Grundig Satellit 300 whinge

Think so? Looks like hand-taped artwork on mylar to me. Even in the '60s, people were using pre-cut pads and tape. Remember Bishop Graphics?

I still draw schematics on vellum, but I use a pencil.

John

The worlds we perceive are clearly different - I suspect that imperfections in perception have more to do with the differences we report than any difference in objective reality (whatever that may be).

=A0 =A0 =A0 =A0 =A0 =A0 ... Presumably

Design reviews don't go on for ever - though sometimes it feels that way - and some details are inevitably glossed over, particularly the ones that might upset the boss. And John seems to be easily upset - he certainly complains about being insulted surprisingly often.

-- Bill Sloman, Nijmegen

n t

one.

es

I'm 69. I can remember back to when we did it that way. Computer-aided design offers the potential to eliminate a lot of human error - not all of it by any means, but at least the printed circuit layout is more likely to be consistent with the circuit diagram, and a TIP29A is less likely to metamorphose in to a TIP29 on it's way from the circuit diagram to the parts list (which didn't happen to me, but to an even better engineer - I was just the guy who found out that the drawing office had screwed up his design).

-- Bill Sloman, Nijmegen

he

til I

ut it

e

or the

(not

hing

re

oken

g had

er,

ne.

s

But you are the boss - as you persistently remind us - so you haven't got time to do it all.

When I was a senior engineer, acting as the technical lead on a project, I didn't get time to check a lot of the detail work, let alone do it - though I did manage to snag some of the really tricky bits.

You may be in a position to goof off from your system responisibilities to do some detailed design, but you can't be doing all of it.

I didn't get to do that when I was being paid a competitive salary - after the move to the Netherlands I found myself in jobs where I was used less intensively, and could get to lay out entire boards.

-- Bill Sloman, Nijmegen

I

it

the

had

We use the Optimum Procrastination Management methodology.

John

I'd thought that's what I was describing. The big technological advance was coloured tape, allowing two layers on a single transparency, and eliminating one source of registration error. Quadrupled material costs, but what the hell.

Always thought that one point in favor of vellum was that it allowed ink erasure...

I use(d) pencil first, then finalize in ink. You would need a lot of confidence (justified or otherwise) to render in ink first. There were even photochemically treated versions that were compatible with blue-print duplication. You could modify the duplicate, through erasure and direct ink application, to create quick revisions and modifications of a basic standard drawing.

Nowadays, you go through any amount of difficulty to avoid picking up a pen or pencil, or to present something that doesn't look automated. It gets kind of ridiculous - easily read and understood CAD drawings being impossible to link to automated processes, and integrated artwork being impossible to read or use. I doubt anyone ever consulted the draughting department, when the software was being developed. Too late now.

Of course, in the end, pencil drawings have the longest shelf life, so those with pencil under the ink may actually end up with only the pencil showing, anyways. Ink-pen, laser and inkjet will have their own unique issues. Aren't too many instances where that kind of shelf life will actually ever be useful, anyways.

RL

Is that similar to "the earliest date you can't prove it won't be finished by" estimation method?

Sylvia

until I

it

the

(not

thing

broken

had

"No work is done before its time."

Never do today what you can put off until tomorrow, because in the meantime it may become unnecessary.

Sylvia.

I have a blueline machine, and it makes very nice copies of pencil on vellum drawings.

I do a lot of whiteboard sketches and photograph them. I recently rigged up a dedicated light to illuminate my whiteboard for photography. Most of my circuit notes and test results get archived as whiteboard photos lately.

I put the whiteboard photos into emails, proposals, and even preliminary manuals. Most of my customers really like it. I think they are tired of spreadsheets and PowerPoint cartoons.

John

Not my experience. Computer-generated artwork includes a lot of information over and above the basic net-list and component data (most of which doesn't get displayed when you don't need to read it) which ought to being being devoted to making the art-work reader-friendly. The gEDA mailing list and tutorials talk about "light" and "heavy" symbols - a NOR-gate would be a light symbol, while an SN7402P would be heavy.

Depends on the software. gEDA was largely developed by the people that used it.

In other words you haven't yet learned how to make humanly readable schematics with a computer-aided design program.

I'm a bit surprised that you persist with white-boards and dye-line machines - something incised into wet clay and subsequently baked would have the advantage of being archivable - Babylonian documents in this format are still legible - and would be accessible to those amongst your customers who are blind as well as dumb.

-- Bill Sloman, Nijmegen

ftp://jjlarkin.lmi.net/

Idiot.

I'm not a bit surprised that you pontificate on everything and do nothing.

How's that oscillator coming along?

John

If it's worth doing, it's worth doing well.

The converse is also true.

--

John Devereux

the

Of course not. Other people do PCB layouts, FPGA design, embedded code, mechanical design, BOMs, and entire product designs. But that doesn't stop me from designing products too. I am pretty much the architecture designer for everything (with a lot of debate) and I write most of the manuals. I think I personally designed about 11 boards this year, so far. One more to go, likely, a VME LVDT/synchro simulator.

I don't run the business; I delegate that.

My company exists to allow me to design electronics.

John

m the

for

until I

, but it

the

n for the

na (not

e thing

wire

broken

ring had

s?

ether,

s?

han

eat

done.

ices

Writing manuals is a specialised skill. One of my better moves in 1992 was to discourage my boss from getting me to write a manual, and putting him onto an acquaintance of mine - David Johnson-Davies

who'd got his Ph.D. in psychology while writing software for Acorn Computers - he was one of the three founders, but got out before they went bust - and realised that he knew how to write good manuals.

My boss came back from talking to him in a particularly good mood - the example manual he had been shown was for an instrument whose commercialisation he'd declined to fund in large part because the manual was rubbish. David Johnson-Davies' rewrite had been a vast improvement, and he did a good job on our project too.

Designers don't tend to be good manual writers - they know too much about how the equipment works, and tend to fail to explain points that are obvious to them. John Morton, when he was at the Applied Psychology Unit in Cambridge, wrote a rather entertaining paper on this and related topics as part of his work on human-computer interactons. He and I have never been able to work out which one it was ...

And most of the design work. Eleven boards in a year is a lot.

Your customers won't see it quite that way.

-- Bill Sloman, Nijmegen

e:

cy.

er

r

Everything? I do have a particular interest in computer-based descriptions of electronic circuits that comes, in part, from being stuck with working with a early example of the breed in the late

1980's. I wrote a long memo back then on the information that such systems needed to capture, and the different ways the various parts of Cambridge Instruments needed to access that information.

Not nothing, but less than I'd like to.

Slowly. Much too slowly.

-- Bill Sloman, Nijmegen Bill Sloman, Nijmegen

until I

but it

for the

(not

thing

broken

had

together,

done.

I write the manual before we design the electronics. It becomes the requirements document; every product needs a manual, so why not do it first? It helps enormously to ensure that the product is usable, and explainable, before it's designed. I also have my prime customers review the manual before we design.

I wrote the manuals, drew the schematics, and participated in the layout of all 11. Other guys did a few more. I did about 200 board designs one year when I was younger, but they were smaller, less complex gadgets than the stuff we do now.

Their function is to finance my design activity and to present me with interesting problems. Lately, they are doing pretty well at it.

John

from the

on for

ked until I

ked, but it

out the

tion for the

tenna (not

the thing

o a wire

s a broken

wiring had

ders?

together,

bles?

s than

great

10% done.

devices

his

ir

and

ngs

l

ups

Then you aren't pushing the technology very hard. If you know exactly what the instrument can do before you design it - to the point of writing the manual - it can't be all that difficult or interesting.

The way the instrument goes together - and how it does what's required

- should be more interesting problems. I was recently re-reading some of my old weekly reports from the late 1980's (which - strictly speaking - I should never have taken away with me) and I was surprised by the way my understanding of the equipment I was specifying changed (and improved) as I was writing the specification. I'd started out with a pretty clear (and essentially correct)idea of what the equipment had to do, and how it could do it, but there was a lot of devil in the details, and there were some fairly dramatic changes as - for instance - it became clear that I could reliably get my hands on faster, bigger ECL static RAM than I'd known about when I made the original proposal.

My memories of that period feel pretty detailed, but it was a complicated machine and the story that I recalled was clearly less complicated than the process I'd been sketching out for my bosses from week to week.

Those weekly reports got circulated to the people I was supervising - as they came on board - as well as the bosses. I circulated it originally to keep myself honest about what I was writing about them, but it turned into a useful team-building mechanism; in the end I had to put a head-line section at the start that was short enough for everybody to read before they went on to check what I'd written about them.

At the end it was taking half a day every week to write the damned report, and I was covering software and mechanical hardware (where I didn't have any direct responsiblity, though I had to know what was going on) as well as the electronics.

I'm sure it does, but if the designs were pushing the envelope half as hard as you claim it wouldn't be all that practical.

The problems can't be all that interesting if you can chew through eleven of them in one year.

-- Bill Sloman, Nijmegen

the

for

until I

but it

the

for the

antenna (not

thing

wire

broken

wiring had

together,

than

great

done.

devices

Look at my products. Do many of them look trivial to you?

When would you figure out what you're trying to do?

My manuals include specifications.

Why not? If you don't know how to take calculated risks, you shouldn't be doing it.

Why would anyone spend months designing one PC board, even if it's interesting?

ftp://jjlarkin.lmi.net/T940_first_board.jpg

ftp://jjlarkin.lmi.net/First_CO2_SPM.JPG

ftp://jjlarkin.lmi.net/TEM2_CTRL.jpg

ftp://jjlarkin.lmi.net/TEM2_FPGA.jpg

ftp://jjlarkin.lmi.net/TEM2_Power.JPG

John