CurveBoards: Integrating Breadboards into Physical Objects | Hacker News Reader

CurveBoards: Integrating Breadboards into Physical Objects

hcie.csail.mit.edu

71 pointsArtWomb21 comments

AT high school breadboards were one of the most hateful pieces of equipment i ever had to use. The problem was that there was almost always a non zero number of defective tracks... depending on the kind of circuit sometimes track resistance could be significant too.

Sometimes the time saved by using a breadboard was heavily offset by the time required to double check each connection with a multi-meter.

Hopefully breadboards these days are higher quality... (or perhaps my school just had very old and crappy ones :-)

Yes, 3M makes a fantastic breadboard with very low contact resistance (https://www.digikey.com/en/product-highlight/3/3m/300-series...)

Get one, treat it well, and never look back.

As a student I would never have sprung for that price but as a professional it’s a no brainer.

As a professional the price is nothing but the value proposition is questionable. We simply don't use our breadboards very often around here, and when we do it's never for sensitive work. Basically, I've never had a case when I've cared who made our random stock of breadboards. Either they're definitely good enough, or they're definitely bad enough that I need to prototype some other way. (It's common to find that the only circuits I am interested in derisking at the breadboard level are the ones that I can't prototype that way, due to parasitics.) But then, I consider an SOIC-8 package pretty big in our line of work, so what do I know....

It's not just you. I haven't seen a breadboard used in professional work in over 20 years.

90% of the time you're going to have to do a PCB layout anyway, so you might as well start there. Even more so since fewer and fewer devices are even made in DIP anymore.

Hoo wow that's some sticker shock. I recently bought some cheap boards for some students I'm working with and they're absolute rubbish - about £2 each for a half size board. They're from Elegoo who make a lot of knock off Arduino-ware and are prominent on Amazon. The tracks are incredibly tight and I had trouble fitting their own Arduino Micro board into it.

Nowadays the cost of getting a board spun up is so cheap, that I don't bother breadboarding much, usually it's just as a means to interface a bunch of breakouts using 0.1" headers. It's useful to keep a bunch of SMT adaptors in stock for testing random SOIC or DFN parts.

Once you figure out simulation and EDA tools, breadboards are pretty much pointless. If the circuit is simple enough to breadboard within half a day, it's also simple enough to simulate effectively. If it's complex (normally due to high speed effects, parasitics and other things that are terrible on a breadboard) then the time spent getting it working in a breadboard is essentially wasted. You never know if it's the design or breadboarding at fault.

Soldered breadboarding (eg veroboard) has its place - at least then you can be certain there's not any electrical connection issues. But PCBs are so cheap and quick even that seems fairly pointless. It's very labour intensive for complex designs which would take less than a day to design and layout in EDA.

Do you have any software recommendations for prototyping digital circuits with dynamic input? Things like buttons and midi. The trial and error process of breadboarding is nice for a tinkerer but I’d like to do it all in software.

It depends what you're doing - for analogue circuits the free software LTSpice is pretty great. It's clunky and you need to follow some tutorials, but it's a brilliant way to sanity check an analogue design before you build it. You can set square wave voltages to see dynamic performance.

If you want to simulate logic then logisim is really useful - you can design your circuit in terms of logic gates and then change inputs to see how it behaves - or generate a truth table. Good for checking that your logic works as expected.

I had to take an electrical engineering course in college. That lab was the worst. You'd set up the circuits in the board and it wouldn't work. You'd ask the TA and he'd say you did it wrong. Check the design, check the pins, everything looks good. Still doesn't work. Two or three more rounds of this, finally get the TA over, he checks it over two or three times before concluding it's the board that's the problem.

Either that or the chips. There were always a few that were bad. So you didn't know if you had a bad board or a bad chip or were using bad paths on a chip.

Similar experience but we also had to deal with randomly failing multimeters and defective components. It was very frustrating but gave me a foundation for software debugging skills that proved valuable, not being able to take any component for granted and understand that even if most times is a human error sometimes the fundamental pieces of infra are also defective. However, sadly, it made me dislike HW design :/

When I was in highschool we never experienced a single faulty breadboard in ~2 years of electronics work. Had a bunch of fun with them.

The 3D design tool workflow for adding channels in the video is really cool! The tools do a lot of the tedious work for you, while leaving the more creative or opinionated choices to the designer.

Filling with conductive silicon is clever too - I wasn’t sure how they were going to add conductivity to a fully enclosed piece (no way to pop in metal bands).

I’m sure there are applications of this that make sense, and ones where this doesn’t make sense due to one constraint or another. Still, this a cool combination of many systems and a creative idea.

In my opinion breadboards can't die soon enough. Too many new engineers think they can use them for prototyping circuits where the parasitic of the breadboard make it entirely unsuitable. Low noise amplifiers, high speed designs, switching power supplies are all terrible applications for breadboard designs. And SMT is basically impractical on breadboards, which limits options far too much.

With the advent of cheap plot and go board shops about 20 years ago and the more recent explosion of unbelievably cheap Chinese vendors, there's just no reason to waste time between sim and layout. The era of the breadboard is over and has been for some time now.

I think there's still a little room for them in education and hobby design, but even there it should be a fleeting use as students learn the better way.

I buy the premise of extending breadboards for actual applications apart from testing circuits.

I recently made a butt triggered pomodoro timer[1], where I had to keep a trigger under... my butt, after messing with velostat and its unreliable resistance; A simple momentary button on a 170 pin breadboard served the purpose! So breadboards of different form factors as physical objects would indeed help to accelerate prototyping.

[1]https://abishekmuthian.com/butt-pomodoro-a-butt-triggered-po...

That's cute. Not clear that a solderless breadboard wristband is useful, but it's cute.

A way to design flexible printed circuits, taking into account that the components are not flexible, would be more useful. Design to a minimum bending radius, and bond to something that enforces that minimum bending radius, like a watchband. Orient the longer components so that the long axis is the non-bending axis.

I could see a wrist-mounted breadboard for prototyping wearables. Jam in a battery and a few SOIC breakouts plus some patch leads, hack up a quick firmware to test out an idea, kind of thing.

I'm not seeing it. The device package sizes are often smaller than even the pin spacing on a breadboard. We generally start with a manufacturer dev board to write the firmware while the hardware is being designed and then manufactured at a prototype shop. The problem with trying to make "breakouts" on packages so small is that they are extremely fragile (if you can even solder to them by hand!) and the poor signal integrity of flying blue wires can often be enough to make the hardware completely unusable.

Those images triggered my trypophobia mildly, even as someone who has worked with breadboards extensively.

A solution in search of a problem.

Maybe helpful for test-fitting how non-flexible components will behave on a flexible PCB, whether they interfere with each other or something else in the case? Maybe? I imagine you'd just 3D model it though and be happy with that. But maybe if you already had the case.

Reply on news.ycombinator.com