When a diode power bridge - discrete devices or four-legged package - fails, it's almost invariably one diode in the 'negative' arm which goes short circuit. Any thoughts on why this should be, rather than any one of the four failing at random ?
Arfa
No it isn't.
Your observation is obviously flawed. All the diodes pass the same current. By chance you've simply experienced a statistically anomalous percentage of such failures.
Graahm
Oh, sorry. I must be more stupid than I thought then. Do you honestly think that if I had not genuinely observed this, that I would be taking the trouble to post the question ? Sometimes, Graham (how long have you not been able to spell your own name BTW ? ) you can be a real contentious prat.
Arfa
And what makes you so sure you haven't just experienced a statistical anomaly ?
I mean, you'd have to believe in black magic or voodoo / whatever otherwise. There's certainly no possible scientific explanation.
What amazes me as much as anything is that you'd keep a record of this !
Graham
Graham:
You obviously don't know very nmuch about electronic componentgs and circuits. It is very possible that there are components that are associated with the bridge, either on the input or output side, that are not completely symmetrical and that might lead to Arfa's observations.
I can think of several components, such as small value bypass capacitors, bleed resistors, that are not cymmetrical, and that could have an effect, especially during surge events. Look at any TV schematic and you should be able to find these components.
Don't dismiss someone else's observations with such disdain when you don't have firm evidence that you are 100% correct. I have been doing electronics servicing for 50 years, from vacuum tubes to IC's, and there are many strange things that I have seen, but closer investigation has usually allowed me to figure out what the likely cause was, sometimes not at all what things seemed to be at first glance.
H. R. Hofmann
Well, all right then. Just *what* qualifies *you* to tell *me* that I'm wrong ? I have probably changed more bridge rectifiers in the 35 years that I've been repairing stuff for a living, than you have ever even seen. Do you think that you are teaching me something by coming out with 'big' words like "statistical" and "anomaly" ? And what qualifies *you* to be so certain that there is "no possible scientific explanation" ? Explain why it's always the neutral terminal screw that comes loose in a plugtop, for instance. Same current flows in both pins.
I don't "keep a record" of this. If you knew the first thing about the practical world of service, instead of just pretending that you do all the time, you would know that this is just the sort of thing that sticks in a proper service engineer's head. As it happens, the reason that I brought it up was that a colleague of mine, just last week, made the comment to me, and I agreed with him. Then, this morning, the very first job on the bench, had a faulty bridge in it, and what d'ya know - it's one of the diodes in the negative arm again that's short circuit.
Whilst the magnitude of the current in all diodes is theoretically the same - and even that might not be quite true if there is any residual DC magnetism in the core of the transformer - the current which flows in the diodes in the negative arm, flows in the opposite direction to that in the positive arm. Also, the arse end of the diodes in the negative arm, typically go to the chassis mass, which may well be tied to power ground (mains earth) so I think that there might very well be a "possible scientific explanation".
Anyone with a better understanding of the real world than Graham care to comment ?
Arfa
It must be my imagination that I'm an electronics design consultant in that case !
Name ONE !
Utter RUBBISH ! There is nothing asymmetrical about any such parts.
Nope.
Why TV ?
If you look closely enough there will normally be a perfectly scientific exaplanation waiting to be found.
Statistical anomalies included.
Graham
Is it ? Is this a 'well known fact' or something ? Are you actually serious ?
Graham
It IS the same.
Any residual magnetism won't affect the load current one tiny bit.
So what ?
No, that would be voodoo thinking.
Check out Kirchoff's Laws.
Graham
An interesting observation; some thoughtful speculation and perhaps a few experiments may be useful.
It would make an interesting science project for a student to accumulate a number of encapsulated bridge rectifiers (I'll leave it to others to outline the statistics) and build a shorting fixture that randomly shorts the device (in a symmetrical fashion, making sure that it is not synchronized with the power line) and run the experiment. If the anomaly is verified, then deeper investigation is in order such as microscopic examination of intact good devices and post mortems on bad ones. What with the _vast_ amount of trashed electronics these days, it shouldn't be difficult to accumulate hundreds of rectifiers for the endeavor.
Also a 'net search on "bridge rectifier" "failure modes" may turn up something.
Regards,
Michael
Arfa Daily wrote: Explain why it's always the
I'd say the reason is mechanical.
The neutral pin shakes about by the action of the (UK) plug being placed in / removed from the socket. This pin probably shakes about more than the earth pin, as the user is more careful to locate that pin first before slamming the rest in.
The Live connection is at the other side of the fuseholder which takes up some of the movement energy.
-- Adrian C
Deadly. I check the mains plug on every bit of kit that comes across my bench, and in at least 50% of cases, one terminal screw will be loose, and that is *almost* invariably the neutral one. There has been discussion about this in trade magazines, so amongst proper service engineers, it would seem to be well known.
Arfa
Ah, so "statistical anomalies" *are* a scientific reason now then. I thought you said 5 minutes ago that "There's certainly no possible scientific explanation". Make your mind up.
Arfa
a
about
seem
There's a simple explanation for that one. The neutral retaining screw is connected to the prong that goes in and out of socket each time and vibrated each time. The live one is decoupled from a lot of that vibration by the fuse.
But why do christmas tree lights always fail to light when reused the next year but were fine before packing away, at least one bulb is always loosened over that 11 months. ?
-- Diverse Devices, Southampton, England electronic hints and repair briefs , schematics/manuals list on
Adrian C's explanation sounds quite convincing.
Graham
Statistical anomalies are simply a fact of life.
It's why if you toss a coin 100 times, it's no freak if it comes up heads 75 times and tails 25 times for sake of argument..
Graham
Vibration....
It is a sad ;-( moment when the feast of Christmas has finished, so the lights are gently, carefully and slowly placed into the original wrapping (cardboard cutouts to hold each lamp are a bit fiddly to do in a rush). When the time comes to have them out again - bash, crash, wallop, ping....
-- Adrian C
Maybe the same reason that it`s always the negative wire from a battery pack that corrodes away?
Ron(UK)
Ok Arfa,
I think now that I have posted physical theories to the last two (plugtop; xmas light) I might have a similar *physical* answer to the diode anomaly.
Sinking away heat from the diode PN junction is more efficient with a large amount of ground plane to connect to which is the case with the negative half of the bridge. These diodes will run with a higher Vf and hence a higher power dissipation than the remaining diodes on the bridge given that both positive and negative halves will be conducting the same current. Where fault conditions exist on the DC side of the circuit, these harder running diodes may be the first to fail?
Dunno.
On the other hand, I tried looking for conditions for thermal runaway in bridge rectifier diodes and came up with the following link
but can't say from that which half of the bridge would be suceptable to shorted failure.
-- Adrian C
I think it was actually Mr Cook who ventured the detailed explanation, with Adrian's being just "vibration", but yes, it does indeed seem reasonable. So you see, with a bit of thought, not everything is as black and white as you suppose. Before seing an explanation that appealed to you as a possibility, you had already started to rubbish me on that one too ...
Arfa
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.