Power Supply Engineer - National Instruments - Austin, Texas

--
Oh, silly me, I thought that what electronic design engineers did was 
to design electronics, with the deliverables being primarily the 
product and secondarily the documentation. 

Unless, of course, the product _was_ the documentation, like what Jim 
does.

Silly, and totally wrong. What we do is produce documentation so that manufacturing (in-house or far away) can buy the right parts, assemble them, test the result, and ship to customers. We also produce the manuals, app notes, firmware, FPGA configurations, BOMs, ECOs, test procedures and (sort of recursively) all the documentation and software for any test sets, fixtures, cables, anything else related.

Our documentation also has to be structured and commented and archived so that we, or someone else, can come in a year or a decade later and understand, modify, or upgrade a design, namely revise the documentation.

Engineers don't build or deliver products. Technicians and amateurs sometimes do.

There is a body of practice about how all this documentation is structured. It varies across companies and industries, and is generally not itself fully documented... it's engineering practice and conventions and passed-on lore. Most companies dont't do it very well. I'm always hearing tales of lost documents, files that can't be opened, schematics that don't match layouts, source code that can't be found or recompiled. I've sold about 4000 temperature controllers because a British company lost the Z80 source code for theirs, and couldn't make some simple changes. I changed a few lines of my

12-year-old code recently, recompiled, rebuild rom images... it was easy.

I do the same: generate documents that other people use to build things. So do civil engineers, architects, clothing designers, script writers, authors of romance novels. We don't make the end items ourselves.

I don't make cookies. Manufacturing does that. Some of my customers get the drawing packages and have the option to make and test the products themselves, for whatever reason, and a few have. The documents are exactly the same.

The big difference is that you sell design services and I never do. You sell your IP once, and I sell my IP many times, in boxes that are manufactured for years.

Look at my web page. Show me something that's paint-by-numbers. Pick a product and tell us how easy it is to do... tell us how you would do it. I don't think you can. Begin weaseling now.

As far as "business demands", it's fun to sell stuff. Customers validate the worth of designs. Revenue provides the people and the tools to develop even better designs.

How's business? Designing anything interesting?

--

John Larkin         Highland Technology, Inc 

jlarkin at highlandtechnology dot com 
http://www.highlandtechnology.com 

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom laser drivers and controllers 
Photonics and fiberoptic TTL data links 
VME thermocouple, LVDT, synchro   acquisition and simulation

read

than

do

aren't.

And it's unfortunately best appreciated after some time has passed. I continually annoy myself by reading my old calculation files and wondering "what the hell was he thinking?"

--
That's not really a valid simile, since Jim hasn't a foundry and 
doesn't churn out IC's for sale by his firm, while you have a 
production facility which churns out product. 

Moreover, Jim's product is _totally_ documentation, while yours (your 
firm's) is finished goods.

An interesting division of labour. I think you'll find that things are a lot more blurred. Even ISO doesn't stipulate who actually does what

- though diligence would suggest that various bodies and their signatures are present where required (considerations of liability).

The days of the secretary or personal assistant are probably gone in this field. It's probably best to be ready to chip in, as required, to get the job done.

In the past, roles overlapped due to experience. Currently it's due to 'under-employment'.

RL

Our engineers don't manufacture stuff. They do work with manufacturing to make sure the stuff is easy to build and inspect and test. But our reaction to such reviews is to change the documentation.

In our shop, engineering documents specify the end result, but seldom say anything about the process; that's up to manufacturing. They have their own process documentation, part of their institutional memory, but that doesn't directly affect us. We only officially care that the shipped products conform to our released documents.

Even ISO doesn't stipulate who actually does what

We're not ISO9000 and don't plan to be. It's been shown to have no effect on product quality and would certainly reduce our productivity and profitability. It's sort of annoying to think that some consultants would come in and claim to know more about how to run our business than we do. Lots of people seem to claim that lately.

Agreed, we do almost all of our typing and breadboarding and cleaning-up-the-lab ourselves. It's hard to keep engineering technicians busy, so we don't have any. I would like a combo tech/parts researcher/Visio drafter/manual editor/pizza-and-coffee fetcher/prototype tester/web crawler/hole driller person, if such a person were possible.

(I used to write manuals in longhand, and have them typed by secretaries. What an inefficient process! And I used to program in longhand, and have people punch cards with source code. Ditto.)

Think so? Things are so complex nowadays, only really good, really smart people can handle high-end designs. Those people are typically over-worked. It makes sense to get them all the help we can, so that they can be more efficient.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators

Our designs are fully documented. Purchasing can send drawings out to contract manufacturers to build; they do, and we usually don't even know about it. A couple of our customers have elected to manufacture our designs themselves, under license, and they do, from our released drawings in the library, same drawings our manufacturing uses.

I doubt that you furnish a full documentation package: schematics, PCB layouts, BOM, assembly drawings, test procedures, manuals, ECOs, enough to build a product. Do you?

Exactly. Our IP is never for sale.

The great thing about doing full documentation is that we - engineering - can move on to something new. Production takes over.

Excellent weaseling!

--

John Larkin         Highland Technology, Inc 

jlarkin at highlandtechnology dot com 
http://www.highlandtechnology.com 

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom laser drivers and controllers 
Photonics and fiberoptic TTL data links 
VME thermocouple, LVDT, synchro   acquisition and simulation

--
You try to muddy the water, as usual. 

Jim supplies documentation based on simulations using mathematical 
models, while the documentation you provide is based upon having built 
real hardware of some kind, having tested its functionality in the 
real world, and then putting words and pictures together so that 
someone else can build the same thing over and over again without 
having to go through the design stage.

I don't think you are doing any serious electronic design at all.

--

John Larkin         Highland Technology, Inc 

jlarkin at highlandtechnology dot com 
http://www.highlandtechnology.com 

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom laser drivers and controllers 
Photonics and fiberoptic TTL data links 
VME thermocouple, LVDT, synchro   acquisition and simulation

Granting your capacity to have silly ideas about a lot of subjects, they may even be right. Aging and eccentric owner-managers are prone to retain antiquated misconceptions, which less powerful figures wouldn't be in a position to enforce.

But you aren't fond of simulating your circuits before you build them ...

--
Bill Sloman, Sydney

t s e d d . e m

uld do

A standard John Larkin put-down. In fact he hasn't got a clue what John Fields does - what John Fields posts here is not all that impressive, but he isn't getting paid for that - and is relying on his fertile and uninhibited imagination to come up with an offensive libel.

--
Bill Sloman, Sydney

Amazingly, some companies did not really know any of the modern ferrites. As if the clock had stopped 30 years ago. They probably still play "Mr.Bojangles" in production :-)

I remember back in the 90's when I painstakingly wrote detailed instructions for a client to make three transformers on their own. Because we could not find anyone who wanted to do it. I could not believe that companies would just leave all that money on the table. Later I told the production manager "Hey, you guys could probably make more profit building these ferrite products than building your usual stuff".

Later I had to almost twist the arm of an engineer at a ferrite place, telling him "Look, this is the ferrite you'll use, these are the wires, this kind of Kapton tape ..." ... "But our calculations" ... "Forget those calculations".

--
Regards, Joerg 

http://www.analogconsultants.com/

do

I don't think you are doing any serious electronic design at all.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators

Simple stuff doesn't need simulation. The really high-performance stuff can't be simulated. Most of our circuits are *designed*, and don't need simulation.

We never simulate more than a small subcircuit of a product, never the entire product. Most of our designs are done with no simulation or breadboarding at all. Over 95% work, and can be sold, first etch.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators

My clock rates haven't really been pushing ferrites that hard, I'm only at the point where they start warming up. But I'm always looking for that edge. Of late, I seem to have come to the realization that there isn't anything that'll go anywhere near saturation past a meg or so without melting -- and by the time you get up towards 10 megs, anything with low losses has really low permeability (Fair-Rite #61 and such) or is lossy by nature anyway (powdered irons). Either way, you aren't going much above, say, 50mT, which makes for an awfully bulky transformer.

Though at least the RF irons are better than the "high-mu" ones like #26 and #52 -- to grenade that #2 core on your antenna way back when, you must've had some nasty SWR... not to mention the kilowatt plus, or whatever it was behind it. :)

Like all Good Engineers, I wear a full beard (ok, maybe just out of immitation, and also laziness, but one can hope right?). I was chatting with one of our sales guys who had just visited an old company. I forget what they made, something mechanical. He said I wouldn't do very well there -- they have strict hygiene rules, shaved face, neat hair, shop suit, I forget if they wore bowties too.

I know that sort of thing was more common back in the day, but most companies these days have evolved past that, even in manufacturing, plus flex hours, more vacation, etc.

I just said, good thing I'm in engineering then. :)

Tim

--
Deep Friar: a very philosophical monk. 
Website: http://webpages.charter.net/dawill/tmoranwms

es.

pect

r

was

hat

mble

t

d.

ived

and

s

and

ell.

t be

s

Jim

ipt

ur

o d t

kie

l.

ent

CB

.

are

our

y

of

e's

would do

g

Certainly not at the moment. We've just moved countries, and this afternoon's job is to fix a few more bookcases to the wall so we can unpack some of the 44 mover's boxes now littering the second bedroom behind me, which have been sitting there for more than a year now - moslty while we were back in Nijmegen.

My new printer/scanner is supposed to arrive tomorrow, but there's a boatload of stuff coming from Nijmegen that won't get here until the beginning of December.

And why would you care? I certainly wouldn't give anything serious to you for nothing.

--
Bill Sloman, Sydney

.

ct

as

t

le

ed

d e o

Or you can't be bothered to get a grip on the stray capacitances and inductances that might let you simulate it.

And you are pushing the state of the art really, really hard, but in tiny increments.

--
Bill Sloman, Sydney

One challenge is during voltage surges. Then you can either cut out electronically which is usually not so desirable, or the FETs become "exotherm" which really isn't desired.

The SWR wasn't that bad. Until after the kaboom, that is. Pushing 1.2kW over a two-incher was a bit much. I should have stacked two but the other 20 Deutschmarks had already been invested in a crate of Pilsener.

I always had a full beard but that wasn't a problem in clean rooms. A mask has to go over it but one needs that anyhow. Ok, in the army I did not have a beard, not allowed there for obvious reasons.

--
Regards, Joerg 

http://www.analogconsultants.com/

would do

If you ever hope to do anything real again, your best strategy would be to find a small company that could use your electronic and physics and chemistry skills, as a consultant, and indeed offer to do something for them for free, to see how the relationship might work. They might take you up on the offer, and would typically counter with an offer of some payment.

There's an old saying from HP: "Pay lags performance." A whole lot of people don't get that concept.

--

John Larkin         Highland Technology, Inc 

jlarkin at highlandtechnology dot com 
http://www.highlandtechnology.com 

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom laser drivers and controllers 
Photonics and fiberoptic TTL data links 
VME thermocouple, LVDT, synchro   acquisition and simulation