Jan,

(1) I've never worked in the telecom industry, though I've worked on communication security equipment that operated on PCM and CVSD hardware, nor have I ever said that I did. You've alluded to this misconception before, and, while it would not shame me in any way to say that I had worked in that industry, I have not.

BTW, telecom equipment, though it has lots of power, that power is often far from perfect.

I've "picked on" you because of the anecdotes you've suggested were symptomatic of RESET problems, when, in reality, you've given no evidence that you, yourself, have ever observed these problems, studied them in any detail, or worked on them in any other way, yet you insist that there's something inherently faulty with RC reset which has served, and continues to serve satisfactorily in millions of circuits that have been in operation since before the inception of these "supervisors."

As you probably know, I'm presuaded, myself, that the real flaw is in the idiotic idea that a RESET should be pegged to the changing Vcc level rather than to a nice, solid, low-impedance GND.

I've never advocated against the use of supervisor IC's, but I have complained that, at best, they solve the more innocuous of the two power transient problems.

Apparently, I'm the only one who's presently interested in solving this purported "RESET problem." It does, in fact, annoy me that BBRAM frequently becomes corrupted and that there's cause to suspect it's during prolonged Vcc decay during power-off. It could just as easily occur during a Vcc brownout.

I've suggested that one could do to alleviate such problems is to use the reset IC to stop the oscillator. I'm thinking that would take a transistor or two, but perhaps not. You, apparently, think that would be too costly and complex. Perhaps you're right, as I haven't implemented this yet and thoroughly tested it for evidence that the "problem" is gone when such a "solution" is present.

So far, all I've proven, at least to myself, is that a MAX1232 doesn't solve the power-down transient part of the problem, as the BBRAM is still occasionally corrupted in its presence.

If I ever get any useable information about actually observed symptoms of MCU and system behavior that inidcate a "reset problem" I can set up and perform an extensive test for it, and, subsequently, test to see whether a supervisor IC actually remedies its symptoms.

The fact is, you pointed out the dVcc/dt issue, though it was already known, yet you've not indicated whether you've ever shown that systems exhibiting "reset problems" had adequate dVcc/dt.

Kai indicated that he used a supervisor to compensate for a grossly inadequate power supply, with satisfactory results. Erik, though his story was purely anecdotal and probably "observed" over the telephone, did likewise.

It seems to me that you're trying to make this an argument about whether one should use supervisor/RESET IC's. That's not what I want. In fact, I'm trying to avoid an argument altogether. I want to gather directly observed facts.

I've gotten the impression, so far, that, while many of the forum participants have "designed-in" a supervisor, few have attempted to discern what the real problem was, or whether it was gone after the supervisor was installed. Isn't any one interested in how effective their "solution" was, or why?

RE