Learn EE at university

[IntegratedValve]

It is not the target of the fundamental-lectures to give you the ability to build a complete Audio-Amplifier or something like that. You should learn the fundamentals and with them (and a book about Audio-Amplifiers) you can build one or you can use (almost) the same fundamentals to build a RF-Amplifier or some digital stuff.

I'm interested in digital stuff so I will specialize that direction. I think I'm not able to build an RF amp with or without any book and I think we will not learn anything about RF amps on mostly digital specialization/module. (the same thing with audio-amplifiers and a lots of other things)

Go for it. Digital is today's and future electronics. No one just out of school can build OpAmp's. It's something that require a huge $$$ only giant companies in the industry can afford.

[IanB]

Agreed. But how much does the analog circuitry represent of modern electronic systems? Mainly power supply, front ends and interfacing with sensors.
The rest is all digital.

Maybe. I'm not an EE, and I know analog has a particular meaning in circuit design, but on the other hand all circuits are analog, even digital ones...

[c4757p]

Yes, but half a graduating class of engineering students won't all find jobs mashing VHDL into their keyboards. Of course you don't build your own op amps, but you do use op amps.

Agreed. But how much does the analog circuitry represent of modern electronic systems? Mainly power supply, front ends and interfacing with sensors.
The rest is all digital.

Pop open a computer and look at the motherboard. There is more support circuitry than computer. And it ain't all digital.

Now, I understand I'm in the minority these days in preferring analog. Most people don't want to spend their lives designing support circuitry, and I don't have an issue with that. What I don't like is this "it's not the cutting edge, so we're going to pretend it's boring old academic bullshit nobody really needs" attitude that is evident in university curricula and students.

Engineering has many specializations, and in my opinion, educators should try to expose students to as many of them as possible so they can discover their interests. Yes, this includes forcing my grouchy ass to do digital stuff (hint: I actually enjoyed working with FPGAs).

[IntegratedValve]

Agreed. But how much does the analog circuitry represent of modern electronic systems? Mainly power supply, front ends and interfacing with sensors.
The rest is all digital.

Maybe. I'm not an EE, and I know analog has a particular meaning in circuit design, but on the other hand all circuits are analog, even digital ones...

True but all this is abstracted. I cannot specialize in machine language or assembly just because all higher level languages and abstract data structures, libraries, APIs, drivers and operating system are end compiled into machine code.

[IntegratedValve]

Yes, but half a graduating class of engineering students won't all find jobs mashing VHDL into their keyboards. Of course you don't build your own op amps, but you do use op amps.

Agreed. But how much does the analog circuitry represent of modern electronic systems? Mainly power supply, front ends and interfacing with sensors.
The rest is all digital.

Pop open a computer and look at the motherboard. There is more support circuitry than computer. And it ain't all digital.

Now, I understand I'm in the minority these days in preferring analog. Most people don't want to spend their lives designing support circuitry, and I don't have an issue with that. What I don't like is this "it's not the cutting edge, so we're going to pretend it's boring old academic bullshit nobody really needs" attitude that is evident in university curricula and students.

Engineering has many specializations, and in my opinion, educators should try to expose students to as many of them as possible so they can discover their interests. Yes, this includes forcing my grouchy ass to do digital stuff (hint: I actually enjoyed working with FPGAs).

And that's why I mentioned there's power supply, protection, and analog signal interfacing circuitry. Generally speaking those are now available as components you can plug in your circuits, and are designed by scientists in the field. The core functionality is mainly digital, and to be proficient in using FPGAs and uP using your own circuit (vs. devkits and Arduino ilks) is much more essential than spending your life in a limited application area. It's fine for the sake of learning but overdoing it is simply a waste of time and effort.

[free_electron]

But how much does the analog circuitry represent of modern electronic systems? Mainly power supply, front ends and interfacing with sensors.
The rest is all digital.


And that's why I mentioned there's power supply, protection, and analog signal interfacing circuitry. Generally speaking those are now available as components you can plug in your circuits, and are designed by scientists in the field. The core functionality is mainly digital, and to be proficient in using FPGAs and uP using your own circuit (vs. devkits and Arduino ilks) is much more essential than spending your life in a limited application area. It's fine for the sake of learning but overdoing it is simply a waste of time and effort.

You ain't gonna be much of a designer if you don't know how to get your signals in and out of that FPGA....
You can be a VHDL wizard but if you are tasked making a system , using DDR5 memory, a couple of PCI-X ports, and two SATa or USB3.0 ports and you have no clue how to do the layout , how to interconnect and terminate these signals properly, how to design the power supply , properly provide it decoupling and sequencing that thing is going to do nothing.

These days designers build SYSTEMS , not just 'i do code, you do analog , and we have a lab rat doing the polygon pushing to make the board'. That doesn't work anymore.

[coodoo]

You can be a VHDL wizard but if you are tasked making a system , using DDR5 memory, a couple of PCI-X ports, and two SATa or USB3.0 ports and you have no clue how to do the layout , how to interconnect and terminate these signals properly, how to design the power supply , properly provide it decoupling and sequencing that thing is going to do nothing.

These days designers build SYSTEMS , not just 'i do code, you do analog , and we have a lab rat doing the polygon pushing to make the board'. That doesn't work anymore.

That is very interesting. Where can I learn something like that? Books?
I don't think we will learn anything about that at university. :-(

[don.r]

You can be a VHDL wizard but if you are tasked making a system , using DDR5 memory, a couple of PCI-X ports, and two SATa or USB3.0 ports and you have no clue how to do the layout , how to interconnect and terminate these signals properly, how to design the power supply , properly provide it decoupling and sequencing that thing is going to do nothing.

These days designers build SYSTEMS , not just 'i do code, you do analog , and we have a lab rat doing the polygon pushing to make the board'. That doesn't work anymore.

That is very interesting. Where can I learn something like that? Books?
I don't think we will learn anything about that at university. :-(

Digital board layout often simply requires time and practice under the guidance of someone who knows what they are doing. When you have 128bit buses buzzing along at a GHz or more, layout is critical.

[free_electron]

That is very interesting. Where can I learn something like that? Books?
I don't think we will learn anything about that at university. :-(

not at university no...

books :

Right the first time by Ritchey
High speed digtal design by Johnson
Anything you can get your hands on from Montrose

[IntegratedValve]

But how much does the analog circuitry represent of modern electronic systems? Mainly power supply, front ends and interfacing with sensors.
The rest is all digital.


And that's why I mentioned there's power supply, protection, and analog signal interfacing circuitry. Generally speaking those are now available as components you can plug in your circuits, and are designed by scientists in the field. The core functionality is mainly digital, and to be proficient in using FPGAs and uP using your own circuit (vs. devkits and Arduino ilks) is much more essential than spending your life in a limited application area. It's fine for the sake of learning but overdoing it is simply a waste of time and effort.

You ain't gonna be much of a designer if you don't know how to get your signals in and out of that FPGA....
You can be a VHDL wizard but if you are tasked making a system , using DDR5 memory, a couple of PCI-X ports, and two SATa or USB3.0 ports and you have no clue how to do the layout , how to interconnect and terminate these signals properly, how to design the power supply , properly provide it decoupling and sequencing that thing is going to do nothing.

These days designers build SYSTEMS , not just 'i do code, you do analog , and we have a lab rat doing the polygon pushing to make the board'. That doesn't work anymore.

What you just mentioned except for the power supply, are basically digital stuff. Interconnecting digital components, memeory, FPGA, clocking, designing board all that stuff is digital. And I'm not talking about "coding"  Using HDL is not coding it's called automated digital synthesis, and it's one stage of the entire design process.

[IntegratedValve]

You can be a VHDL wizard but if you are tasked making a system , using DDR5 memory, a couple of PCI-X ports, and two SATa or USB3.0 ports and you have no clue how to do the layout , how to interconnect and terminate these signals properly, how to design the power supply , properly provide it decoupling and sequencing that thing is going to do nothing.

These days designers build SYSTEMS , not just 'i do code, you do analog , and we have a lab rat doing the polygon pushing to make the board'. That doesn't work anymore.

That is very interesting. Where can I learn something like that? Books?
I don't think we will learn anything about that at university. :-(

No books on this planet on how to create your own FPGA/uP embedded system from scratch. It's an art mastered at big companies or try it urself thing by deciphering lengthy data sheets and design guides.

[coodoo]

books :

Right the first time by Ritchey
High speed digtal design by Johnson
Anything you can get your hands on from Montrose

Thank you. I think the "Right the first time" is very understandable.

No books on this planet on how to create your own FPGA/uP embedded system from scratch. It's an art mastered at big companies or try it urself thing by deciphering lengthy data sheets and design guides.

I'm sad about this.