Originally posted by Kguru (a) In my experience low-cost light materials mean low durability.
In my experience as an engineering product manager ...
- Cost isn't just about materials
- The weight of materials doesn't not translate into sturdier materials (especially not in the past 5-10 years)
- Pentax's fab in Vietnam is actually newer, so it sports newer technologies that are now "commodity"
Cost is more than just about input, but many aspects of Quality Assurance (QA).
You can ...
- Use lower costing components of less exacting tolerances and still get most excellent results
- That requires you to discard samples that are poor results of the process.
- How many samples depends on how you define your tolerances and allowances
From what I've seen, Pentax is ...
- Getting good yields of excellent quality, despite using commodity components of low-cost in the process
- Setting their tolerances and allowances too low, which causes them to ship 20-25% of samples that are less favorable
That might be a factor of their QA decisions. I.e.,
- Pentax may only look for major defects and discarding those
- Inconsistency and "harder to notice" IQ issues may not be tested, because additional testing costs money
QA can be the "biggest cost" in any product fabrication -- especially the more testing it involves.
E.g., Even though NASA has switched to commodity hardware and software since the '90s, QA costs are still high.
NASA once tried to slash QA budget 90% like they did hardware and software development, this was the infamous Mars Polar Lander.
The result of killing 90% QA was that one contract software developer in Texas was using metric and another software team at another contract in Colorado used Imperial units -- in two different routines that eventually interfaced.
It turned into the biggest embarrassment in NASA's history.
Post-review showed that basic engineering and QA practices were utterly ignored, because of the sheer cost to implement them.
Originally posted by Kguru (b) In similar FL & price bracket current Sigmas and Tamrons weigh nearly twice as much as the 50-200.
The DA 50-200 is already a
very, very small and compact unit to begin with.
I don't care what material they used,
the design is very small.
So your "weight" argument is already biased in that regard.
Originally posted by Kguru Does that mean Sigma & Tamron are so far behind in lightweight technology, or because Pentax use lower durability (compared to S & T) materials?
It could be many things.
But yes, Pentax's Vietnam fabrication facility (shared with Tamron I believe?) is actually circa 2003 technology.
The massive strides in polycarbonate and other optical technologies would easily blow away an older fabrication facility in cost v. performance.
New fabrication facilities are expensive to build (and likely a reason why Pentax shared the cost with another company).
Just ask AMD. They just can't keep up with Intel.
Back when Intel had not done a complete architectural redesign of x86 since 1994 (NetBurst was not a completely new design, just a "refit"),
AMD's 5 year newer architecture (1999/Hammer v. 1994/P6) kept them ahead, even though they were 18 months behind Intel in fab.
Now that Intel actually redesigned a full and new architcture (2006/Core), AMD's lack of being anywhere close to Intel's fabs really hurts.
Originally posted by Kguru (c) If Pentax has created this lightweight yet durable material for the 50-200 why don't they use it for lenses like the 31mm for example, to make them less heavy?
Redesign costs. You can't redesign things overnight.
Furthermore, older fabs still have their uses, for old designs that still sell well.
So not only is it that you can't redesign things overnight, but you still have old facilities that can still make those older designs.
Likewise, AMD and Intel also have older fabs, where AMD makes EEPROM and Geode NX (embedded Athlon) and Intel makes EEPROM and XScale (embedded ARM) at 90-130nm processes.
But as Pentax redesigns new, prime lenses, they will ship more, cheaper, commodity polycarbonate designs.
It will be interesting how they balance fab, QA and other costs on their DA* lenses.
Originally posted by Kguru PS: By durability here I don't mean against breakage, I'm talking about micrometre wear that can easily upset a good copy.
I know.
And if that is the case, then Boeing wouldn't stop using Aluminum, let alone Titanium, in favor of new composites.
I mean, we don't still use Titanium all over aircraft, only limitedly.
It's more than just about cost, a lot of it has to do with better materials for the job.
And the techniques to do so are now commodity, with more on the way.
Engineering product management 101 here.
It costs real money and time to build modern fabs, redesign components and the fact remains that you have older fabs and older designs that are still "good enough" to pump out products.
Not everyone always works on the "sexy, new" product, and it's often more difficult to maintain the old one.
As I always point out, NASA's greatest engineering lifecycle achievement wasn't the moon -- the Apollo "product" lifecycle of only 5 years.
NASA's greatest engineering lifecycle achievement was the Space Shuttle -- a product of, now, 30 years.
NASA actually had zero (0) design failures on the Shuttle, a system of 100,000 parts.
What NASA had were two (2) (actually more than 2) "material review" (MR) QA oversights.
One was when the sealant on the O-Rings were changed in 1985 -- a change due to new EPA regulations on the chemicals used.
The MRs failed to catch the fact that the O-Rings were now suseptible to weather and environment, where they were not before.
Unfortunately, no one sold the old sealant, and that's when it came down to an availability/feasibility issue.
The engineers did not properly assess the risk, and the result was proven in the sheer statistics of launches from 1985+ (but not before the change).
The other was when the insulation on the external tank was changed in 1997 -- also a change done by the EPA (back in 1987 which took effect in 1997).
The new, CFC-free insulator has severe tensile strength issues, and can "break apart" instead of holding together.
This is an extreme issue for not merely just the Shuttle, but our ballistic missile arsenal which uses rope to "pull off" insulation as the missile rises.
I've seen this crap first hand in White Sands, and it's scary that we can't use CFC insulation anymore where tensile strength is a requirement.
The EPA has given NASA an exception on CFC insulation since 2003, but just like the sealant prior, you can't find CFC-based insulation these days.
It was easy to find suppliers when a design was only 5 years old.
When you have a product that is made for 30 years, materials become a real issue.
NASA only has so much money to throw QA at a problem, and the Shuttle is getting too old and needs a complete redesign.
Unfortunately, Project Orion isn't it.
Prior, NASA thought it was the Lockheed Martin VentureStar prototype.
But that was canceled because the cryogenic tanks required lightweight materials not invented yet, and NASA eventually gave up (and Lockheed stopped using their own money).
That was not even 10 years ago now and, ironically, not only have they invented the new materials in that time, they work better than aluminum now.
Which is why many in NASA are calling for a refocus on VentureStar, and a cancellation of Orion.
This is just sample insights into the "real world" of an engineering product manager -- from materials to QA to lifecycle and product lifetime.