The most important changes in electronic design over the past 25 years?

I've always chortled, "I don't care what a structure is, or how it's made... give me a Spice model and I can design around it" >:-} ...Jim Thompson

le as

.highlandtechnology.com  jlarkin at highlandtechnology dot com

Excellent,

I want to add the great analog parts now availble for cheap low noise opamps

George H.

re

ttdesign.com

Yeah, digital continues to swallow more and more of the world.

George H.

Embrace, George. Embrace.

Of course FPGAs grew out of PLDs and FPLAs, which had been around a decade by then. I used a register with an FPLA in the feedback to build my own controller in the mid-'70s (it was the "brains" for what was essentially a DVM).

Let me know when they replace the repair bench. :)

Using AoE2 as an example, the only glaring omissions I recall are:

1. Besides Larkin, who actually uses 68k anymore? Yeah, a few, but PIC/AVR and ARM pretty much dominate today. (Indeed, the 68k dominated in its time, but that time has passed. That said, perhaps the PIC/AVR/ARM market is mature and, by the time most people read the next editions of these textbooks, will be obsolete!) Few use MSI anymore; CPLDs and FPGAs are so much easier.
2. High performance parts. GaAs, GaN, PHEMTs, and for power, SiC too. Even superjunction devices didn't exist at the time.
3. High bandwidth signaling standards; signal integrity. Highly important today (you can't get too much out of the FPGA from #1 without using up all your pins, unless you use LVDS or PECL or..), and forms an intersection between analog and digital which one cannot ignore.

Anecdote regarding #3: at school, I once attended a recruitment presentation from Plexus. The speaker asked us, "Does anyone know about signal integrity? If so raise your hand." Out of the about 100 EE students in the room, I was the only one. Needless to say, signal integrity is not covered in any undergrad courses.

Tim

Those old Intel ceramic chips make excellent x-acto knife sharpeners.

Did anybody else run ECA? That was a non-Spice text-netlist simulator. Ran fast, always converged, let you parameterize anything (capacitance, resistance, anything) versus anything else.

I hope someone does. I have about 8,000 on hand. A reel of 5000 of RM73B2AT683J, and around 3000 68K .25W 5% Carbon Film

We're using ARMs in new designs, but I have a lot of 68332-based products that still sell well. Once in a while I have to make a code change, and it's easy. All the build tools, even for 15-year-old designs, just work.

I wonder how many people will be able to easily change their code and rebuild 12 years from now.

Yeah, a few, but

I can't imagine any architecture replacing ARM for a long time. The reasonable future is multicore, *lots* of cores, in x86 and ARM.

I think multicore will change things.

The real, earth-shaking tech change coming up is a whole new way to write software. Software is the great misery and hazard of today's electronics. I'm thinking something non-procedural, state-machine based, vaguely like LabView. The other revolution that we need is computer-related education, teaching kids how to write reliable, readable, maintainable code.

Printed circuit board fabrication techniques and the software to design printed circuit boards.

Assimilate. You will be...

Liquid Immersion Lithography.

The move to HDTV and digital broadcasting.

High bandwidth wired connectivity for the masses.

Large form factor fast refresh LCD displays with LED backlighting at a reasonable price.

And finally... all the Dick Tracy Wrist Radios folks all have now. just a bit too big for the wrist. And we still call the "phones".

We had Racal-Redac software for digitally generating printed circuit artwork at EMI Central Research back on 1978, which is almost 35 years ago. It wasn't user friendly, and the draftsmen that drove it needed a refresher course if they'd been doing something else for more than a week, but it was a relatively quick way of generating artwork for digital circuits, if too clunky for analog layouts.

It was probably the Racal-Redac MAXI PCB package which was introduced around then, but I never actually saw it in operation though - IIRR - we did check out one of the layouts it produced for us before sending it out to get turned into a board.

The biggest change is not needing a huge library with datasheet books due to internet. Ohm's law won't change for at least another 50 years :-)

Grin... do I have any choice?

I went to a local maker-place for electronics night. I brought my Rigol scope and DMM. Everyone else there took out their lap tops, and started tapping keys.

I'm thinking I should bring my soldering iron next time, I've got a few 'kit' projects I can work on. :^)

George H.

ttdesign.com- Hide quoted text -

I thought the repair bench HAD been replaced - no exchangeable parts inside. Throwaway menatality.

Have you see the minimum replaceable size of an automobile system block? We're talking scratch a bumper and replace the front end.

As an undergrad, about 1980, I worked with a millimetre-wave radio astronomer (Prof. Bill Shuter, may his tribe increase). His 110/115 GHz radio telescope used a GaAs FET as the first IF, and it wasn't a particularly shiny new rig IIRC. So they've been in some sort of use since the 1970s.

Cheers

Phil Hobbs

It's a ghost!

Cheers

Phil Hobbs

"Let's get out of here, Scoob!"

-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics

hobbs at electrooptical dot net