Page 172 - Troubleshooting Analog Circuits

Letters to Bob I59

LMC662 have the lowest leakages in plastic DIPS, which are made by an automated
process and are untouched by human hands. The TO-99 and CERDIP packages are
not nearly as good as the plastic minidips which do R repeatably.
17. “Protection diodes can rectify high-frequency noise.” I have never seen this one.
Boy, Roy, you must live in a nasty neighborhood for ambient RF noises. You can
probably run a transistor radio without any batteries.

18. I forgot to even mention thermal probes. We use such probes more for characterizing
than for troubleshooting. Often, a soldering iron does the troubleshooting job faster
but more crudely.
19. The LF198 data sheet mentions that you should not let the Sample input move too
slowly. Are there other S/H circuits that are touchy? Ones that do not mention this
fact in their data sheets?
22. I rarely work with wrapped-wire stuff, but I bet a lot of people get fooled by bad
daisy-chaining of power-supply runs and lousy power-supply bypassing-whether
for linear or digital ICs.
23. I rarely find batteries necessary, but. in extreme cases, they are useful.
24. I rarely find notching out 60- or 120-Hz interference necessary. I usually subtract the
60-Hz noises visually from a scope trace.
25. Yes, portable preamps are often useful.
26. Just as I was saying back in Chapter 8-it’s silly to assume that the CM error is
linear.
27. Design reviews are a good idea. But even if the circuit design is perfect, I find the lay-
out to be pretty critical. So a beer check by all your buddies is awfully important, too.

RAP

Dear Bob:

As a practicing technician for many years, I want to comment on one or two things
I read in your series and perhaps pass on an experience or two.
On page 130 of the August 17, 1989 article (new Chapter 7) you mention the pos-
sibly harmful side effects of drawing base current out of a transistor. Most (if not all)
of the switched-mode power supplies I have come across appear to do just that-to
switch the transistor off more quickly by removing carriers from the base.
This technique appears to work well in practice. I have used the technique in many
of the inverters I have designed, and (on the face of it, at least) there doesn’t appear to
be any component deterioration over time. I usually use some form of reverse-
voltage limiting to ensure that the base-emitter junction doesn’t undergo zener break-
down.
Referring to your comments on p. 132 of the same article, I have to disagree on the
advisability of plugging MOS ICs into sockets with the power applied. I consider it
inadvisable because power may easily be sourced to the chip via its input and output
pins in this situation. Some manufacturers forbid this by implication in the “Electrical
Characteristics” section of their data sheets, and I have witnessed device destruction
being caused by this practice. I am also unable to agree with your comments on not