New on here, need a job, most complex way to light an LED?

[drew_the_mutant]

Like the title says, I want to make a demo board to take to interviews. I recently finished a college course, got very good grades, but the course was meant for people far younger than me. I want to demonstrate that I actually know what I am doing, as most of my personal projects are a big mess of wires and too fragile to transport. I thought a fun way to do this might be a needlessly complex LED lighting circuit.

An analogue button, (really a pot), connected to a microcontroller, connected to another microcontroller through a serial link, powering the LED through a MOSFET connected to a 5V supply regulated from a 9v battery is the sort of thing I was thinking of.

Any better ideas?

[duskglow]

I suggest not doing that.  If you're going to take a demo to an interview, make something that's actually useful.  It's not too much of a stretch from what you're planning to make a clock or something.

[drew_the_mutant]

True, a clock would be more useful. It would also be easier. Plans for clocks are on the internet. I would have no way to prove it is my own work.

[duskglow]

The point is not to make a clock, the point is that to show that you can make something useful, rather than just something weird to show you can do it.  The closer to "production" quality you can make it, the better you'll be able to show yourself off.

[Skimask]

Like the title says, I want to make a demo board to take to interviews. I recently finished a college course, got very good grades, but the course was meant for people far younger than me. I want to demonstrate that I actually know what I am doing, as most of my personal projects are a big mess of wires and too fragile to transport. I thought a fun way to do this might be a needlessly complex LED lighting circuit.

An analogue button, (really a pot), connected to a microcontroller, connected to another microcontroller through a serial link, powering the LED through a MOSFET connected to a 5V supply regulated from a 9v battery is the sort of thing I was thinking of.

Any better ideas?

I think you should get a "shake-light", generate the power to charge up a couple of super-caps before you can send the signal to turn on the led first.

And at the other end, maybe spin a motor to turn a generator which will be required to light a bulb which triggers a photosensor to move a servo to push a force sensor which triggers a solenoid to push down on  a switch.

oh the possibilities get stupid...

[Stonent]

Make sure you bring the chemicals to mix to make a battery

[drew_the_mutant]

"The point is not to make a clock, the point is that to show that you can make something useful, rather than just something weird to show you can do it.  The closer to "production" quality you can make it, the better you'll be able to show yourself off."

The point is to demonstrate that I have more knowledge than my qualifications suggest.  Suggest a project, that uses my skills, that I can complete in the same time, that is the whole point of this thread.


"I think you should get a "shake-light", generate the power to charge up a couple of super-caps before you can send the signal to turn on the led first.

And at the other end, maybe spin a motor to turn a generator which will be required to light a bulb which triggers a photosensor to move a servo to push a force sensor which triggers a solenoid to push down on  a switch."

When I said complex, I meant electronic stuff. I don't have the machines to build mechanical stuff.

[EEVblog]

Of no relevance of course, but the most complex way to light a LED is this:
http://www.maxwell.com/products/microelectronics/docs/hsn1000_rev3.pdf
Nuclear bomb goes off, LED lights up.

[drew_the_mutant]

Nuclear weapons are not really suitable. If I was willing to use them, I wouldn't need a job.

[AndyC_772]

If you're looking to demonstrate that you have skills which will be useful in a commercial environment, make sure it uses parts which are obviously relevant. The point is not that the board is particularly useful in its own right, but that the functional blocks resemble those which a real, commercial product would have.

So, for example: start with a switched mode power supply of your own design, and use it to power an ARM Cortex processor. Add a couple of sensors for, say, temperature and light level - something which doesn't require connection to external hardware to demonstrate. Log the measurements from those sensors to an SD card, display the last few readings on a graphical LCD, and provide an Ethernet interface which allows logs to be downloaded to a PC. If you can come up with a genuinely useful product that uses the same types of hardware, then do so.

Whatever you do, design it using the best PCB layout tools you have access to, get a proper PCB fabricated, and make it look saleable. Use surface mount components, and make sure you can solder them well. A few wire mods are OK as long as they're done neatly - but nothing says 'amateur' like stripboard, loose wires and DIL packages.

[cthree]

Here's an idea:

Instead of building something you think somebody else would want you to build, build something that you want to build and that you find interesting.

The point of having a portfolio is to show YOUR work. If your potential employer thinks it's stupid then you probably don't want to work there. How much enjoyment are you going to get from building things you think are lame? How many months or years are you going to waste in a dead end job when you could have been doing something you loved to do instead.

If you like games, make a game. If you like programming FPGAs, make something clever with one. If you like wireless data, build a sensor network. If you like robotics, make a robot. The point is you should build what you like to build and build lots of it.

Unless you make a time machine then no one blinking LED project is going to impress everyone. You have just as much of a chance that it will kill your chances as get you the job. If you make a time machine, you don't need a job. Make a lot of different things to show your versatility.

Also, if I may, I suggest you not build something and bring it to an interview. Instead, design and build something and post it on youtube. Start a channel and post all your projects. Let your interviewer see you work and you definitely won't go unnoticed. It's not easy, this is an optional extra, but if you can manage it you would open up a huge number of doors for yourself whether you are employed or contracting, the videos may even get you a job you didn't expect.

[vk3yedotcom]

Strongly suggest against 'most complex way to light an LED' approach.   

Good applied engineering is all about the simplest/cheapest/best way to do a job.

If the simplest is a 9v battery and a 470 ohm resistor, then why should I be interested in something more complex?  Especially if it's my company's resources and profit margin at stake?

The point is not to make a clock, the point is that to show that you can make something useful, rather than just something weird to show you can do it.  The closer to "production" quality you can make it, the better you'll be able to show yourself off.

I'd go further.    Maybe show that you have:

* Defined a problem, requirement or need

* Surveyed how others have done it

* Evaluated the shortcomings of these approaches or an extra feature you'd like

* Designed and constructed something that offers some benefit (eg simpler, cheaper, better) based on what was learned above

Let's take the 9v battery/resistor/LED approach as an example, pick it apart and see if we can make it better.  This isn't hard as it has many problems.  Eg 9v batteries are too expensive.  The 20mA's LED current may be too high. The LED might not be bright enough.  The resistor may dissipate (waste) current.  The upshot after all this refinement may be a bright LED torch that runs nearly forever on an AA battery (maybe using an efficient switching regulator and better LEDs).

I wouldn't necessarily suggest an LED torch as a project, but this approach is more useful than designing a convoluted way to do the same thing without regards to simplicity or efficiency. 

Another possibility is to find a circuit that someone's already done, build it and make it better/cheaper/simpler.   Then document it (preferably as you go along).  It might not be as 'creative' as starting something from scratch.  But it doesn't need to be to be worthwhile if there's some other benefit. 

Read here about 'Madman Muntz'.  The engineer who made his millions from snipping out unnecessary parts in TVs.  http://en.wikipedia.org/wiki/Madman_Muntz

[ejeffrey]

Definitely advise against a rube-goldberg apparatus for the sake of complication.  Unless you can explain how the design solves some (even made up) requirement, few people are going to care that you made a serial protocol to connect two micro-controllers.  That isn't to say you can't do some variation on your idea, just that you should have a reason why the LED needs to light, and why you need 2 microcontrollers (or whatever).  Hell if you have to say that the LED needs to be mounted on the inside of the warp core, so it has to get both power and signaling from a photocell illuminated by device outside the containment field.  Sure, you may be accused of playing into nerd stereotypes, but at least you have a reason for doing what you are doing.  If you want to show that you can solve problems and make products, you need an example problem to work with.  Obviously a real problem you solved would be better, and if possible the result should be more impressive than lighting an LED.

Also, try to focus the sales pitch on one aspect.  A whole bunch of individually simple ideas chained together but that otherwise have no relationship doesn't show much.  For instance, in your example you have a 5 V regulator to generate the supply for the LED.  Unless that has some relevance, leave it out of your description, even if you leave it in the design.

[MasterOfNone]

Disclaimer:
I haven’t been for an interview for many years and I don’t interview potential employees. Also I don’t work as a hardware design engineer anymore because I’ve changed to software engineering.

Comments:
There is some great advice above but the OP didn’t say he just completed an Electronic Engineering degree, he said he just completed a college course which was aimed at people younger than him. Based on this I’m going to assume he may not be going for design jobs initially.
I’m assuming what he was trying to do was creating something that demonstrates he has a broad range of knowledge that covers microcontrollers, communication protocols and analogue electronics etc. Maybe this is what you need to do for interviews in areas like testing, repair or PCB layout etc.
Also it’s possible that in some areas of electronics having a professionally finished board isn’t as important as showing the potential employer that you are keen to learn (on your own time) and you already have a good grasp of the main areas they need in the available role. If this is the case then maybe the complex LED controller wasn’t such a bad idea.  Maybe best approach is to build the system so that it clearly looks like a demonstrator rather than a useful product. The system could be built from a series of small PCBs which are neatly interlinked and mounted on a bit of Acrylic/Perspex using stand-offs. The PCBs could demonstrate key electronic knowledge areas, while the interconnections between the boards could demonstrate a different communication protocols or interfacing techniques.
My example:
Board1: Pot to Microcontroller. Output Parallel Data and Data/Ready on Port.  (Knowledge=Microcontrollers/ADC/Parallel bus)
Interconnect 1: Synchronous Parallel data bus (protocol i/f = Bespoke parallel)
Board2: Parallel to serial converter built from Shift Register & Oscillator (Knowledge=Digital Logic)
Interconnect 2: Synchronous Serial data (protocol i/f = Bespoke Serial)
Board3: Serial data (& Clock) into Microcontroller port. Output SPI. (Knowledge=Microcontroller/SPI)
Interconnect3: SPI interface (protocol i/f = Commercial Serial interface)
Board4:  SPI DAC and Transistor based amplifier to drive LED. (Knowledge=SPI/DAC/Analogue electronics).

The system could use a Battery based Power Supply with a switching regulator IC.

[AndyC_772]

I don’t interview potential employees. Also I don’t work as a hardware design engineer anymore

I do - and, potentially at least, I'd be impressed by someone who had built an interesting little project in their own time. Whatever the details, the fact that the device exists at all shows genuine interest in electronics beyond just having done a course and needing a job for the money. That's a really good start.

Beyond that, I'd be looking for familiarity with parts and techniques that will be useful in the job and which I won't have to teach. For example, if the project is built on a real PCB, I know the candidate has already started to learn about board layout and how to get one fabricated, so that's one important topic where I won't have to teach the absolute basics. If it's built using surface mount parts then that's something else I won't have to introduce, and a lot of time that won't have to be spent teaching surface mount soldering on scrap boards. If it's got a PIC or AVR on it then that's great, but if it's using an ARM processor, a proper FPGA, or anything else that's a bit more grown-up than the kinds of chips hobbyists use, then that's a major plus point and would really help the candidate stand out.

To anyone with even the most basic knowledge of digital electronics I can explain SPI in 5 minutes, so a device that uses SPI just for its own sake isn't impressive at all. But, if that same device just happens to use SPI to connect an FPGA to an LCD display, and uses that display to show something interesting on the screen, then that's much more noteworthy.

[MasterOfNone]

I don’t interview potential employees. Also I don’t work as a hardware design engineer anymore

I do - and, potentially at least, I'd be impressed by someone who had built an interesting little project in their own time. Whatever the details, the fact that the device exists at all shows genuine interest in electronics beyond just having done a course and needing a job for the money. That's a really good start.

Beyond that, I'd be looking for familiarity with parts and techniques that will be useful in the job and which I won't have to teach. For example, if the project is built on a real PCB, I know the candidate has already started to learn about board layout and how to get one fabricated, so that's one important topic where I won't have to teach the absolute basics. If it's built using surface mount parts then that's something else I won't have to introduce, and a lot of time that won't have to be spent teaching surface mount soldering on scrap boards. If it's got a PIC or AVR on it then that's great, but if it's using an ARM processor, a proper FPGA, or anything else that's a bit more grown-up than the kinds of chips hobbyists use, then that's a major plus point and would really help the candidate stand out.

To anyone with even the most basic knowledge of digital electronics I can explain SPI in 5 minutes, so a device that uses SPI just for its own sake isn't impressive at all. But, if that same device just happens to use SPI to connect an FPGA to an LCD display, and uses that display to show something interesting on the screen, then that's much more noteworthy.

Great answer however the OP may not be going for design types of openings initially. How many of those design of jobs would you give to a new starter without a Degree (or HND/HNC)?

[AndyC_772]

Whether the job opening is for a design engineer or an entry level repair technician, I'd still be looking for evidence that the candidate is genuinely interested in electronics for its own sake. That's the real benefit of bringing some gadget along to an interview, it shows a degree of enthusiasm for the subject.

It does, however, also make a statement about the candidate's knowledge level, and that may or may not have the desired effect. A board with a PIC and a few LEDs is great if you're applying for a technician's job, because it shows me you're keen, you have some good basic background knowledge and the ability to learn new skills. On the other hand, I'd really hope that anyone applying for a design engineer's role could do that already, so bringing along that same board is really telling me more about what the candidate doesn't know.

[cthree]

I don't think the point is to prove design ability or specific technical knowledge. Knowledge can be googled and skills learned. That stuff is all in your resume and part of the standard interview process (or at least that's what they think). If you are going to try and earn extra credits for initiative and creativity you should show some or why bother? There are literally tens of thousands of basic circuit designs available online, and in datasheets, and in app notes. Gazillions. Copying something you saw on youtube or duplicating a design someone else did from an app note probably doesn't help your cause.

My first assumption is going to be "this guy copied this from youtube" and I'm not going to waste a second proving it. I'll just figure I'm right and move on. That brings me to my last point in the previous post.

If you post your design, start to finish, on the web, on YouTube for example. If you upload the code to github. If you have feedback from other people validating your design and the mistakes you made, it offers credibility to the work. If you have others who then contribute to the work it adds even more credibility.

If you copy somebody else and post it as your own without credit you can be sure that some dickhead is going to point out how you ripped off the design and post a link to the original work. Transparency proves originality. Send me to youtube to see how you built it, what you were trying to do and what you ended up making in the end. Describe the goals, the restraints, the compromises. Talk about what components you chose and why (cheap, handy, specs, whatever) assemble them and then test the the project to prove it works. Show your methodology and if it's not ideal explain what tool you would use if you had it (and what you used instead).

Simply showing you can assemble a working circuit which has no context is pointless and redundant. What you want to say is: I can think for myself. I can solve problems. I can work independently. I can also work well with others and collaborate. I love what I do.

If they hire some other shmuck because of the school they went to or how they did on an interview puzzle or because they made a useless and expensive blinking LED then you don't want to work for them anyway. But, as I said before, do something fun and interesting for you because guessing what someone else wants is hopeless.

Anyway, my advice is probably completely wrong, give it whatever merit you think it deserves. I haven't had a job interview since 2002 and I didn't get the job so there you go.

[cthree]

One other quick point, if whatever you do answers the question "What is this guy going to do when I DON'T tell him how?", you've nailed it.

[MasterOfNone]

Whether the job opening is for a design engineer or an entry level repair technician, I'd still be looking for evidence that the candidate is genuinely interested in electronics for its own sake. That's the real benefit of bringing some gadget along to an interview, it shows a degree of enthusiasm for the subject.

It does, however, also make a statement about the candidate's knowledge level, and that may or may not have the desired effect. A board with a PIC and a few LEDs is great if you're applying for a technician's job, because it shows me you're keen, you have some good basic background knowledge and the ability to learn new skills. On the other hand, I'd really hope that anyone applying for a design engineer's role could do that already, so bringing along that same board is really telling me more about what the candidate doesn't know.

I agree.

Whether the job opening is for a design engineer or an entry level repair technician, I'd still be looking for evidence that the candidate is genuinely interested in electronics for its own sake. That's the real benefit of bringing some gadget along to an interview, it shows a degree of enthusiasm for the subject.

It does, however, also make a statement about the candidate's knowledge level, and that may or may not have the desired effect. A board with a PIC and a few LEDs is great if you're applying for a technician's job, because it shows me you're keen, you have some good basic background knowledge and the ability to learn new skills. On the other hand, I'd really hope that anyone applying for a design engineer's role could do that already, so bringing along that same board is really telling me more about what the candidate doesn't know.

I like the Youtube I idea, but I’m not sure how would get your potential employer to watch it. If you send them your resume (CV) with a covering letter, they may have to look through many others before they narrow them down to a few for the interviews. I’m not sure if they would spend any time looking at Youtube videos at the initial stage just because it’s on a covering letter. If you have a fix amount of time for an interview you probably can’t expect them to watch some youtube video on your tablet during it, they will want to engage with you and ask you questions. So how do you get them to watch it?
I think the OP came up with the whole idea of the complex LED controller because it isn’t practical so you wont find lots of designs for it on the internet. 

One other quick point, if whatever you do answers the question "What is this guy going to do when I DON'T tell him how?", you've nailed it.

I don’t think he expects to be told how to design/implement a solution, I think he’s just looking for ideas.

[cthree]

He did say he wanted to take it to an interview which infers that his résumé has been examined in detail.

Sorry if I wasn't clear, the "question" is a rhetorical one in the mind of the employer.

[ftransform]

God I hate some of the people posting in this thread. I love the study done that said almost all screening techniques used by interviewers are utterly pointless, as I suspected, and that people who do interviews are usually useless cogs in broken machines. 

[duskglow]

God I hate some of the people posting in this thread. I love the study done that said almost all screening techniques used by interviewers are utterly pointless, as I suspected, and that people who do interviews are usually useless cogs in broken machines.

It's all a game.  Everyone realizes it except for the people who don't know how to play it.

[ftransform]

its like witch doctors practicing voodoo rituals