug programs
by making random changes and seeing what hapepns. I do try to understand
what's going on.
And I must admit that my first reaction when I discover I don't know
something that would be useful to know is to try to learn or understand
it.
guess when you swap *one* board is still better then
tossing the stuff in t=
he
bin, because it does not work. It isn't always that black and white.
Ture enough. Another alternative, given that we don't depend on these
machines for income, etc in most cases would be to put the machine on one
side until we have the skills to repair it. I have many devices that I've
not repaired yet because, for example, I need to learn how to cut a gear.
I actually have the tools to do that (and to make the form cutter), I
need to spend a lot of time working with them and practising.
For me, that's one of the great things about this hobby, all the extra
skills I've had to learn.
-tony