[rescue] Slightly OT: Bad Cap Saga

[Curious George]

--- On Mon, 8/18/08, Geoffrey S. Mendelson <gsm at mendelson.com> wrote:

> On Sun, Aug 17, 2008 at 03:22:07PM -0700, Curious George
> wrote:
> 
> > Can we get terms straight?  Who's the "manufacturer" in
> > your descriptions?  I see many manufacturers in a typical
> > product -- most of the board assemblies are manufactured
> > by one of more *vendors*.  The mechanical assembly may be
> > done by yet another house.  It may then be rebranded and
> > sold to the "Manufacturer" from which the consumer buys (note that
> > a consumer need not be John Q Public).
> 
> Each one is a manufacturer, they do something to put it
> together. They buy a raw material, whether it's aluminum
> foil to make the capacitors, or
> the assembled boards and wire them together.
> Each step along the way, someone builds, and possibly tests each unit. 

Yes, that was my point...

> Generally no one sells their product with a warranty.

Well, we know the *final* manufacturer who sells to John Q Public
does!  Are you saying that the rest of the chain is all "caveat
emptor"?  So, the poor fool who walks into a Wal-Mart and "takes
a chance" on some cheap piece of kit is actually getting a
*better* "deal" than the rest of the folks in the chain?  :-<

> It's too costly to sell a mass produced item with one and
> expect them to come back for repair.

You don't have to "take it back for repair".  E.g., most
of the items that I've designed are *examined* when they
fail "under warranty" (i.e., to figure out *why* so we can
determine if there are component or design issues to be
remedied to reduce those costs in the future)

> They either provide extra components, or a "reduced" price
> based upon the anticipated failure rate, or tell you "what you
> get is what you get", and plan accordingly.

How (seriously) do you come up with an "anticipated failure
rate" when you are using components outside their specs?
Rely on past similar "abuses"?  :-/  It would seem too costly
to *test* a marginal design too try to come up with a reasonable
confidence interval for this figure -- especially with the short
product development cycles nowadays...

> > At some point, "you" (whichever of these "manufacturers" you happen
> > to be!) have to have some faith in the quality of "your" product.
> > Or, are you saying their are manufacturers who just slap things
> > together and never bother to see *if* they work??  :-/
> 
> Exactly, the cost is too high.

OK, I can buy that for "cheap" consumer kit.  (e.g., $100).
But, are you saying LCD monitors are just "produced" with
no testing or other means of quantifying your quality?
Ditto for PC's?  (I'll have to start looking at some of the
"failed" medical instruments and see if they are suffering
the same sorts of "bad cap" failures!)

> Some companies do it, for example
> Tandy did it with the original TRS-80 (and it may have been the
> last time too), and Ten-Tec still does it with their radios.

The common thread in your examples seems to be "cheap" (as in
"inexpensive" and "not generally thought of as high quality")

> I see articles in Tele-Satelite magazine showing
> manufacturers in the far east who check their units,
> some every one. But that's a different
> market than the 10 UKP DVD players ASDA had for Christmas
> last year (or was it 2006?), 

But $20 DVD players are different than $200 LCD monitors
or $500-1K computers!  People almost *expect* the failure
("Well, it was only a $20 player, anyway... let's go buy
*another*...").  I'm not sure that extends to *everything*
made "for the (John Q Public) consumer".

> > I've seen PC power supplies (covered by the PC's warranty),
> > motherboards, LCD monitors, "flat screen" TV's, etc.
> > all with the same sorts of problems.
> > Note that none of these are "disposable" consumer kit.  Most have
> > price tags high enough that the end user *will* be annoyed by "early
> > failures".  (e.g., the three LCD monitors I fixed today were from an
> > institution user -- if they are seeing large number of failures
> > you can bet that information gets back to the manufacturer... in none
> > too pleasing terms!)
> 
> Sure, what do they do? They complain to the board manufacturer who 
> might do something or not. Most likely they will give them a few
> extra boards for free, or promise not to do it in the next batch,
> or most likely, since they stopped making those boards a year ago or
> longer, just smile and say "sorry".

So, in effect, these folks (the people who ultimately sell the
items to you) have placed their company's fate almost entirely
in thehands of these suppliers -- and are *comfortable* with
that relationship?  :-(

Every time I've had to meet with a Japanese supplier there have
been *lots* of reassurances ($$) exchanged prior to beginning
any sort of serious relationship (since this was potentially
a significant "dependancy" being established... if the Japanese
company just saw us as "one of many customers", we weren't
very reassured since an entire product line could end up
depending on how faithfully that Japanese company held up
its end of the arrangement!)

> > And I am sure they would gladly sell you a *PC* today and again
> > in 6 months!  Problem is, would *you* be willing to buy it *and*
> > accept the fact that the 6-month-old unit was just "worn out"?
> 
> Some people do that. Often not in PC's, but in consumer
> electronics, all the time.

But, these same problems are manifesting *in* PC's.  So, why
is it OK (to the cheap DVD seller) to one seller and *still*
OK (to the PC vendor)?

> Look at the public relations disaster Apple had when the
> first generation iPods failed after a year due to battery
> problems.
> 
> Now how many people keep their iPods for a year? What about
> the cheap MP3 or "MP4" players that are everywhere? Does
> anyone care if they die in a year? It's cheaper to go to
> Wal-Mart and buy a new one with more
> memory, more features, etc and toss the old one.

Again, these are inexpensive devices.  What happens if the same
quality issue presents with large LCD TV's?  Or, PC's?
I wonder how that sort of problem would be met here in the
US if it afflicted these new "HDTV converter boxes"?
(imagine 1 out of 5 ATV user losing their TV reception because
of a flaw in that design... (I'm told several boxes have
noticeable firmware bugs))

> > Sure, that's common practice!  It's still a losing proposition for
> > the "manufacturer" if "enough" units are returned.  (there are costs
> > involved in handling the return, complaint, etc. -- many of those
> > can't be contained... e.g., a lawsuit pops up and you suddenly
> > can't factor the cost of *that* into your $75 fee!
>  :> )
> 
> And exactly what do people sue you for? It's hard to
> sue someone because a gaming device failed out of warranty.
> Look at all the threats Apple got after the iPod battery
> fiasco. How did they resolve it?  You could return your iPod
> for battery replacement for $99. How much did that cost
> them? How much money did they make on each of those
> repairs?

I seem to recall Gateway having a big problem with Plasma
screen TV's -- clearly a "consumer item".  Yet, *they* were
sued for failures that were "out of warranty".  And, I
believe it was probably a big issue in their demise.
I think this reinforces my "what is 'cheap'" notion as
a threshold for pain for most consumers.

Perhaps if iPods cost a few kilobucks, Apple's response
would have been less palatable (for all parties).

> How many people did not bother and bought new iPods, or
> replaced the battery themselves, absolving Apple of all guilt.
> 
> > And, it depends a lot on who your customer is and what your
> > relationship with them happens to be.  For "consumer" (John Doe)
> > product, you can afford to screw the user as, for most companies,
> > you'll probably never do business with him, again.  Or, he won't
> > remember/know he's doing business with the company who previously
> > screwed him, etc.  (or, he may be naive enough to believe the
> > product was *supposed* to only last N months... )
> 
> Well, with cheap consumer goods, it's exactly that. 
> 
> > OTOH, if you have long term relationships with your customer
> > (recall, customer may be a *company* -- even one that remarkets
> > your product or incorporates it into something of their own),
> > then you really want to make sure that relationship sours as
> > a result of them thinking they *were* screwed.  :<
> > 
> > <shrug>  In my market, it's easier just to Do The Right Thing and
> > not chase customers who follow the low-ballers around...
> 
> I assume you are not a member of the Walton family. :-)

No.  I just don't like threats of lawsuits when a device fails
and a customer's production line goes down for several *hours*
(note:  not *days* or weeks!) and he's looking at how he's
going to recover those losses -- and *you* (me) are the only
likely candidate!  :<