Is electronic engineering dead?

I don't agree - while you see  a great many published designs executed in this way this is because their is little protectable (or saleable) IP embedded in them.

Real products made commercially still (often) use original hardware design to gain cost and performance advantage. But having invested a lot of money in a design it's not usually a good plan to publish it.

The picture shows an example (a current project):

There are two processors and an FPGA but there are filters ADCs, bipolar DACs, switching supplies, differential amplifiers etc. None of these are copies of reference designs (although obviously based on the huge body of analogue design knowledge available in books, on the internet etc.).

And while I can happily show you a picture of the pcb I'm not ready to share the detailed design.

There will be work for good hardware designers for a long time - and I suggest that such work is much more resistant to attack by AI  than any kind of software work.

MK

I don't agree - while you see  a great many published designs executed in this way this is because their is little protectable (or saleable) IP embedded in them.

Real products made commercially still (often) use original hardware design to gain cost and performance advantage. But having invested a lot of money in a design it's not usually a good plan to publish it.

The picture shows an example (a current project):

There are two processors and an FPGA but there are filters ADCs, bipolar DACs, switching supplies, differential amplifiers etc. None of these are copies of reference designs (although obviously based on the huge body of analogue design knowledge available in books, on the internet etc.).

And while I can happily show you a picture of the pcb I'm not ready to share the detailed design.

There will be work for good hardware designers for a long time - and I suggest that such work is much more resistant to attack by AI  than any kind of software work.

MK

Oh my goodness I hope not! I worked in medical research as an electronics development engineer and nearly every project required novel solutions.
I relished the challenge of working on demanding problems that often required unconventional approaches to achieve the final goal.
Developing equipment for use in a medical environment of course has the added constraints of stricter safety standards, and whilst (as our questioner points out) standard circuit elements are employed, but it's the design engineers job to use her/his skill pull these together to provide a satisfactory outcome.
That's what I got out of bed every morning for, the excitement of pitting my skills against what seemed an insurmountable electronic puzzle.
It's the creativity that made me love the job.

Daveb

Out of curiosity, how do you chose the ICs, chips... to use in such medical devices?
In most datasheet that I have read over the years, there is always a text saying "this device is not for human health, we won't be liable for any injuries/death..." even for the littlest component.

Well you have to use the chips that are available of course but its the careful development of the rest of the design that mitigates these safety issues. That's the challenge and skill of being a development engineer I guess. Yes I would lie in bed pondering these issues but still couldn't wait to get to work the next day. Pretty sad aren't I?

Daveb

Hello, My job for the last 38years is repairing electronics. when dealing with domestic equipment which can sometimes get circuits. But when i am doing industrial modules i have never got circuits. so using data sheets and my own test points works. one thing that gets me is when designed they put small caps running 24/7 in a standby power supply that should be away from heat. Where i have found bigger caps of the same value last longer.

then a motor driver ic with multiple legs would die because it has high voltage then dust then moisture.

When you plan power supply or Driver circuits do you think of the life of caps in the machine  

running 24/7 there is 8760 hours in the first year of warrantee 

how many caps used are better than 2,000 hours.

acdc90 said:

When you plan power supply or Driver circuits do you think of the life of caps in the machine  

Actually, yes.

The board in the picture will have an easy life - it's planned existence is that, after development and testing, it's powered up life time will be << 1000 hours.

But that is unusual.

You can make your caps last longer by running them cool and de-rating by large factors for voltage and ripple. 1000 hour rated cheap caps often last a very long time in low stress applications.

The realities of power design often result in large stress on capacitors but when the cost of a big 10000 hour 450V cap can exceed £100 you can see why they don't end up in washing machines.

Another thing to watch out for is that the capacitance of ceramic capacitors (especially Y5U and X5R types) can be as little as 20% of nominal value at the full rated voltage and less than 50% of nominal at 50% of rated voltage.

Good point about the dust !

Just been working on a design with 80V supply and lots of tracks -  0.15mm track and gap rules definitely not suitable !

MK