Any Maxim discussion Forum Members?

You can always get samples, even when you can't buy parts. So sometimes we have friends+family from all over the country request the max number of samples, which is something like 8.

They deliberately use non-standard pinouts, so you can't use anybody else's parts when they won't deliver.

Customer abuse is a strange business model.

John

/IME

I had a one of their reps in the other day who *claims* they're changing the way they do business. This was after he asked about companies he shouldn't even bother us with.

Yet they've been using that business model for decades.

In all this time it has been a weirdness of Maxim. It is like part of management is incapable of thinking of production runs of less than an old style "boat" load of about 50 2 inch wafers. With modern 8 inch and 12 inch lines a single wafer in the production line produces a boat load of devices. Moreover the processes have changed so much in the past 30 years that essentially each wafer can have its own process.

ld

on

es,

ls

fo

ps

at

t,

way

othing

none

x has

of

A proprietary part has whatever pinouts the engineer picks. I suppose you could make an exception for opams and comparators. However, there are all sorts of issues in choosing pinouts. You might want to shield an input with a ground or power (ac ground) pin, stuff like that. There is also the issue of making a die fit multiple packages. Sometimes for performance issue, a custom packaged is used.

And they still all get all of Maxim's mailshots to this day, right? :)

They were one of the first manufacturers to offer easy samples, for small customers starting up like myself. And that got them their design-ins here. But I have been designing them out again for the past

hing

=A0 =A0 =A0...Jim Thompson

I have left in the lurch - several times, as have friends. Perhaps we should ask how many engineers have used a high-performance Maxim part, one not replaceable by another manufacturer's product, and _not_ had it discontinued?

Maybe we should say "prematurely" discontinued, but whenever this has happened, it's been a disaster.

Some manufacturers, like Analog Devices, solve the discontinued-part issue by make a large quantity of parts and selling (?) them to Rochester Electronics.

I read an interviwe with the CEO of Maxim, where he said the unique pinout thing is deliberate.

John

That's truly ignorant. Unique pin-outs cause NO ONE to consider your part as a drop-in replacement. Sheeeesh!

...Jim Thompson

hould

n, a

ge on

on

lines,

tools

info

had

er

apps

ge at

f it,

ed way

t nothing

use none

ing

d at

;) )

onix has

ere

rt of

at

=A0

can

Which CEO? Do you have a link?

=A0 =A0 ...Jim Thompson

=A0 | =A0 =A0mens =A0 =A0 |

=A0 | =A0 =A0 et =A0 =A0 =A0|

=A0|

=A0 =A0 =A0 |

If you are doing a second source or an improved version of an existing product, you copy the pinouts.

I seriously doubt Maxim is scrambling pins.

ld

on

es,

ls

fo

ps

at

t,

way

othing

none

x has

of

I used Intel as a foundry (a long story). I know about each product having it's own process. (In the day, a factorial was run and the best process was determined from the factorial analysis/) Few companies can tweak the process to the chip. It's hard enough for these people to get it right just following the standard formula.

Maxim never had 2 inch wafers.

Maybe someone can post an example of scrambled pins? A quick peek shows pin-outs consistent with industry standards.

...Jim Thompson

Nonsense. Pinout is done by the layout guy, usually aligned as requested by the end customer.

...Jim Thompson

pose

, there

ield

your

=A0 =A0 =A0 ...Jim Thompson

=A0 =A0 =A0 ...Jim Thompson

=A0 =A0 ...Jim Thompson

=A0 | =A0 =A0mens =A0 =A0 |

=A0 | =A0 =A0 et =A0 =A0 =A0|

=A0|

=A0 =A0 =A0 |

It depends how you work. Many engineers chip plan before it gets to layout. There are often performance issues. For instance, the mask designer doesn't know which pins have the lowest inductance. The mask designer doesn't know which pins needs the most isolation. For instance, the LX pin (flying end of the inductor on a switch mode chip) shouldn't be next to the reference pin since there will be C*dv/ dt coupling. For a power device, there are electromigration issues that wouldn't be considered by mask design. I could go on and on, but pinouts are the responsibility of engineering if you do it right.

For standard products, the end customer doesn't even exist technically. Now granted, you have an idea what they (the market) can use, but you won't be sending them a spec unless you are a complete fool. (Non-disclosure means nothing to most people because they know you won't waste your time suing them. Remember, you have to prove damages to sue.] If you are doing your job right, you design a chip that the market didn't know it needed.

Same here, lots of design outs, no design ins. The usual trigger is a frantic call from a client in a line stop situation. Because they can't buy the parts.

FWIW, non-standard pin-out is IMHO a bad idea. I will not design in a part with non-standard pin-out unless there really is no other choice. And I usually find another choice, that's what my clients pay me for.

Nor could it be said any more clearly.

can

I guess you would be surprised the chips I did there would do from a few hundred K per year to just under a million units. [The big sellers would do millions per year. Kind of luck of the draw as to what was assigned.] Clearly, parts are shipped. I wonder if you guys are getting stuff designed by Gain Technology. I have no idea what the deal is today, but when I was there, Gain Technology would design basic stuff like comparators, op amps, etc. Let's just say I wasn't impressed. I can't say the Arizona and Texas designs are up to par with how things were done in the valley.

But did you have a non-Maxim option for a comparator like this? In the early 90-s I used their MAX901 (just TTL, yours must have been way faster - this one was 10-20 nS or so, quad) and it was the only such option on the market by a great margin.

Around the same time, I used their MAX500 quad DACs - again, they were way ahead with these in terms of usefulness for my purpose then.

I must confess I have not designed in many of their products lately myself for the same reasons everybody else here states (hard to get, too unique); but are we not being unfair to them, they must be much smaller than say ADI and NSC.

Having been able to seduce so many of us with unique parts they have designed is not such a bad thing, I suppose. Sure, there have been flops - I remember I designed in a power supply monitor, MAX791 or sort of, and had to bypass virtually every function it had to make my board work, removed it from the second revision and replaced it with discretes... :-) .

Dimiter

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

------------------------------------------------------

Original message:

What's with this sudden burst of anti-Arizona and anti-Texas? Did poor-little-old ex-Maxim employee get his feelings hurt ?:-)

I don't know that anyone impugned Maxim's designs, just their shoddy delivery... n'est ce pas ??

As for Gain Technology, I don't know much about them... they approached me once, quite a few years ago, to do some designs, but apparently didn't like my pricing ;-) (And they're just a few miles from me.)

As for "the valley"... anyone who "lives" there deserves it... too frigging many people.

...Jim Thompson