...*unable* to ship (via) US mail?
That is not believable. USPS will pickup and allows accounts like
just like FedEx, UPS, etc.
Furthermore someone there could just as easily deliver packages to a
local USPS office.
So there is really no excuse to *not* use USPS if a customer wants that.
To make a profit, *LISTEN* to the customers AND keep the customers happy.
On 9 Aug 2006 16:07:44 -0700, "Rebecca of Maxim Dallas"
It's getting to be conventional wisdom among EEs that one should avoid
designing Maxim parts into products. It doesn't help that pinouts are
deliberately made non-standard.
I've been personally burned four or five times. At this instant, we
can't ship over a million dollars worth of product because Maxim has
slipped their promised delivery date on MAX5205's. Do you know where I
can score a couple of hundred?
I've been having friends and family request samples, which are
available. Every 8 dacs we can scrounge, we can ship a $54,000 laser
I'm curious which parts you think have deliberate non-standard pinouts.
I can tell you in the decade I worked there, this was never an issue.
Pinouts are carefully chosen to minimize noise, ease PCB layout, etc.
To keep costs down, Maxim doesn't make a die per each package type (in
general), so pinouts are designed to work across many package types.
[Back lap specs change from package to package, so often wafers get
allocated to one particular package at the expense of another, which in
turn causes shortages in a particular package type.]
At no company did I ever deliberately make pinouts to be
non-standard. If anything, if a part exists already with defined
pinouts, you make it a point to use THOSE pinouts in order to steal the
sockets with your hopefully superior product.
For one chip I did at a company I better not mention, a customer leaked
us a datasheet of a competitors part. We actually changed our part to
match their pinouts since our part was still in design. They had an
interesting feature that I added as well. As luck would have it, our
part came out before their part, but with the competitors pinouts and
features. While we were doing high fives, I suspect somebody at the
competition was pounding their fist on the conference table.
Then the problem seems to be that design engineers do not think the same
way he does. Usually that isn't a good thing. Most of us never design in
a sole-sourced part unless there is no other alternative. Usually there is.
Did you model your laser modulator to be plug compatible with another
company's laser modulator? I'd hate to think each company had their own
standard. That would raise the price, and making money is bad.
On a data converter, I'm guessing a good wafer run yields about 10k
parts. Profit should be about a factor or 3 or 4, so break even is only
say 3000 parts. I can't see the factory turning down such an order.
BTW, I still own stock in the company and am certainly not pleased with
the performance since I know it is something easily fixed with a good
purge at the top. I can name the names of people running the show that
really have to go, not that I think it will happen. I've read the
posting on Yahoo and just laugh at how accurate they are, especially
regarding the fab.
Note: the Peter principle is not a theory, but a fact.
Maybe TI will buy the company and clean it up. It woud take very little
effort to make Maxim a money making machine again since an outside
company would get rid of the top level of management, which is 90% of
On 11 Aug 2006 20:56:58 -0700, firstname.lastname@example.org wrote:
It's a rackmount chassis full of timing, gate pulse generators, PLLs,
self-test, multisection lithium niobate modulators, power supplies, a
4 gs/s arbitrary waveform generator, ethernet interface, and
distributed amplifiers... full custom.
They didn't turn it down. They promised 12 week delivery and didn't
deliver. They slipped it another 4 weeks. When Maxim does that, we get
scared, because additional slips often follow.
I always liked Burr-Brown for their quality, integrity, and
availability of parts. I was afraid the TI would mess them up, but
they haven't. TI is deadly serious about analog, and they are doing it
I am using a number of Burr-Brown parts (and a lot of TI parts too) in
various designs, and I'll agree TI is deadly serious about doing it
Maxim should be *very* worried, because TI is second sourcing a lot of
their RS232 parts, amongst other things.
I get their weekly e-newsletter and it seems a new one or two are added
On 12 Aug 2006 09:47:59 -0700, the renowned "PeteS"
I actually designed one in a couple of weeks ago. But only because TI
had it. So maybe it could possible benefit them, but only if they can
compete with TI on price and have the parts available.
"it's the network..." "The Journey is the reward"
email@example.com Info for manufacturers: http://www.trexon.com
You can have rework on wafers or some machine needs repair, but
generally wafers won't slip 4 weeks. A run gets outright scrapped. When
it slips a few weeks, the problem is nobody is watching the product on
the backend (Packaging and test). As I've said before, these are
problems that can be fixed, but you need to remove some totally baffled
people that should have long been fired or retired. I have little
tolerance for manufacturing screw ups, assuming the design is robust.
When I do a board level design (generally for myself ;-)), I use TI
parts since I can get them from Digikey or Mouser. TI is the last big
company that has any respect for analog.
And then the next worry is whether that could be the final slip into the
land of Unobtainia.
They seem to take that part of biz serious. I know an engineer who works
in what they call the Burr Brown group, cranking out new stuff. Pretty
encouraging. In the days before TI acquired them I usually shunned Burr
Brown because their pricing was often sky-high. Now it's much better.
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.