Re: Disc caddy

Andy Hewitt <wildrover.andy@xxxxxxxxxxxxxx> wrote:

Rowland McDonnell <real-address-in-sig@xxxxxxxxxxxxxxx> wrote:

It's a rubbish one, that's what.

Yeah, but more rubbish than I'd imagined could be possibly made.

Such a bloody simple device, and just
about the only moving part on the thing, and it's crap.

Not just crap, but gratuitously and unnecessarily so.

Absolutely, and I won't be buying any more Acoustic Solutions products
(I also had a DVD player that went back for refund after 4 months, the
spindle in the drive motor kept slipping, and causing the discs to jam
on the platter).

I had one of their later inkjets (after the breeze-block models), and
wasn't too impressed I have to say.

We have an HP970cxi. It stopped working properly a very long time ago.
No two sided printing. An internal part snapped and flew out in normal
operation - the printer was just sat there, printing away, and then came
the nasty noise and I found a bit of snapped plastic. And then ISTR it
packed up completely. I think the much older HP DeskWriter 520 is still
working fine - it was when I retired it.

My next printer will be something other than an HP... Well, I say that,
but the performance of the HP LJ 1320n on the filing cabinet behind me
as so far been unimpeachable just so long as you don't mind the PS
emulation giving that 'orrible graphics banding. It was bought for text
output, so I don't.

On the other hand, the HP supplied software caused nothing but hassle,
HP's customer support is fucking *awful*, including the information it
supplies for setting up and troubleshooting.

I mean, I had a problem caused by faulty instructions from HP. So I
told 'em about it. They explained that what I'd run into was a
well-known problem, that I should not have followed the instructions on
HP's Web site because they were wrong, they told me what I should do
because it was a standard issue with that particular software and Macs
that they knew all about and had done for ages.

And they told me that they'd updated the information on the appropriate
Web page, where I'd got the wrong instructions from.

I checked.

They hadn't changed a damned thing - just plain lied to me.

<shrug> *** 'em.

So far, my take on the quality I've experienced with printers, in order
of preference:


1) HP Laserjets - we have a 5N and a 6M (at my Mum's) that are lovely -
as you say, not good for graphics, but great for text. Toners can be got
very cheap.

2) Canon inkjets - very impressed with the print quality, and value for
money, consumables reasonable, built quality not as good as I hoped
though, this one is already (18 months old) suffering with banding in
photos. Prints a duplexed booklet without any problems though.

3) Epson inkjets - superb print quality, especially photos, but the
heads are far too prone to blocking, and although the consumables are
cheap enough, they need to be to cope with the head cleaning you need to
do.

4) HP inkjets- poorly made, poor print quality

Lexmark (all) - complete ***, poorly made, and *very* expensive
consumables.

I've also had experience of various printers in my work environment -
usually needing to work on about 2-3000 pages a month. I have been very
impressed with the Kyocera laser printers we had at one time, they just
kept going and going. The Oki 24 pin dot matrix printers we had were
almost indestructible (just about every MOT testing machine has these on
them, even now).

The 7100 was built well - I had one, and the innards were built like a
Brunel bridge. I think the next generation went the wrong way though for
sure.

As soon as IDE HDDs turned up in Macs, I got suspicious. Solid
mechanical construction means nothing regarding the quality of the
electronic design, or the quality of the electronic and
electro-mechanical parts.

Indeed. Most of the 7x00, and all of the 8x00 and 9x00 series were SCSI
based (according to MacTracker).

[..]
Hmm, looks like a general tendency for the GPU manufacturers to use
*** fans then.

Reckon so - a bit rich, given the prices they charge.

Indeed, and an upgrade cooler of better quality isn't that expensive -
expecially when you consider the difference in reliability.

<puzzled> Erm. Are you saying that it's standard practice to replace
these fans? If so, how does one go about such a job?

Pretty much, yes - at least it's sugegsted quite a lot in many forums
I've seen elsewhere (in particular in the Apple 'Discussion' forums)

It was easy enough on mine. I fitted a Zalman copper cooler kit, for £25
you get a lovely copper/aluminium cooler assembly with a silent fan (can
also be wired to run at one of two speeds). On my Radeon card (after
unpluggingm the card from the Mac of course) it was a matter of undoing
two screws, and unplugging the fan, cleaning the old thermo paste off,
apply the new (comes in the kit), and screwing the new until in it's
place. The new one came with a couple of mounting brackets to fit
different hole spacings. Took about 10 minutes.

Mine also came with a set of VRAM heat sinks too, which just stuck onto
the chips.

I didn't wire mine back to the onboard controller, but wired it with a
'Y' connector to the DVD drive supply.

I got mine from here:

http://www.quietpc.com/gb-en-gbp/products/nvidia-vga

(what I'm thinking is that if it's cheapish and easy enough, I might
just replace the dodgy fan myself rather than go to the bother of
sending it back and running the risk of further damage, etc)

IIRC, that is something that should easily be within your abilities :-)

[..]
It's spring now. ;-)

When I posted that, there was snow falling outside here :-(

<grin> Sounds like spring to me.

Well, we know it's summer when it floods.

We don't get floods - it's just that the entire area seems to have been
marsh before they built the houses, and the gardens keep trying to
revert. Soggy, but the nearby river is a good 20 yards below ground
level at my front door.

And the various dips and hollows in the area that *do* flood don't wait
for summer. One thing, though - down by the railway station and right
next to the river I referred to, the water company installed a huge
underground tank as a flood buffer.

I don't like stuffy rooms. Windows stay open here - well, some of them
- unless it's very breezy.

That's quite a bit of the time here in Scarborough.

I'm on the other coast - we get wind too, but I'm a few miles from the
actual edge of the land so it's probably less bad.

I'm about a mile from the North Sea. Very soon we'll probably start to
get the 'Sea Frets', and won't be able to see across the garden after
about 5pm.

Ah. I'm - erm - maybe 8-12 miles from the Irish sea, but rather closer
to the Dee and Mersey estuaries in their tidal parts. We get the fresh
air without the worst of the coastal battering.

That makes no sense to me. Graphics processors don't use that much
electricity compared to CPUs. Electrical power going to heat is what
matters with this.

There are reports of 600W PSUs going pop after fitting the X800XT cards
into older G5s.

??!!??

????!!!!!!!????

How much power do the things use?

No idea, but it must be substantial (for a graphics card).

`For a graphics card'? Well, that's not really the point: it's `for the
PSU' that matters.

Mine went not long after fitting the 9800XT.

I'm astonished. No properly designed power supply should permit itself
to be overloaded.

That's the problem though, these aren't.

I don't see how they could have designed 'em without protection of the
sort I describe. I can understand that the protection might have failed
- but not putting it in at all? That's crackers.

I'd /assume/ over-current, over-power, and over-heating protection as a
matter of routine in any PSU - and maybe an over-voltage clamp to shut
down the PSU in event of an internal fault to stop it frying what it's
feeding.

The G5 PSUs are actually a known weakness, and that coupled with the
higher demands of high performance upgrades is most likely exploiting
the weakness that's already there.

But the damned things should be nigh-on indestructible with respect to
load. It's quite easy to design in the protection I describe, and one
of the things you're supposed to deal with when designing a PSU is to,
well, deal with that sort of thing.

<shrug>

I dunno. I mean, it's not like the firms that sell the chips that
control PSUs don't publish sensible designs on the flippin' data sheets
and in the application handbooks.

There is an Apple Care cover for some models of G5 PSU failures (of
course mine wasn't covered when it went bang) - a poor quality batch of
capacitors is the problem I believe.

Which is hardly a design fault, is it?

[..]
It's the ones that are getting upgrades that seem to be suffering more,
i.e. fitting a card that was not originally intended for the machine.

You really mean fitting a card that was intended for the machine because
it's a graphics card built to go in a slot of that type.

Er, yes, although cards of such performance weren't around when the G5
was originally designed.

Irrelevant: all that the PSU cares about is electrical load.

I'm only guessing, but these newer cards could
be exceeding the original design specifications.

<puzzled> With these buses, surely there are power supply specs?
Unless the card exceeded the bus spec, there should be no problem - and
if it does, the only problem that should occur is the PSU failing to
supply enough juice and possibly shutting the machine down in protest.

We know that Apple are
quite good at cutting corners - usually where most users aren't going to
notice (upgraders are still pretty much in a minority).

Remember, the early G5s have a poorly implemented SATA controller, and
problems with USB2 data rates.

I don't know the details in those cases.

[..]
Yes, seen things similar to that, and was most likely what was inside
the old system we used in 1980.

Nothing higher-tech was available then, was it?

Probably not. At least not at a level that was affordable in public use.

I think surface mount was still experimental back then. I don't think
anyone had started to investigate pin arrays and those natty solder ball
thingies. Of if they had, there was no sign of it in consumer products
(what's the betting that sort of thing was in military radar come 1974,
eh?)

That's what I was thinking - worth noting that the 'stealth' planes were
flying in 1982 (just as an example of how advanced the miltary are with
technology), yet we didn't see them (sorry) until much, much later.

Stealth planes were flying long before 1982. The first modern stealth
plane was the Blackbird series of spyplanes (and the interceptor
prototypes) and I can't be bothered to look up the date, but wasn't it
the late 1950s that the first of 'em took to the air?. The F117A
stealth fighter was preceeded by a bunch of curious stealth flying
testbeds.

You say `affordable in public use', but one of the reasons for the
half-sized data buses available on the outside of many late
microprocessors back then was because it was bloody expensive on the
packaging and in any case, 0.1" pin pitch DIL packages of huge length
are quite easy to break. Expensive and fragile is what those old DIL
packages microprocessors were.

Surface mount's cheaper, as are all the other fancy `hundreds of pins'
packages we've got these days.

Aye, but new technology always commands a premium price, just because it
is new technology, it doesn't matter if it is cheaper to make.

<puzzled> Nonsense. The whole point of a great deal of new technology
is simply that it's cheaper - in many cases, that's the /only/ point to
it.

(E.g., Jap bikes are all modern technology; Harley-Davidons are a bit
old-fashioned with respect to tech. And you get a better bike from the
Japs, with more of everything except weight and cubic capacity. Okay,
okay, the V-Rod and associated bikes are pretty good engine-wise. But
it's a German engine - as far as I can tell, H-D has made one serious
upgrade to its standard basic engine design since the 1930s and they
called that the `Evolution' engine, to try to emphasis how small the
changes were... At least Triumph added an extra cylinder to its 1930s
design (I lust after a T150V Trident - with modern carbs and tyres,
thanks), and tried to get a Wankel working (IIRC, the engine which got
used in the Norton Wankel range was originally an NVT project)).

I mean, a 1340cc Harley will do about 110mph, yes? About the same as my
wife's GS500 or well-sorted Triumph 500cc bike of the 1950s. And a Jap
bike of similar size to a 1340cc H-D - let's pick the Suzuki Hayabusa
with its slightly smaller engine purely at random <cough> - can do about
200mph with the speed limiter removed.

Surface mount is cheaper and smaller and more reliable than what we had
before. Those solder bump array packages are one of the cheapest, most
reliable, and generally best ways of fastening a lotsofpins IC to a PCB
as far as I'm aware.

If you want lots of pins, you just can't do it in any practical fashion
with an olde-tyme DIL package. I handled 64 pin DIL packages - oh, you
wouldn't want more, you really wouldn't, but it was do-able. But do you
get it? 48 pins of DIL is unwieldy; get up to 64 pins, and you know
you're approaching the giddy limit.

Okay, if we're talking surface mount, the first neat gadget you design
and flog with surface mount in it might well be sold at a premium -
well, why not? It'll be so much smaller and (culturally) cooler than
the older competition, that your sales will be assured even at the
higher price, thus increasing your profits. And you'll have a new
production line to pay for, too. But once someone else has a surface
mount gadget in competition with yours, the price comes down very fast
because it can come down an awful long way.

But here's another example: back in the old days before ICs, a firm sold
an operational amplifier with staggeringly low input currents. My other
half has the book at work, so I can't look up the details - but I do
recall the price charged for this operational amplifier was $200, and
they were flogging them in the 1960s. The parts cost was somewhere
around $5 or less, IIRC - as everyone who bought one new damned well,
but they were so much better than anything else anywhere that they paid
up not exactly happily, but with grudging respect (grudging, 'cos they
were getting badly gouged).

IC op-amps came out, and as soon as people worked out how to use them in
circuits with similarly low input currents, the discrete transistor
monster price device was instantly obsolete due to being far too big and
a couple of orders of magnitude too expensive. Although I think by that
time, it was out of production because the firm that made it couldn't
get the parts any more, and couldn't re-design it to use new parts
because the designer had left the firm years since, but his design was
so incredibly clever that no-one else could quite figure out how it
worked.

(it used a parametric amplifier which depended on, amongst other things,
a phase shift caused by the separation between the two PCBs, and that
was just one of the things they did figure out on the way to failing to
re-design it)

They are
cheaper now, because they are more common.

<puzzled> Eh?

Rowland.


--
Remove the animal for email address: rowland.mcdonnell@xxxxxxxxxxxxxxx
Sorry - the spam got to me
http://www.mag-uk.org http://www.bmf.co.uk
UK biker? Join MAG and the BMF and stop the Eurocrats banning biking
.

Loading