What's the hi-side current sense chip du jour?

Old American S&M wisdom: If you don't take care of your customers, someone else will.

any chip

ould be

ed

at the

ike it

parts

rdly

=A0A

at

ly

=A0 =A0 =A0 =A0 =A0... I haven't been

It may be a matter of logistics, but if you claim a part doesn't work, the company wants to hear about it. It might be a way to get a free eval board (hint hint).

A lot of time the customer screws up and throws in the towel too quickly. I had a case where a battery charger didn't work. The board had made the rounds through apps and eventually it became my problem. I looked at the batteries and my first though was they were not Sanyo. A bit of bench testing showed these batteries could not be quick charged, even though the datasheet indicated they could. Sent the board back to the customer with Sanyo NiMH, and sent the dud batteries back to the manufacturer. Customer happy, Sanyo happy, the other guy, kind of pissed.

Incidentally, my one and only recall was due to a customer gating the part with a P-fet rather than using the shutdown pin. I assumed since a shutdown pin was provided, it would be used. The customers design philosophy is everything gets shut down with a Pfet pass device, period end. There was a strange latch-up in the part due to being started up with the pass device and voltage present on certain pins Latch-up QA tests are basically static, or perhaps done with a Tek curver tracer, which has a smooth sweep. Everything post that fiasco got the P-fet pass test. [Slow start up, essentially hand cranking the power supply knob, was always part of the bench test.]

Regarding Maxim management, don't hold your breath. The board is useless. Certain people will have to retire. I can name them and shame them, but it's not my job and I don't need the lawsuit. A corporate raider could hire a few ex-employees and get the low down on who to get rid of, then make a bundle.

chip

be

it

When the MAX9690s started to fail, Maxim sampled me 3500 pieces of the

ftp://jjlarkin.lmi.net/Break2.jpg

ftp://jjlarkin.lmi.net/OnBoard.jpg

John

ly any chip

should be

lowed

ed at the

t like it

eir parts

xim

hardly

blem

nd. =A0A

r a

ive

t or

sure

are

us)

ard

are

llow

or

tion

d in

he

that

ently

ke

=A0 =A0 =A0 =A0 =A0 =A0... I haven't been

he

he

ns

nd

at

What exactly failed on the part? Was it a packaging problem?

any chip

should be

the

like it

parts

hardly

Maxim subbed the fab to some shop in Minnesota that apparently had a dirty oxide problem or something. Over time, some carriers would diffuse somewhere and pinch off a resistor, at which time the comparator would get erratic. The parts failed at high temperature, and over time - like a year maybe - the "fail" temp would creep down towards room temp. The weird thing was that you could bake the parts and raise that "fail" temp back up for a while. All this was very confusing.

They discontinued the part without warning, in fact when shipments were long overdue. They then announced their "drop-in replacement"

- the MAX9691 - which is the only comparator I've ever heard of that has BACK-TO-BACK DIODES BETWEEN ITS INPUTS. Imagine what that did to a capacitor-linear-ramp/dac time delay trimmer circuit.

grrrrrr.

John

early any chip

ing should be

d slowed

tayed at the

on't like it

their parts

Maxim

or hardly

problem

round. =A0A

for a

drive

gous

ault or

is.

assure

as

rdware

KA

rious)

t hard

hey are

t allow

rator

tuation

used in

y the

as

If that

not

nt

equently

l.

s I

like

me.

=A0 =A0 =A0 =A0 =A0 =A0 =A0... I haven't been

he

t the

s the

tions

ts

gn

to

, and

ned

the

s,

he

that

that

ly

I can see this happening. My original comment was about parts being non-functional right out the chute.

But the 9690 was a bipolar part. Hard to believe oxide has that effect.

I've only used bipolar inputs in bicmos for ASICs. The clamping under those conditional is not that unusual, then again, I'm setting up the operating conditions. If you yank a differential pair far apart, two ugly things can happen. One could be a polarity reversal in the comparator, which depends on a lot of things. The other more sneaky problem is the offset of the differential pair shifts. The whole reason you went through the pain of using bipolar inputs (speaking from a BICMOS point of view) is the low untrimmed offset off the differential pair.

That was a Gain Technology design. Those guys spent way too much time in the Arizona sun and fried their brains. While you end users may think you are the judge of component quality, your best source of how well a part works is usually the test engineer that handles it. Trust me, the test engineers didn't like the Gain Tech designs. The term cowboy was used often.

nearly any chip

should be

slowed

at the

like it

parts

Maxim

hardly

problem

ground.  A

drive

or

assure

hardware

serious)

hard

are

allow

situation

in

that

Consequently

I don't know the physics. The behavior was as stated. Whatever went wrong, it could be annealed out.

Clamps are fine for opamps; not for comparators. Don't be shocked, but comparator + and - inputs are sometimes driven to *quite different voltages*! On purpose!

Yeah, what the hell do we know? When a comparator output is entirely wrong, I should ask a process engineer if that's OK.

John

Especially since they obviously wrote or said "drop-in replacement". The engineer who ECO'd that statement should have his head examined.

Old American business rule #2: A product is only as good as the customer says it is.

I thought rule #2 was, "When the boss is wrong, walk." ;-)

...Jim Thompson

If the end user doesn't see it, does it matter?

Why shouldn't it (be foreign)? ATE crap is the test engineer's job.

Insiders are *supposed* to make a bundle. That's why they took all the risk.

I've found a number of horrible bugs in chips, some of which the chip designers were unaware, some of which they knew about and did their best to keep a secret.

John

A couple I can specifically remember are:

(1) The LT1246 datasheet should by now be updated to state that the chip goes quite insane if you pull the Vc all the way to ground. One diode drop above ground seems to be enough to prevent this.

(2) Early LT1028 datasheets cut off the noise graph before the ugly rise to a peak that happens near 300KHz.

Other than simple op amps, Gain designs were terrible. This comes from some stupid management that thinks all engineers are the same. I assure you the stuff designed in the valley (Maxim HQ) was thoroughly flogged before shipping, not that the occasional bug doesn't get through. I don't consider anything the Arizona boys designed to be Maxim products, but unfortunately, the Maxim brand is on their designs.

I'm not sure if LTC still has that Singapore design center, but when it was opened up, they sent one of their employees to run it, which insures it was done to the valley's "home of the gurus" standards.

One thing management doesn't get is that a customer judges your company by the crappiest stuff you are willing to put your brand on. This should be obvious, but Dilbert is more than just a comic strip. The old HP really understood this, and nothing they sold was junk. [If a connector reached the outside world, it had to do 25kv ESD. If you couldn't deliver, they would add parts to make sure that happened. Obviously, some test instruments can't do that, but certainly their consumer gear was done to that standard.] Carly came along and Chinesed up the computer line with crap, and nearly sunk the company. The current HP stuff seems to be more under control, but how do you win back the customers that got burnt.

The test engineer sees the chips that are buggy. The typical test flow is something like this. The person running the tester loads the program, tests a golden unit, then tests the real thing. When the yield is low, testing is stopped (really just put aside) and the test engineer determines if it is an ATE problem or a problem with the device. [Remember, the wafers have had their test pattern flogged, so the wafers are known to be good prior to testing.] Once it is determined the product sucks, it goes back to product engineering at most companies, but often the design engineer at Maxim. [Maxim uses fewer product engineers than most companies.]

At this point, you are saying this is just "inside baseball" and who cares. Well, if the product is buggy, delivery slips. The designer, who has other work, eventually gets around to the problem, but what if the part was designed by the Arizona boys. Hey, they got paid for their (cough cough) skills, and won't spend time on improving their crappy engineering.

The bottom line" so called unobtainium Maxim parts.

If a Texas Instruments bought out Maxim, you would be amazed at how quickly the problems would go away. Cash out the useless management that just get in the way, and put in more competent and hopefully less managers. Many of the duds have left, but a few problem managers are still there.

That's downright dishonest.

One Hittite switch data sheet states boldly that it works from DC to 4 GHz. Turns out their in-house definition of DC is 100 MHz. [1]

THS3062 is great, as long as you don't ask it to amplify sine waves anywhere near a third of its swing or slew rate specs. If you do, it latches up and gets hot enough to put a hole in your finger. A little time spent phoning tech support (six weeks in our case) will verify that this is maybe true, in not so many words, and that it's really your fault.

LM35 is a very sensitive latching pulse detector, as well as a temperature sensor and general-purpose oscillator.

Mini-Circuits MMIC data sheets should state that "fab houses and gross performance are subject to unannounced, random, radical change" and "we don't actually know much about these parts." Optek lasers, ditto.

Don't even get me started on voltage regulator capacitive-load stability.

John

Yes, your point is? Yes, I've been on both ends of this.

What does the inside politics at Maxim have to do with anything? No one can actually get their parts anyway.

That's not Maxim's problem. Crap, they "sell" stuff they've never actually made. They are incompetent from the top to the bottom. Yes, that is managements fault. So be it. Just avoid Maxim at all costs.

You could have picked someone better than TI to use as your example. We're using TI parts that, according to the spec, don't work. No surprise that we've been having problems, huh.

That was my observation about Millaway, as well.

I just did a sweep of my PC, eliminating all the unnecessary start-up crap. MOST was hp garbage, not even necessary for my setup, like resident software to run a network printer that was attached to a specific PC... it's not, it's a stand-alone network printer :-(

...Jim Thompson

AD8212?

John

AD8211 has a better cm range.

John