EEVblog Electronics Community Forum
General => General Technical Chat => Topic started by: eecook on June 26, 2019, 12:19:08 am
-
I stumbled upon this thread on quora about the programming growth stages:
https://www.quora.com/What-are-the-growth-stages-of-a-programmer (https://www.quora.com/What-are-the-growth-stages-of-a-programmer)
and naturally wonder what you guys think it would look like for a well rounded Electronics Design Engineer.
I think I would personally rank pretty low so I'll leave it to the masters who have walked and keep walking "the path".
-
Firstly they'd need to have worked in the service industry to have first hand experience on the impact bad design has on ease of serviceability.
-
(https://www.samaa.tv/wp-content/uploads/digital_news/2012-09-18/fat-and-getting-fatter-u-s-obesity-rates-to-soar-by-2030-8065.jpg)
Sorry, couldn't resist...
-
First it's an engineer egg, then it's an engineer larva, then it becomes an engineer nymph and after feeding on vast amounts of sarcasm and coffee it will moult a final time to become a full size engineer.
-
(https://www.samaa.tv/wp-content/uploads/digital_news/2012-09-18/fat-and-getting-fatter-u-s-obesity-rates-to-soar-by-2030-8065.jpg)
Sorry, couldn't resist...
:-DD
-
First it's an engineer egg, then it's an engineer larva, then it becomes an engineer nymph and after feeding on vast amounts of sarcasm and coffee it will moult a final time to become a full size engineer.
There's no butterfly in the process then....
-
(https://www.samaa.tv/wp-content/uploads/digital_news/2012-09-18/fat-and-getting-fatter-u-s-obesity-rates-to-soar-by-2030-8065.jpg)
Have you been spying on me ?
-
I like the French approach. You have to have been qualified as an electronics technician prior to becoming an electronics engineer. Certainly working in servicing especially to component level or even board level opens ones eyes to the importance of serviceability. I have seen some shocking designs by R & D staff (especially mechanical engineers) who paint themselves and their company into a corner because they ignore design-for-service, design-for-test and design-for manufacturing; costing a massive amount of money and time in the long run.
A well rounded electronics engineer swallows their pride and takes design reviews as being constructive. The real learning curve comes from listening to those with experience, especially with safety and EMC issues mitigation. Good engineers leave their ego at the door. I have know too many who believed the inference they were told at uni that they are a superior being. Another good trait is to have people skills. Again too many lack this basic requirement, some because they are slightly autistic, others because they are socially inept because of their upbringing. One fresh grad I worked knew everything or so he thought because he had two degrees and I only had one, which overrode my 35 year experience. He once told me seriously that I cannot pass too much data through USB WiFi dongle because the main chip will literally melt :palm:.
A well rounded electronics engineer will be the jack of all trades, but master of maybe one or two. In my opinion, a well rounded electronics guy will know how to code properly and know how to write documentation with the correct grammar and spelling. In addition, they should know the twelve times table and do basic mathematical calculations in their head.
-
1 to 15 years - Thinking you in know everything. Bitter at all those above you for being slow and out of date. This is a good time to transfer to a management position (where you can really piss off the good engineers with half baked ideas and the clout to force them to run with it)
15 to 30 years - Realizing how little you know. Relearning basically everything. Becoming a good engineer.
30 years + - Becoming increasingly bitter because no bugger will listen to your sage advice.
-
I like the French approach. You have to have been qualified as an electronics technician prior to becoming an electronics engineer.
Is that really the case in France? I could see the benefits, but it is the first time I hear about this. And I thought that the Bologna reform 20 years ago has standardized the academic tracks and degrees across Europe somewhat. Would you have a link explaining this approach? Does it apply to other engineering disciplines as well?
-
Dunno about others, but one of the main steps for me was that my cynicism gland fully developed.
-
Thank you for the replies guys! great input!
-
Sorry, couldn't resist...
This hits *way* too close to home...
-
I like the French approach. You have to have been qualified as an electronics technician prior to becoming an electronics engineer.
Is that really the case in France? I could see the benefits, but it is the first time I hear about this. And I thought that the Bologna reform 20 years ago has standardized the academic tracks and degrees across Europe somewhat. Would you have a link explaining this approach? Does it apply to other engineering disciplines as well?
I'm sorry to say that it's not the case. To become an engineer in France you have to graduate from an engineering school, which is usually a different path from getting a university degree (due to european harmonization, it's slowly evolving, but for now we still have these separate educational tracks that many other countries don't have). There is no requisite to have been a technician prior to getting the "engineer" title. It's basically a diploma, but it's not directly a master's degree, as it's the case in a lot of other countries.
Except in regulated engineering fields (such as civil engineers), you can have an engineering position in a company without having an engineering degree (as long as the company is OK with that), but you can't claim that you hold the title.
In many engineering schools, the very first internship you have to do in a company must often be at a technician level, so maybe that's where this idea came from. But that's just part of the mandatory internships you have to do, and that's only a couple months. You don't become a qualified technician after that. And you can work as an engineer right away once you have graduated.
-
30 years + - Becoming increasingly bitter because no bugger will listen to your sage advice.
Indeed!
Like when you want to use a 555 as a SMPS controller, or design a state-machine solely with discrete 7400-logic.
Now seriously, there is a grain of truth in hagster’s comment.
My pet peeve is that many designers nowadays are completely ignorant of design for manufacturing and/or serviceability:
“why would anyone require test points on the board? They take valuable board real state, which I have to fill with leadless semiconductors and 0201 passives.”
-
Somewhat related to the topic is how to work with management, project management, manufacturing, and other departments within an organization. After some amount of technical development, it becomes necessary to take on responsibilities in the infrastructure of engineering to support the engineering process.
At this same time one can also move into a specialization or two.
Take a little time to study and read up on a few management-related topics. A few that I liked:
"The Mythical Man-Month" by Frederick Brooks, second edition.
"Peopleware" by DeMarco and Lister, third edition.
"The Design of Design", also by Frederick Brooks.
-
I like the French approach. You have to have been qualified as an electronics technician prior to becoming an electronics engineer. Certainly working in servicing especially to component level or even board level opens ones eyes to the importance of serviceability. I have seen some shocking designs by R & D staff (especially mechanical engineers) who paint themselves and their company into a corner because they ignore design-for-service, design-for-test and design-for manufacturing; costing a massive amount of money and time in the long run.
A well rounded electronics engineer swallows their pride and takes design reviews as being constructive. The real learning curve comes from listening to those with experience, especially with safety and EMC issues mitigation. Good engineers leave their ego at the door. I have know too many who believed the inference they were told at uni that they are a superior being. Another good trait is to have people skills. Again too many lack this basic requirement, some because they are slightly autistic, others because they are socially inept because of their upbringing. One fresh grad I worked knew everything or so he thought because he had two degrees and I only had one, which overrode my 35 year experience. He once told me seriously that I cannot pass too much data through USB WiFi dongle because the main chip will literally melt :palm:.
A well rounded electronics engineer will be the jack of all trades, but master of maybe one or two. In my opinion, a well rounded electronics guy will know how to code properly and know how to write documentation with the correct grammar and spelling. In addition, they should know the twelve times table and do basic mathematical calculations in their head.
The old PMG/ Telecom Australia used to have a system of "Cadet Engineers" where they would pick up EE students who had done well in their first year, & then pay for the rest of the the course.
The "Cadets" didn't get to "dance & sing like Smurfs" in the break between Uni years, but were expected to work in the system, as well as taking "in-house" courses.
Many of these "Cadets", & arguably, the best, were people who were already internally qualified as Senior Tech/Technical Officer, with several years "under their belt"in these positions.
In both cases, "Cadets" were exposed to "real world" Electronics Engineering.
Seeing how various manufacturers do things is a good grounding in the area in which they wish to specialise.
I have since run into EEs who have no idea of how reputable manufacturers design & construct equipment, & hence produce designs with silly problems that an experienced Tech can point out in an instant.
And don't they just hate that!
-
To answer the original question, I think there's a direct parallel between what's described there and all other engineering professions, obviously not in the things listed for each stage, but in the thinking patterns. It's impossible not to develop them if you are going to successfully develop complex systems.
Also, I think that the first answer on quora is incomplete. Black belts are always relatively young for programming - what's after? Unfortunately, I think for electronics most people who get such responsibility to get to black belt level are a bit older, so there'd have to be a lot fewer people above that level. We can only hope that one of them lurks on this forum to give us the answer.
-
To answer the original question, I think there's a direct parallel between what's described there and all other engineering professions, obviously not in the things listed for each stage, but in the thinking patterns. It's impossible not to develop them if you are going to successfully develop complex systems.
Also, I think that the first answer on quora is incomplete. Black belts are always relatively young for programming - what's after?
As a bjj blue belt, I can tell you that a black belt has an intuitive grasp of what he is doing. During sparring things will come out as second nature with not much need to focus for them, and they can tell from miles who is at what level. It is not about physical power (or brain power in our case), although it will go long way. Also, there is an old saying that states that "a black belt is nothing more than a white belt who never stopped training", that mindset requires the realisation that even at that level you will be well aware that you still have a lot to learn and that requires humility, which I don't see a lot in the ego-driven software engineering world, specially with youngsters. My point is I don't think there are a lot of young programmers who hold a "black belt" out there.
Unfortunately, I think for electronics most people who get such responsibility to get to black belt level are a bit older, so there'd have to be a lot fewer people above that level. We can only hope that one of them lurks on this forum to give us the answer.
I agree 100%. Still waiting for that comment to pop up.
-
the following article may be of interest to the OP.
https://www.electronicdesign.com/analog-amp-mixed-signal/bob-s-mailbox-advice-young-engineer-julie-resistors-key-less-acceleration-an (https://www.electronicdesign.com/analog-amp-mixed-signal/bob-s-mailbox-advice-young-engineer-julie-resistors-key-less-acceleration-an)
-
its like playing poker with management
-
30 years + - Becoming increasingly bitter because no bugger will listen to your sage advice.
Indeed!
Like when you want to use a 555 as a SMPS controller, or design a state-machine solely with discrete 7400-logic.
Now seriously, there is a grain of truth in hagster’s comment.
My pet peeve is that many designers nowadays are completely ignorant of design for manufacturing and/or serviceability:
“why would anyone require test points on the board? They take valuable board real state, which I have to fill with leadless semiconductors and 0201 passives.”
the way even how most professors hammer cost into students its no surprise. they know they can get faster raises, promotions, vacations, etc if they do shit cheap because management is satisfied.
-
I think id break it down in to these:
1) Script kiddie: Where they can put together something from instructions online and copy paste in some Arduino code and eventually get it working but having no idea how any of it works.
2) Brute force experimenter: They then try to modify a design off the internet to do something else. Doing it by pretty much randomly tacking stuff on in non sensible ways, sometimes blowing something up, and eventually making it kinda do what they wanted with terrible performance. The design in the end being cringe worthy to any proper engineer and abuse of components everywhere (What do you mean a LDO needs cooling? its working fine)
3) Copy paste hero: They start to build up a collection of useful circuit snippets and gained experience on how they work so they stack these snippets of schematic together to make it do what its supposed to. Results are now more predictable but the designs might not exactly be optimal.
4) Ex-Optimist: They start to see how the world of electronics is not perfect. Those resistors suddenly don't just have resistance but also capacitance and inductance. Paths traces take on a PCB do actually matter for performance. The real world problems of EMI compliance and manufacturing sets in. They start to study datasheets carefully to chose optimal parts. Nothing is perfect or ideal anymore, just components that are less non-ideal. This is likely where they start to realize how much they don't know.
5) Proper engineer: They learn to overcome the imperfections of real world electronics, avoiding the common mistakes. Still scratching there head on the rare weird ones but usually eventually figuring them out after enough cursing and consulting google.
6) Gray beard: Having collected so much experience about how non perfect the world of electronics are they now know about most pitfalls and can foresee them in the design process. They have picked up lots of little tricks for unconventional use of components. They can look at a young engineers design for a few seconds and list out a bunch of problems that the rookie engineer does indeed end up having after shrugging off the cranky old mans critique of his 'brilliant design'.
-
An additional line:
7) Sage Hermit: Complains that electronics is not the same like it was in the past, talks about designing a whole ALU unit with discrete TTL gates, opamps should be powered from +/-15v supplies, knows all the relevant frequencies for analog TV: IF, chroma subcarrier, audio subcarrier, field rate, line rate for both PAL and NTSC. Looks with disdain at FPGAs, and boasts that real engineers should program in FORTRAN. An inverter consists of a DC motor coupled to an AC generator. Speed is controlled via a field reostat.
-
(http://devlog.datarealms.com/wp-content/uploads/2008/12/manprogress.jpg)
-
An additional line:
7) Sage Hermit: Complains that electronics is not the same like it was in the past, talks about designing a whole ALU unit with discrete TTL gates, opamps should be powered from +/-15v supplies, knows all the relevant frequencies for analog TV: IF, chroma subcarrier, audio subcarrier, field rate, line rate for both PAL and NTSC. Looks with disdain at FPGAs, and boasts that real engineers should program in FORTRAN. An inverter consists of a DC motor coupled to an AC generator. Speed is controlled via a field reostat.
Ha! I had a couple of professors like that! I remember one who used to talk about how he built entire op-amps out of discrete transistors , another one complained about the fact that nowadays you can buy integrated audio amplifiers and yet another one who made mixers out of discrete transistors. This last one absolutely hated SMD components and made his boards as big as he possibly could. He had an RF generator laying around his lab with nixie tubes as a display and it was fully functional which was awesome. Interestingly enough he was also a karate black belt and his last name was a female first name. Such a character!
-
Ha! I had a couple of professors like that! I remember one who used to talk about how he built entire op-amps out of discrete transistors
Like this? ;)
(https://cdn.evilmadscientist.com/catalog/emskits/741/kitv2/imglrg/1@2x.jpg)
https://shop.evilmadscientist.com/productsmenu/762#
-
Ha! I had a couple of professors like that! I remember one who used to talk about how he built entire op-amps out of discrete transistors
Like this? ;)
(https://cdn.evilmadscientist.com/catalog/emskits/741/kitv2/imglrg/1@2x.jpg)
https://shop.evilmadscientist.com/productsmenu/762#
Brilliant!!!
-
An additional line:
7) Sage Hermit: Complains that electronics is not the same like it was in the past, talks about designing a whole ALU unit with discrete TTL gates, opamps should be powered from +/-15v supplies, knows all the relevant frequencies for analog TV: IF, chroma subcarrier, audio subcarrier, field rate, line rate for both PAL and NTSC. Looks with disdain at FPGAs, and boasts that real engineers should program in FORTRAN. An inverter consists of a DC motor coupled to an AC generator. Speed is controlled via a field reostat.
Lots of us never reach that stage. We are delighted that we no longer have to deal with all the minutiae of designing from the most basic elements. We really enjoy being able to approach problems at a higher level. We relish being able to solve much more complex problems, without forming large, costly and difficult to manage teams. I see many youngsters who have a deep interest in doing things from scratch. This is great, if they want to learn and gain deeper insight than only approaching things at a high level gives. However, many seem to see doing things the old ways as a noble goal in itself, which I can't relate to.
-
I know the first stage. Use as many parts as possible to show you are a professional. And use the absolutely newest chips that won't be available three years after production to show you are into the latest electronics.
-
An additional line:
7) Sage Hermit: Complains that electronics is not the same like it was in the past, talks about designing a whole ALU unit with discrete TTL gates, opamps should be powered from +/-15v supplies, knows all the relevant frequencies for analog TV: IF, chroma subcarrier, audio subcarrier, field rate, line rate for both PAL and NTSC. Looks with disdain at FPGAs, and boasts that real engineers should program in FORTRAN. An inverter consists of a DC motor coupled to an AC generator. Speed is controlled via a field reostat.
Lots of us never reach that stage. We are delighted that we no longer have to deal with all the minutiae of designing from the most basic elements. We really enjoy being able to approach problems at a higher level. We relish being able to solve much more complex problems, without forming large, costly and difficult to manage teams. I see many youngsters who have a deep interest in doing things from scratch. This is great, if they want to learn and gain deeper insight than only approaching things at a high level gives. However, many seem to see doing things the old ways as a noble goal in itself, which I can't relate to.
Well there are often retried engineers that really knew there stuff back then but have not kept up with all this new fangled microcontroller stuff. But nothing wrong with messing about with the old electronics you know in your shed.
In my experience a lot of newer engineers tend to go towards the "Copy paste hero" type from my list. Electronics are getting quite a bit more complex in there fundamental workings and ICs are treated ever more as black boxes as the datasheets only give you the information of how to make it do the one thing it was designed for. So todays new engineers just accept the fact that things are complex and glance over any unimportant details. They wire the chip up according to the datasheet and call it a day. How does it work? Who cares as long as it works. The sort of stuff is becoming ever more popular with arduinos where they take a software library and use it blindly, if it doesn't work then just use a different library.
I have to admit its really easy to get caught up in "when you know how to use a hammer every problem looks like a nail" behavior where non optimal but already known solutions are forced against a problem rather than researching the proper way to do it. But the issue is made worse when you don't understand the inner workings of your black box solutions.
-
Well there are often retried engineers that really knew there stuff back then but have not kept up with all this new fangled microcontroller stuff. But nothing wrong with messing about with the old electronics you know in your shed.
In my experience a lot of newer engineers tend to go towards the "Copy paste hero" type from my list. Electronics are getting quite a bit more complex in there fundamental workings and ICs are treated ever more as black boxes as the datasheets only give you the information of how to make it do the one thing it was designed for. So todays new engineers just accept the fact that things are complex and glance over any unimportant details. They wire the chip up according to the datasheet and call it a day. How does it work? Who cares as long as it works. The sort of stuff is becoming ever more popular with arduinos where they take a software library and use it blindly, if it doesn't work then just use a different library.
I have to admit its really easy to get caught up in "when you know how to use a hammer every problem looks like a nail" behavior where non optimal but already known solutions are forced against a problem rather than researching the proper way to do it. But the issue is made worse when you don't understand the inner workings of your black box solutions.
The problem is that the field of electronics has broadened incredibly and still isn't slowing down. It's easy to look at people going the black box route as lazy, but you could argue it's actually what you need to get anywhere in an increasingly complex field. You can't be a specialist in everything. The trick is to know just enough about what's going on in the black box to understand and prevent strange behaviour without having to fully comprehend every aspect of it. We see the same in software development. Software is increasingly abstracted in ever more layers piled on top of each other. This has definite drawbacks, but also allows for effective development without having to build a computer from the transistor up.
-
I know the first stage. Use as many parts as possible to show you are a professional. And use the absolutely newest chips that won't be available three years after production to show you are into the latest electronics.
oh man better stick with that 741 huh ::)
might have to look at simple graphs in 30 years to find a replacement part
what is this vast array of obsolete chips you speak of?
are you still running nixies on circuits with barretters (maybe the janitor can coble some together in the broom closet with no one is looking if slyvania goes out of business).
sometimes I see cars on the street with old bulbs and I imagine a horse and carriage with torches in front of it...
what you are doing is running a feudal enterprise (yea we are gonna go out of business because on of the boards might need to be respinned) LOL. Or perhaps you are hooking up third world military forces (that don't have any fucking gunsmiths) with ammo? jesus
-
The problem is that the field of electronics has broadened incredibly and still isn't slowing down. It's easy to look at people going the black box route as lazy, but you could argue it's actually what you need to get anywhere in an increasingly complex field. You can't be a specialist in everything. The trick is to know just enough about what's going on in the black box to understand and prevent strange behaviour without having to fully comprehend every aspect of it. We see the same in software development. Software is increasingly abstracted in ever more layers piled on top of each other. This has definite drawbacks, but also allows for effective development without having to build a computer from the transistor up.
Yeah the field is getting way too big to know everything and one guy is not expected to know how every chip works down to the transistor level.
I'm talking more about the habit of not even trying to understand it in the first place. For example like taking a switchmode regulator IC, plonking down the same value components around it as the example schematic from the datasheet (Or maybe what a online automated tool like TIs WebBench spits out) route the PCB just like any other board and done. But all trough out this process not knowing how a buck converter even works. Why does it need that diode to ground? Why does the inductor saturation current need to be so much higher than the output current? Why is it 22uH and not 2uH? What about the weird RC mess hanging off the COMP pin? Pfft why would i need to know any of that. The people who designed the chip have to know this boring crap, im just using it to make my 5V rail and the datasheet already has the feedback resistor values for that in a convenient table. Just throw it together as the schematic shows and put 12V into it, done!
Sure you can't understand every little detail when you got a job to do on time, but one should at least make an attempt at understanding how it works. Sometimes its just a matter of copy pasting the fancy word from the datasheet into Wikipedia and reading the first two paragraphs and glancing over the helpful pictures. For example seeing how a switchmode converter works can really help you layout the PCB in a more optimal way so that it spews out less EMI.
-
oh man better stick with that 741 huh ::)
might have to look at simple graphs in 30 years to find a replacement part
what is this vast array of obsolete chips you speak of?
are you still running nixies on circuits with barretters (maybe the janitor can coble some together in the broom closet with no one is looking if slyvania goes out of business).
sometimes I see cars on the street with old bulbs and I imagine a horse and carriage with torches in front of it...
what you are doing is running a feudal enterprise (yea we are gonna go out of business because on of the boards might need to be respinned) LOL. Or perhaps you are hooking up third world military forces (that don't have any fucking gunsmiths) with ammo? jesus
What is wrong with you, coppercone?
Seekonk raised an absolutely valid point: Chips meant for consumer products can have very short lifecycles, which don't meet the needs of many industrial/professional product designs, intended for significantly longer product lifecycles. Young designers are often unaware of this; more experienced designers look for e.g. microcontrollers which the supplier has positioned for automotive use, and which come with availability guarantees for 10 years or so.
Having to respin major, complex PCBs every few years is technically possible, but economically a major drag -- you do want some time to develop new stuff too... Seekonk mentioned three-year life cycles being too short for this market, which is simply true.
Why do you have to build a "30 year" strawman argument and ridicule him? If you don't have experience with product design for industrial, scientific or professional markets, there's always the option of keeping your mouth shut.
-
I know the first stage. Use as many parts as possible to show you are a professional. And use the absolutely newest chips that won't be available three years after production to show you are into the latest electronics.
oh man better stick with that 741 huh ::)
might have to look at simple graphs in 30 years to find a replacement part
what is this vast array of obsolete chips you speak of?
are you still running nixies on circuits with barretters (maybe the janitor can coble some together in the broom closet with no one is looking if slyvania goes out of business).
sometimes I see cars on the street with old bulbs and I imagine a horse and carriage with torches in front of it...
what you are doing is running a feudal enterprise (yea we are gonna go out of business because on of the boards might need to be respinned) LOL. Or perhaps you are hooking up third world military forces (that don't have any fucking gunsmiths) with ammo? jesus
If you remain an engineer for a few years you might look back and wonder just how dumb you were to write something like that.
One of the biggest pitfalls young engineers fall into is believing vendors. They believe all parts in development will actually reach the market. They don't. They believe parts will fully meet the preliminary spec. They seldom do. They believe there are unlikely to be horrible delays while the new device is respun because of a problem so bad it can't be lived with. Huge delays are actually common. They believe the vendor is committed to the part. If its a consumer part the vendor will typically drop it if they don't start to get market traction pretty quickly.
Try looking in a piece of industrial electronics. Most of it looks like it was developed years ago, when it might have been released very recently. The main reason is the avoidance of here today gone tomorrow parts that might greatly simplify the design, but aren't guaranteed to be around for the 10+ year production cycle of the product. This problem prompts some silicon vendors to develop specific programs where they nominate certain parts in their range for long term users, and make promises that there will be availability for 5 or 10 or more years, and that provision will be made with a company like Rochester Electronics to ensure supplies are available for repair work even further out. You see this mostly with complex parts, which are typically the ones with the shortest production cycles, but you also see it for simpler parts where customers will only design in if they feel future supply is secure. At the top of the complexity tree you see specific parts (SKUs) from Intel, AMD, and the FPGA vendors nominated for long term use. In the middle you'll find things like specific variants of MCUs, memories, data converters, etc. nominated for long term use. You'll find niche analogue parts nominated for long term use. You can't buy most Intel Core 2 processors any more, but if you look at their long term support program you should find one or 2 SKUs that just don't go away. Without this the industrial designs would be really screwed, unable to use anything complex that they can trust.
Learning this real world practical stuff is one of the key differences between a fresh engineering grad and an experienced engineer. A truly cynical fresh grad is WAY ahead of the curve. :)