Worst PCB Ever?

[funkyant]

This was a post on Reddit, but was just too good not to share here:

https://www.reddit.com/r/electronics/comments/6efm66/worst_pcb_ever/



 

I used to work as a lab assistant for a local state university and have seen some amazing and some crappy projects come though.

The image above was supposed to be a "function generator" for a 400 level final project!  One of the FOUR STUDENTS assigned to this train-wreck came over to me and asked for help "diagnosing" what was wrong with their circuit board.  Not only did they not fundamentally understand that circuit traces could not be overlapped, but they had been trying to repair the trace they broke by using solder as wire. 

Normally, I am an extremely sympathetic person when it comes to situations like this, but they waited until the last minute, did not ask any questions beforehand, did not ask to have their circuit checked before they ordered it, and did not attend any of the many after-class PCB/soldering meetups that I often arranged.  It is the worst circuit I have seen to date, and figured I would finally share it with all of you.

[grumpydoc]

Wow

[P90]

project for modern art class?  LOL

[Zero999]

Did they fail or have time to redo it?

It looks like it was made in house, are the holes through plated? If so, how was it done?

I'm surprised the university got boards made in house, wouldn't it be cheaper to get them done externally?

[JPortici]

was that a fritzing job?

[funkyant]

Did they fail or have time to redo it?

I'd make a pretty solid guess that even if they respun it, they still would have failed 

[X]

Did they fail or have time to redo it?

It looks like it was made in house, are the holes through plated? If so, how was it done?

I'm surprised the university got boards made in house, wouldn't it be cheaper to get them done externally?

Some unis actually have their own small-scale silicon fab. Not likely to put Intel out of business, but when research-grade results are required, you'll be surprised at what length unis will go through to ensure they can control the process, which is very handy when you want to keep the number of independent variables under control.

[CJay]

Did they fail or have time to redo it?

I'd make a pretty solid guess that even if they respun it, they still would have failed 

Fourth year project with four students working as a team, I bloody hope they failed if that was the level of their subject knowledge.

[funkyant]

Did they fail or have time to redo it?

I'd make a pretty solid guess that even if they respun it, they still would have failed 

Fourth year project with four students working as a team, I bloody hope they failed if that was the level of their subject knowledge.

It's the:

they not fundamentally understand that circuit traces could not be overlapped

that's really the kicker! hahahahaha

[Jeroen3]

When art students try electronics.

[amyk]

part of me wants to think that this was done as a joke. Then again, given some people I've met...

[tatus1969]

hard for me to believe this is real. i have to show this to my 14yr old son

[Ash]

But it worked in SPICE.....     

[igendel]

What's wrong with crossing traces? Electrons that come from the left continue to the right because of momentum, electrons from the top continue to the bottom... it's simple physics people! 

[Cerebus]

ot boards made in house, wouldn't it be cheaper to get them done externally?

When I was at university (a relatively small one with ~3500 undergraduates) the university had about 4 student machine shops (excluding the specialist ones for metallurgy and materials science) and two prototype scale PCB shops. How do you expect people to learn if they don't get hands-on experience? If you don't, you end up with people who design things that can't be efficiently manufactured, or even can't be manufactured at all.

If you've got the size to have frequent need of some facility it's often cheaper to have it in house; after all, you're not paying someone's profit. Let alone the usefulness of having experienced staff on hand whose job it is to help you get the best, cheapest job done, not the most profitable or easiest to sell. Even if the economic benefits are marginal or even slightly negative, the convenience, access to processes for teaching and access to experienced technicians can make up for the economic angle with those intangible benefits.

Aside from teaching related facilities as above, the university also had one main and two satellite professional print rooms as it just made commercial sense to have that much printing capacity. Heck, even the student union had its own print room, that kept an ageing Rotaprint 30/90 in daily use.

[Dielectric]

Just  checking in to make sure it wasn't one of my boards.   

[EEVblog]

 

[mdszy]

I need to hang a picture of this up in my lab, that way every time one of my own boards fails to develop/etch properly, I can look at this and be thankful I didn't make something like this!

[CJay]

What's wrong with crossing traces? Electrons that come from the left continue to the right because of momentum, electrons from the top continue to the bottom... it's simple physics people! 

At certain frequencies that... No.

Just no 

[GreggD]

I have a hard time believing this but if true then not just the students but the school should find something else to do. Hope there are no student loans to be repaid.
I want to see the other side of the pcb.

[691175002]

I saw a lot of this kind of thing when I was at university.  If you aren't interested enough in the subject to do projects on your own time you just never connect classroom knowledge to the real world.

Its more prevalent (but less obviously hilarious) in computer science.  If you ask a graduating class to produce a simple application 80% of them will start by implementing a linked list from scratch in C.

[Zbig]

I saw a lot of this kind of thing when I was at university.  If you aren't interested enough in the subject to do projects on your own time you just never connect classroom knowledge to the real world.

Its more prevalent (but less obviously hilarious) in computer science.  If you ask a graduating class to produce a simple application 80% of them will start by implementing a linked list from scratch in C.

No. There's no sensible explanation for this other than them being dumb as a bag of hammers. I knew this thing wouldn't make any kind of sense when I was 10. And I'm not really that smart, either.

[tszaboo]

Not even close. The worst PCB I saw was some programing board for some old Motorola microcontroller. Obviously it was autorouted. And it had a lot of logic chip on it for the wierd timing programming requirements for the MCU. It had a bunch of NC pins.
...
...
...
Every NC pin was connected together by the autorouter.
And of course you dont spin the board again, you solve this with the scalpel. And if you fixed the wrong track, you fix your fix with a bodge wire.

[mdszy]

Not even close. The worst PCB I saw was some programing board for some old Motorola microcontroller. Obviously it was autorouted. And it had a lot of logic chip on it for the wierd timing programming requirements for the MCU. It had a bunch of NC pins.
...
...
...
Every NC pin was connected together by the autorouter.
And of course you dont spin the board again, you solve this with the scalpel. And if you fixed the wrong track, you fix your fix with a bodge wire.

Pardon my own idiocy... But what's the issue with NC pins getting connected together? If they're not internally connected to anything, what does it matter how they're routed?

[Zbig]

Pardon my own idiocy... But what's the issue with NC pins getting connected together? If they're not internally connected to anything, what does it matter how they're routed?

"Don't connect" often really means "don't connect" and not "don't care where you connect this as it's floating anyway". It's often some factory test connection or such.

[rfdes]

This smells like a bogus post.  i.e. fake. 

[Muxr]

That student definitely does not have the knack.

[tszaboo]

Pardon my own idiocy... But what's the issue with NC pins getting connected together? If they're not internally connected to anything, what does it matter how they're routed?

"Don't connect" often really means "don't connect" and not "don't care where you connect this as it's floating anyway". It's often some factory test connection or such.

Not only that, but the guy decided to call a bunch of pins NC. Like there was a counter, and most of the outputs were NC...

[mdszy]

Pardon my own idiocy... But what's the issue with NC pins getting connected together? If they're not internally connected to anything, what does it matter how they're routed?

"Don't connect" often really means "don't connect" and not "don't care where you connect this as it's floating anyway". It's often some factory test connection or such.

Not only that, but the guy decided to call a bunch of pins NC. Like there was a counter, and most of the outputs were NC...

Ohhh, they weren't NC pins on the device, but were marked as NC in the schematic and all shorted together? Yikes!

[CJay]

Even if they were NC it's always a good idea to check and see what the datasheet suggests for them, sometimes you find out that you can improve circuit performance by connecting them to ground, VCC or whatever.

[Bud]

NC pins can be used by the manufacturer for internal testing. As the other post said, check the datasheet what it says. If says nothing leave them alone and do not connect together.

[jonovid]

looks like schematic diagram went straight to pcb, oowps!  forgot the pcb layout!    no altium or EAGLE used here. 

[skarecrow]

What's wrong with crossing traces? Electrons that come from the left continue to the right because of momentum, electrons from the top continue to the bottom... it's simple physics people! 

No, crossing the traces would be just like crossing the streams.  It would be bad.  [emoji51]

Sent from my XT1565 using Tapatalk

[grifftech]

What's wrong with crossing traces? Electrons that come from the left continue to the right because of momentum, electrons from the top continue to the bottom... it's simple physics people! 

No, crossing the traces would be just like crossing the streams.  It would be bad.  [emoji51]

Sent from my XT1565 using Tapatalk

this works

[skarecrow]

What's wrong with crossing traces? Electrons that come from the left continue to the right because of momentum, electrons from the top continue to the bottom... it's simple physics people! 

No, crossing the traces would be just like crossing the streams.  It would be bad.  [emoji51]

Sent from my XT1565 using Tapatalk

this works

OMG they crossed the streams! Egon was wrong!

Sent from my XT1565 using Tapatalk

[X]

One of the Reddit users made a good point about the possibility that this was purposely rushed:

Probably an example of a clever strategy. If you don't have enough time to do everything you have to decide what you can dump. Projects take a lot of time and don't generate a lot of marks. Even if you can't find someone to fix it for you there might still be some marks in it and since you didn't spend any time on it those marks are basically free.

... and heck ... maybe you will get really really lucky and the thing will work in some sense. Sometimes when you are underwater, luck is all you got.

One of the most important life skills you learn in engineering school is how to do a half assed job quickly.

Perhaps the students aren't so dumb and they have excellent time management skills and can work within constraints.

[xrunner]

Mixing modern art and electronics - what's the big deal? 

[janoc]

I saw a lot of this kind of thing when I was at university.  If you aren't interested enough in the subject to do projects on your own time you just never connect classroom knowledge to the real world.

I can concurr - I did teach at a  uni and the things we got to see from time time ... mind boggles.

Its more prevalent (but less obviously hilarious) in computer science.  If you ask a graduating class to produce a simple application 80% of them will start by implementing a linked list from scratch in C.

Well, I had undergrads give me blank stares and then claim that my blackboard calculation is wrong, because "it is not what the calculator says!". Guess what - I have used fractions. Elementary school stuff. Total Chinese to them.

Or something like operator priorities - no idea. Many students I had kept punching numbers into calculators until one combination "worked" and gave the expected result. Thank God for the modern calculators that can handle such difficult problems like multiplying before adding for them.

Another such horror was a computer science master final year (!!!) student in a computer vision course who couldn't multiply two matrices. When confronted with it (it is freshman year stuff), he claimed he has never needed it - even though computer vision pretty much cannot be done without them and he has passed several other courses where they were needed.

So I can perfectly believe that the board posted by OP is real. Not everyone going to university and studying an engineering curriculum actually has the mental faculties for it. However, many are often very good at "winging it" just well enough to pass when there are too many students or the teacher/TA doesn't really care (exam resits mean extra work).
 
Teaching is not rewarded, at most unis you don't get any credit for it, so many profs will do only the bare minimum required. If you don't and try to care about what your students are doing you are actually shooting yourself in the foot, because then you don't have time for research and writing papers - which are directly tied to your lab's budget and the extension (or not) of your contract or chances at tenure.

Speaking from experience - spent 4 years as an assistant professor, had two semesters where I have been running 10 different courses with no TAs (insane!), the rest were 3-5 courses (normal load is 1-2 courses max). Then was told that "sorry, but your contract will not be renewed, we don't have an associate position for you". The guy who got it instead was a terrible teacher, but he was churning out papers like a printing press (and knew who to rub shoulders with). I had two papers published in that time - with 10 courses on my back I had barely time to prepare the classes and go to bed in time to get at least few hours sleep before another day of lecturing.

The result of this are university grads that are barely able to add two numbers together ...

[john_c]

When art students try electronics.

Mixing modern art and electronics - what's the big deal? 

Whoa, whoa. I'm an artist. Please don't make comments like that predicated on the idea that artists are stupid. Whoever made that PCB is a garden-variety dummy.

[Zero999]

Did they fail or have time to redo it?

I'd make a pretty solid guess that even if they respun it, they still would have failed 

What CAD software did they use? M$ Paint?

[mdszy]

Did they fail or have time to redo it?

I'd make a pretty solid guess that even if they respun it, they still would have failed 

What CAD software did they use? M$ Paint?

Almost looks to me like they just used a sharpie and a straightedge or something haha.

[CJay]

Did they fail or have time to redo it?

I'd make a pretty solid guess that even if they respun it, they still would have failed 

What CAD software did they use? M$ Paint?

Almost looks to me like they just used a sharpie and a straightedge or something haha.

Hey, I've used both, I've even transcribed PCBs from magazine layouts using straight edge and a bowspring compass. All of them were better than that thing...

[ZeTeX]

How can such an idiot be at University
how 

[Cerebus]

Almost looks to me like they just used a sharpie and a straightedge or something haha.

Except that the annular rings on the through holes look really neat and even, suggesting that it was produced with a actual PCB CAD package. If that's the case, how the frig did they manage to make such a pig's ear of the rest?

[691175002]

How can such an idiot be at University
how 

I've had a masters in EE tell me that its impossible for the battery charger to be outputting 2A at 12V because the benchtop supply it was being powered from was only showing 1.2A of output (at 24V).  He simply could not understand that a device could output more amps than it was taking in.

This was back when I still thought these people could be reasoned with, so I wasted about half an hour explaining that amperage is not power, and playing with the voltage knob to show that more amps are drawn at lower voltages.

The discussion was heated and went nowhere.  To be fair his research was in computer vision, but this is the kind of stuff you see a lot of.

[Muxr]

How can such an idiot be at University
how 

As a soft. eng I've worked with some PhD's who were very mediocre coders. I mean they were not terrible, just not anything special. And some of the most brilliant guys I've worked with were college dropouts.

It really comes down to passion. Passion and hard work is like 90% it imo. A formal education can help but if you don't have the passion for it it doesn't mean all that much.

Jim Williams is a great example.

Also some people end up going to a wrong school. They don't know what they want to be before they go to uni. Like after they enroll they realize that's not the field they want to be in. I feel bad for those who don't find the field they enjoy. I was lucky I knew exactly what I wanted to be when I was like 10 years old.

[amyk]

One of the Reddit users made a good point about the possibility that this was purposely rushed:

Probably an example of a clever strategy. If you don't have enough time to do everything you have to decide what you can dump. Projects take a lot of time and don't generate a lot of marks. Even if you can't find someone to fix it for you there might still be some marks in it and since you didn't spend any time on it those marks are basically free.

... and heck ... maybe you will get really really lucky and the thing will work in some sense. Sometimes when you are underwater, luck is all you got.

One of the most important life skills you learn in engineering school is how to do a half assed job quickly.

Perhaps the students aren't so dumb and they have excellent time management skills and can work within constraints.

You usually get 0 marks for not doing something. These students should be getting negative marks.

[P90]

How can such an idiot be at University
how 

I've had a masters in EE tell me that its impossible for the battery charger to be outputting 2A at 12V because the benchtop supply it was being powered from was only showing 1.2A of output (at 24V).  He simply could not understand that a device could output more amps than it was taking in.

This was back when I still thought these people could be reasoned with, so I wasted about half an hour explaining that amperage is not power, and playing with the voltage knob to show that more amps are drawn at lower voltages.

The discussion was heated and went nowhere.  To be fair his research was in computer vision, but this is the kind of stuff you see a lot of.

powered from 3-phase 120Hz DC... :-)

[james_s]

This has got to be a joke. There's no way a group of students made it to 400 level classes without being able to do better than that.

[P90]

yes, it's obviously a joke...

[StillTrying]

If the crossing tracks really were joined to each other there'd be a round blob at the crossing points, so it should work OK.

I think it's real.

[Someone]

This has got to be a joke. There's no way a group of students made it to 400 level classes without being able to do better than that.

yes, it's obviously a joke...

You've probably never taught students then, those lagging behind and scraping their way through the prerequisites can have astoundingly poor understanding of the material and get by with combinations of "group work" where they don't contribute anything that helps the group (or hinders it), "tutoring" assistance with their assignments, and in extreme cases "substitutes" in tests. The really surprising part is they don't see any moral issues with it and will justify how they are entitled to all the assistance they need to pass.

The best example I can provide was a final year electronic major, who given all the resources of the laboratory couldn't measure or even describe current. You can't even laugh about it at that point.

[mdszy]

This has got to be a joke. There's no way a group of students made it to 400 level classes without being able to do better than that.

yes, it's obviously a joke...

You've probably never taught students then, those lagging behind and scraping their way through the prerequisites can have astoundingly poor understanding of the material and get by with combinations of "group work" where they don't contribute anything that helps the group (or hinders it), "tutoring" assistance with their assignments, and in extreme cases "substitutes" in tests. The really surprising part is they don't see any moral issues with it and will justify how they are entitled to all the assistance they need to pass.

The best example I can provide was a final year electronic major, who given all the resources of the laboratory couldn't measure or even describe current. You can't even laugh about it at that point.

I was about to rage about "How could someone like that even be allowed to get a degree?!!!" but then realized: they're completely unhireable and will just end up getting a crappy job unrelated to their degree , which they likely are in massive debt for!

[Someone]

If the crossing tracks really were joined to each other there'd be a round blob at the crossing points, so it should work OK.

I think it's real.

Thats the sort of thinking that would lead to this. You could produce clever laboratory work where the students would need to simulate the design and then measure the real world and determine why it differs, very few would ever be able to come up with the answer. I could fully imagine a group of clueless students (having failed to attach themselves to competent partners/groups) sitting around the schematic in simulation and just dropping that onto a board without a DRC or any forward annotation.

[jwm_]

My guess about how this came about is that they attempted to split up the work.among the 4 person team. One person got schematic, another got theory and one dude got the make the PCB task. PCB guy probably slacked to the last minute, claiming he was on top of it, not asking for help because he already said he started it, until the night before it was due. Whipped this up, but wasnt the schematic or theory guy so didn't realize it wouldn't work. Brought it to his team members a half hour before it was due and they were like.. WTF. But soldered the components in anyway to turn it in so as to not get a zero, horribly embarrassed the whole time and shooting daggers at the PCB guy.  Not a fan of group assignments.

[james_s]

I rarely liked group assignments either. With the right group they worked really well, but it was too easy to end up with some slacker who was just along for the ride.

[CJay]

I've mentored quite a few students on hardware and software design. The bottom line is, anything can happen with undergrad students.
Some can't write a hello world in their 4th year (yeah, I know they are EE, not CS/CE, but come on).

On the other hand, there are some quite bright undergrad students that can take on a fairly large project with new technology on his or her own, or even manage a small team to solve an even bigger problem.

I found the best students were always those who took the courses to certify their skills, I.E. they were long term hobbyists who were probably, in knowledge terms, already at degree level before they started and just needed to be taught how to apply it.

They were the 'self starters' who, when tasked or even just shown a new technology, devoured it if it interested them.

One of my responsibilities in a long left job was to manage university students on work experience placements, they were almost without exception, utterly bloody useless, most freely admitted they'd taken the course because of the financial prospects and not because of any interest in it.

The very best, one who stands out in my memory (I'd love to know what happened to him) was a guy called Andy Papageorgiou who was in the middle of designing a Transputer based machine that got published in one of the electronics magazines of the time (mid 80s, if anyone has scans...).

[Red Squirrel]

lmao at the overlapping traces.   I think they were so used to seeing it in circuit diagrams where if you don't put a dot it means the lines arn't touching, that they actually took that literally in the PCB design. 

[Zbig]

I've had a masters in EE tell me that its impossible for the battery charger to be outputting 2A at 12V because the benchtop supply it was being powered from was only showing 1.2A of output (at 24V).  He simply could not understand that a device could output more amps than it was taking in.

I've had a teacher in my technical high school look down at me in disgust because, during the lab class, I haven't hooked up an ammeter just after the PSU (as the instructions called for) but took the readings off the meter conveniently built into the PSU instead (only some newer ones they've got had them). She said something along the lines of "and how on Earth do you think the current would have ended up inside the power supply!?". She just couldn't wrap her head around the idea that the current doesn't magically appear on the positive output terminal and then disappear to another dimension into the negative one. It was just too much for her to comprehend. Jeez, how I hated that woman. I think it was also her who insisted that the optocoupler working principle as explained in the book was just a kind of a "metaphor" to make understanding of the concept easier as she figured it would be impossible and/or silly for an actual LED to be shining on an actual phototransistor inside the IC package

[CJay]

I think it was also her who insisted that the optocoupler working principle as explained in the book was just a kind of a "metaphor" to make understanding of the concept easier as she figured it would be impossible and/or silly for an actual LED to be shining on an actual phototransistor inside the IC package

Your opto coupler comment piqued my interest, I've always taken it as read that there was a LED and phototransistor inside the package, probably as two separate die with some sort of transparent filler between, now I need to go and chisel one open to find out for myself

*edit*

Ah, the 'net to the rescue, x-ray images are brilliant. Though there's a 4N35 that's going to meet the Dremel this evening anyway...

[TerraHertz]

While I agree with everyone saying there are plenty of completely clueless electronics undergrads, I still think this is someone trolling. The original post is this:
   http://imgur.com/gallery/i8MEXct
Where there's zero evidence this was real.
If the photo was taken in a university lab environment, why not include some context in the frame?

[Cerebus]

I was about to rage about "How could someone like that even be allowed to get a degree?!!!" but then realized: they're completely unhireable and will just end up getting a crappy job unrelated to their degree , which they likely are in massive debt for!

Yeah, the problem is that unrelated job is "Management" and you get to call them "Boss". 

[pitagoras]

What's wrong with crossing traces? Electrons that come from the left continue to the right because of momentum, electrons from the top continue to the bottom... it's simple physics people! 

They could have even fix it with miniature traffic lights in the corners

[igendel]

What's wrong with crossing traces? Electrons that come from the left continue to the right because of momentum, electrons from the top continue to the bottom... it's simple physics people! 

They could have even fix it with miniature traffic lights in the corners

Now that's some original, out-of-the-box thinking! You're hired! 

[madires]

Promising candidate for a senior consultant at Accidenture 

[ZeTeX]

What's wrong with crossing traces? Electrons that come from the left continue to the right because of momentum, electrons from the top continue to the bottom... it's simple physics people! 

They could have even fix it with miniature traffic lights in the corners

And add a hospital in case one of the electrons crash into another

[janoc]

I was about to rage about "How could someone like that even be allowed to get a degree?!!!" but then realized: they're completely unhireable and will just end up getting a crappy job unrelated to their degree , which they likely are in massive debt for!

You don't realize that these guys are not there to get hired as EE (or another type) engineer. They are there to get a degree, often choosing the school only because it was easiest to get in or simplest to get visa or having cheapest tuition. Then they go home and become managers.

Not joking - had exactly this happen with some international students. Getting western education is a matter of prestige there, regardless of whether the student needs it or wants it. So they "study". All they really want is the graduation paper, in most cases they will never touch the subject they have studied in their lives anymore. One such guy got almost expelled for cheating, totally incompetent, barely scraped by but colleagues let him graduate. Now he is a C-level exec at a major company in Singapore laughing at how stupid we were working for the pittances we were paid.

I have met even a French guy (so it is not just international/third world stuff) who has been *proud* (and bragging about it) that someone else wrote his thesis and did his final project for him (under his "management"!). He was dumb as a brick, of course, anything he touched with his management "skills" turned into a major clusterfuck. And that fraud has almost managed to get a tenure at a fairly prestigious technical university by bullying ("managing") others to credit him for work he didn't do. He would have got it if there wasn't for one prof who actually knew him personally and put a kibosh on it by vetoing the process.

[tatus1969]

What's wrong with crossing traces? Electrons that come from the left continue to the right because of momentum, electrons from the top continue to the bottom... it's simple physics people! 

They could have even fix it with miniature traffic lights in the corners

Don't underestimate the abilities of these tiny little electrons, they have gone through worse than that. Below is a diagram showing their tricks with a BJT. Don't believe all that theoretical explanations with semiconductor junctions and charge carriers, this is how they really do it!

[pitagoras]

What's wrong with crossing traces? Electrons that come from the left continue to the right because of momentum, electrons from the top continue to the bottom... it's simple physics people! 

They could have even fix it with miniature traffic lights in the corners

And add a hospital in case one of the electrons crash into another

Problems arise when there is no electricity, traffic lights would not work, so we need also little policeman to take over the situation.

[Red Squirrel]

What's wrong with crossing traces? Electrons that come from the left continue to the right because of momentum, electrons from the top continue to the bottom... it's simple physics people! 

They could have even fix it with miniature traffic lights in the corners

And add a hospital in case one of the electrons crash into another

Problems arise when there is no electricity, traffic lights would not work, so we need also little policeman to take over the situation.

Or just make it a 4 way stop, you can even put "no left turns" and "no right turns" signs to ensure they go the right way.  Violators are just seen as noise and are the reasons for error correction further down the line, so you need a correctional facility in the circuit.

[German_EE]

"Or just make it a 4 way stop, you can even put "no left turns" and "no right turns" signs to ensure they go the right way."

Which is roughly what a microwave circulator does, I've always thought that a little bit of magic was involved with them.

https://en.wikipedia.org/wiki/Circulator