1. Home
  2. General Electronics
  3. Any other suggestions? (was Re: Anybody able to ID this SM diode for me?)

Any other suggestions? (was Re: Anybody able to ID this SM diode for me?)

D> > > >

> > > > MAYBE there will be further damage. I'm not as pessimistic as you seem > > to be about this thing's chances. Worst case scenario: I spend a buck > > or two and a few minutes, and it remains dead. Best case: It returns > > to > > life as good as new. I'm game to try it! :) > > > > I'm quite aware that what I'm looking at is just the *VISIBLE* damage. > > But scrapping this rig (which, in its day, was the absolute > > top-of-the-line - What am I saying "in its day"? The thing was still > > the best modem I've ever met until the hit that killed it. > > "v.Everything" meant exactly that...) because it *MIGHT* have more > > damage is an idea that's simply repugnant to me. > > I wonder, these days, what the old V.Everything actually runs at. > Has it actually done anything that isn't V.90 in the last few years?

Actually, it was flashed up to v.92 within the last year or so. Although I mostly (like 99.99% of the time) use it to hook to my ISP, which, due to being on the end of a 20+ mile run of mixed copper and glass, generally means 26400 connects, I like having the ABILITY to "go anywhere, do anything" that it represents, even if I don't actually

*USE* that ability often.
> >> A recent job I sorted saw the whole computer fried including the > >> screen from a strike which killed the fax machine as well as the PCI > >> modem! > > I'm glad the rest of the system is OK! I have uploaded a couple of pix: > >
formatting link
(33kb) > >
formatting link
(36kb) > > The fax was a Philips machine, the modem was a noname PCI internal. > You can see the similarities with your description.

FWIW: There's another reason why I'll *NEVER* own an internal modem - WAY too much potential to take out the entire machine. With externals, a hit will take out the modem, and MIGHT (but IMExperience, never has) take out the serial port. With internals, a hit can easily wipe the entire machine - there's no "insulation" or "buffer" between the modem and the computer like a serial port provides - The juice from the strike is going DIRECTLY to the internals of the computer, with basically nothing at all standing in its way. Sorta like the difference between getting tagged by brushing up against 110 (or in your case, 220) with your hand, vs. grabbing one wire from the mains in each hand - The former hurts, sure, but it isn't likely to kill. The latter is a real good way to get an appointment with the coroner.

Looks like that gear you played with took a much "heavier" hit than my modem did. Aside from having the "two pin" edge of the diode blown off, the damage in mine is visible mainly as 6 of those little surface-mount resistors wearing "blisters" (no visible charring or anything like that

- just the paint is blistered up a bit) and a sort of "haze" on the coating of the board at the point where it looks like the arc jumped from one of the phone jack pads to a nearby pad that appears to be associated with a jumper block whose purpose I once knew for sure, but have now forgotten. (I have a foggy memory of that jumper block having something to do with "Synchronous" vs "Asynchronous" communication from modem to phone line.)

Anyway, those jumpers were all "open", but the trace that got jumped to is hooked into a series of 4 resistors (marked "332" - I'm weak with surface-mount devices - should I, as I expect, be reading those as "3300 ohms"?) that are wired back-to-back-to-back-to-back, then into the cathode side of the exploded diode, which connects the resistors and jumper block directly from the negative side of the bridge rectifier that turns the juice from the wall-wart (nameplated as putting out 20 VAC @ 500 mA, measured as 21.6 VAC) into 27-ish volts of DC. (The "-ish" is because I never got it to "settle" - it bounces around from 26.8 to

27.5 VDC according to my DMM - since there's an LT1271 regulator a bit further downstream, I have *SERIOUS* doubts about a 0.7V variance on the rectifier pins being significant to a switch-mode regulator that's rated for an input voltage of +5 to +30 VDC)

On the "update" side of things, I managed to locate a "5D" diode (actually, several dozen of them) on one of the bazillion broken/obsolete/partially stripped boards that we've got laying around here, and after replacing the blown one, found that there's (Surprise...) nothing getting from the phone line into the "guts" of the modem. I say "Surprise..." because two of the cooked resistors (both marked "501") are connected almost directly (through a toroidal coil I expect is a noise choke, and then through a MOV that bridges the outputs from the coil) to the input pins from the phone line - One on each trace that the MOV is on. So, I got out the magnifying glass and started hunting for "501" (500 ohms, right?) SMD resistors. Of course, Murphy's law kicked in, and I was unable to find a pair of them. So, going with the hunch that a 30 ohm difference wouldn't be enough to mean anything, I grabbed a couple of similar-sized resistors marked "471" (470 ohms, right?) and replaced the two on the board (After removing them from the circuit, both of them read as wide open on my DMM. In-circuit, I couldn't get a reading that I'd call reliable) and tried again.

This time, what came out of the speaker sounded like crap, but at least it was *SOMETHING* - I couldn't pick dial-tone out of the static, but hey... I'm now better off than I was before... At least *SOME* kind of noise is coming out of the speaker, which tells me that I've fixed one (of who-knows-how-many-yet-to-come...) problems!

I'm suspecting that the 30 ohms I wrote off as "no big deal" might be more important than I realized, since the "static" coming out of the speaker now is strongly suggestive of the kind of distortion and noise you'd get from *MASSIVELY* over-driving a speaker. Obviously, it's not yet working...

--
Don Bruder -  dakidd@sonic.net
Reply to
Don Bruder
Loading thread data ...

I have seen a lot of sterios trashed by lightning . Few were reasonably fixable .

tim

-- ....( remove the " 6 " to e-mail reply )

info.

Reply to
Tim Kettring

Gee Whiz! Ain't the internet great? Thanks for such an incredibly useful and informative answer, Tim! What would I have ever done without your input?

Look, seriously, I know the thing is likely dead. That doesn't mean I'm going to shitcan it automatically. For all anybody knows at this point, I may be able to fix it completely by finding the one critical three-cent piece that got burnt and replacing it. Or, alternatively, I may never get it to anything approaching functional. Either way, I'm quite aware that lightning makes a mess of electronics, and replacing hit items is usually easier than fixing them. That doesn't stop the cheap bastard in me from trying.

Now, if you've got something helpful to say on the topic of diagnosing/repairing a lightning-hit modem, I'd love to hear it. But your top-posted one-line response that even an idiot already knows, complete with nearly 150 lines of original text retained, isn't doing any good whatsoever.

--
Don Bruder -  dakidd@sonic.net
Reply to
Don Bruder

I forgot to say that in both pix, the phoneline sockets are at the top of the picture.

Yep, I like externals, too. So much so that I put one *inside* my computer! One of the pages in the Miscellanea->XP1800+ article shows how I did it!

Yes, me too! But 332 does tie in nicely with 3,300 ohms.

I wonder if these really are resistors? Although I don't know what else they could be - anyone?

Agreed.

It may also be due to those 501s not being resistors, or there being a bad resistor that is now falling apart and injecting noise.

--
Graham W   http://www.gcw.org.uk/ XP1800+ Page added, Graphics Tutorial
WIMBORNE   http://www.wessex-astro-society.freeserve.co.uk/ Wessex
Dorset UK  Astro Society's Web pages, Info, Meeting Dates, Sites & Maps
Change 'news' to 'sewn' in my Reply address to avoid my spam filter.
Reply to
Graham W

If it is a 2-terminal device marked with nothing more than a 3- or

4-digit code like this, then it is almost certainly a resistor.
Reply to
Lewin A.R.W. Edwards

That's what I thought... The parts in question are have black (although they do seem to have a tiny bit of greenish tint to them) bodies, a narrow, obviously solder-colored band at each end, and three clearly visible digits - "501" - which I interpret, using the standard practice that goes with decoding conventional color-banded resistors, to be "5" followed by "0" followed by "a single 0", or 500 ohms. There is zero possibility that the "5" is actually an upside-down "2", and I'm reading it backwards - It's quite clearly a "5", and the "1" is most definitely a "1" (it has a serif, albeit a small one) but only when oriented so that the "5" is really a "5". Any other way of trying to read this thing gives gibberish. (unless, perhaps, "20L" is supposed to be read as something like "20 followed by 12 zeroes ohms"...)

--
Don Bruder -  dakidd@sonic.net
Reply to
Don Bruder

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.