On the risk of getting OT, but desoldering ICs is certainly on-topic...
> > Good luck! Meanwhile, I'll try to get that blown 74123 out of my RX211
> Best method for replacing chips is to cut the pins of the broken
> part and then take out the single pins with a desoldering pump.
> Works good even on multilayers with a good soldering iron.
Yes, I have not only heard this hint, but also tried to follow it. At
least with my tools, I always have trouble getting to the pins to cut.
There is not much room between a pin, the board, and the IC body. For
the corner pins, it's quite easily doable, but when it comes to the
other ones, you have remainings of neighbor pins that also get in the
way.
Using a pump is mandatory, no question about this. I sometimes can
improve results by using some screening maze from a coax cable to take
up solder that my pump would refuse to eat. I'd love to try one of those
electric desolder pumps: it must be a great advantage if you have a
continuous vacuum instead of one shot (even if in the case of this chip,
a one-shot may be appropriate ;-). Jochen, you are in Germany as well:
are those ELV7000 series kits still around somewhere?
(For international readers: ELV used to be a German magazine for
electronics hobbyists, and they created a series of very interesting
devices, like voltage and frequency generators, measurement equipment
etc., all to be homebuilt by their readers. However, at that time, I
lacked the money for the kits. They appeared to be very good, but not
very cheap.)
Another point is the attribute "good" attached to "soldering iron". I do
not have one of the common little stations with a transformer and an low
voltage iron, and when I mention that my iron is a 70W type, people
usually get scared. However, it is the only type I ever saw that is
temperature controlled and adjustable, even though it looks like a cheap
standard thingie and has no transformer with a stand and a sponge etc. I
am happily using it for more than 20 years now, and the only trouble
with it is the PVC insulation of the cable: it's fading now and will
soon need replacement.
I think if I use not too high a temperature, I can save the copper leads
from getting off the epoxy, and it wouldn't matter if desoldering takes
a bit longer: the bug is dead anyway.
> > I suspect replacing the 74123 with a 74LS123 will change the timing? -
> Depends on the purpose of the gate. If it is not used to generate
> timing it should be no problem to replace it with a LS type. (IMHO)
Haven't looked up what the other half of it is doing. The bad output is
used to sample a signal coming from the address decoder and MSYN (master
sync), IIRC. This signal is latched into a D-FF under control of the
one-shot to generate SSYN (slave sync). Sounds much like bus timing,
huh?
I think I'll spend a socket on this chip, even if it wasn't there
originally. This way, I might try a LS123 from my stock first, then a
standard 123 if the LS type doesn't play.
Meanwhile, I found formulas to compute the pulse duration, and, yes,
they differ between LS and standard, not only in coefficients, but even
in the form of the expression. In case somebody cares:
74123: t = 0.32 * R * C * (1 + 0.7 / R) with 5k < R < 50k
74LS123: t = 0.45 * R * C with 5k < R < 260k
I might even want to read the UNIBUS specs...
--
Andreas Freiherr
Vishay Semiconductor GmbH, Heilbronn, Germany
http://www.vishay.com
Received on Thu Mar 07 2002 - 07:10:08 GMT