Power Supply Engineer - National Instruments - Austin, Texas

rote in

But I eventually managed it. It turned out to be impossible - in the Netherlands - after I'd turned 62. I think I came close once or twice, but never got an offer. That doesn't make krw any less of a nitwit, of course.

have.

So why did you indicate in your writing that he was?

in

It is _not_ impossible. But one has to be pro-active about it and broaden the horizon to other countries and often also other continents.

I personally know people who consult or contract on hardware design and they are older than you. In fact I enjoy working with them because hey have such a wealth of experience. I do not have to explain EMC risks to thme, they know.

in

I could get all the interesting work that I wanted, if I didn't already have a job. I've even told Bill how to do it. He doesn't seem to be interested enough to try.

pay,

have.

We occasionally do a review of a product's history, units sold over time and repair rates, to estimate MTBF and to look for any specific failure hot spots. A lot of our products have million-hour MTBFs, often a factor of 10 or more better than we compute from MIL-HBK or Bellcore methodology. Parts are really good these days.

e in

ers, and

rt of

rback

pay,

e.

ad

dge

I'd like to work for ASML too. You are a sub-contractor to a serious firm - the people who commissioned you are working in a serious environment, and will have to try and persuade you to perform at the appropriate level. I don't envy them. That aspect of my job at Cambridge Instruments could get a bit frustrating, but they have got the advantage that they have got you working in an area where you have an established expertise..

wrote in

Which is what I've just done - at last - by virtue of my wife's relocation

My last job in England involved getting the Affinity Sensors machine through the emission tests certification - it failed, but we knew exactly why, and fixed the problems and passed when they re-submitted a few months after I'd left.

Quite a lot of the projects I was involved in at Cambridge Instruments involved minimising emissions and coping with EMC problems. Lots of differential signalling and receivers, and the occasional transmission line transformer aka balun.

wrote in

er

It's not lack of interest, it's lack of the kind of unrealistic self- belief that you need to be able to behave that way.

ote in

in

ofit

has

of

clue

I have.

I didn't. He is irrationally optimistic about the quality of the stuff he sells, but clearly he does listen to reason when it comes working out when the product is good enough to ship. In fact his claimed development schedule of one new product a fortnight is more in line with the Japanese policy of continuous incremental improvement, where you wait to see how the customers react to each new improvement before getting on with the next one - than any kind of exuberant build- something-great-from-scratch approach, which I've been caught up in from time to time.

and

of

Powerpoint-paperback

pay,

I've designed 6 different boxes for this project, about 15 different PC boards, and every one worked first etch. All you need to do is be careful.

This is one of the boards.

The 12-bit ADC above the FPGA is clocked at 250 MHz. It's sort of shocking to think that the data is only valid for a nanosecond or so around each clock.

wrote in

All you need to believe is that you can be friendly, helpful, and reasonably competant. Is that beyond what you can imagine?

Most people like to talk about what they do, if the audience seems interested and intelligent. I get a lot of business by asking non-electronic people about their physics or optics or whatever, then suggesting how good electronics can help them. The concept here is pretty simple.

in

have.

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

You claim you didn't, and then in the same line you do it again. That takes the cake.

Do you honestly believe you can judge the quality of someone's products telepathically from 6000 miles away without ever seeing them live?

At dime a thousand available volumes it is going to take some time, plus their replication rate is rather high.

?-)

..

hat

ever

s.

nd

to

That doesn't stretch my self-belief at all. I've been friendly, helpful and rather more than reasonably competent all my life, though I do get testy when people claim that I can't get work because I haven't been trying hard enough.

You missed the bit about getting through the door to talk to them in the first place. It's relatively easy when they know they've got a problem, and they've got some kind of idea that you might be able to help them solve it.

Telling them that they've got problems that you can solve is a rather different interaction, but that's basically what cold-calling is all about.

I've got to spend an hour sometime next week getting out to an employment agency in Pennant Hills - a remote suburb of Sydney - and another hour getting back after I've talked to them. I suppose it's time well-spent - there are other agencies, but this one seems a bit more inclined to advertise high-tech jobs, and claims to know the local semiconductor firms, such as they are.

Back when I was involved, they called the mahines wafer steppers, because the mask was stepped across the face of the resist-coated silicon wafer to put down repeated copies of the same image to create one of many successive layers of detail in the integrated circuits being manufactured.

Feature size is now down to 28nm, on a device that may be a cm across. If the successive images are to register to this accuracy from one exposure to the next, you require rather better temperature control of the silicon and the mask than could be offered by a champagne cooler.

rote in

gsters, and

s of

sort of

read

aperback

ried

the pay,

he

used

use.

s to

head

judge

in

ofit

you

zed

ge

n.

And not too ambitious.

Why? At 250MHz everything repeats every four nanoseconds. That's about

The most evil design I ever ran into was running TTL latches at

I got to improve its sub-nanosecond output jitter by resynchronising the outputs to a 200MHz clock in ECLinPS and shifting the resynchronised outputs back to TTL in an ECL-to-TTL converter. Fun project. My version of the board had its own hand-tuning built in - you could pick which edge of the 200MHz clock you used to re- synchronised the output.

Still a bit evil, but less so. And it did reduce the jitter.

wrote in

e in

profit

old

e has

y of

t clue

f? I have.

^^^^^^^^^

Look at the words. He's claimed that his products are "insanely good". If that doesn't translate exactly into "irrational = insane optimism" then I'm not a native speaker of English. He's condemned out of his own mouth - or more precisely - out of an old post of his here, which you could find without too much effort.

And do keep in mind that I do have a sense of humour, even if John regularly fails to detect the joke.

Don't remind me! I'm surrounded by them, here in the Deep South. :)

Hint: The champagne cooler is not the part I designed. It comes as an "accessory" in that league.