Group Actions
Engagement
  • Author Author: kkazem
  • Date Created: Date Created
  • Views 2118 views
  • Likes 10 likes
  • Comments 4 comments
Related
Recommended

Does Anyone Still Breadboard Anymore?

kkazem

Many years ago, I did a lot of breadboarding as that was the only way to determine if a circuit design would work and it could be debugged and improved. In those days, all parts were thru-hole. But these days, with many parts only available in SMT packages, it's almost not worth it to breadboard these days as we have options. Here are three good ones:

1. Do a PCB layout. There are still some free tools or low cost. You can design in any SMT part you want and optimize the layout. And once it's done, there are firms that will make you a professional PCB, up to 4 layers at 4" x 4" x 0.063" with double-sided silk screen and solder mask for $8 for 5 pieces. How can you beat that. That gives you a much better quality finished circuit compared to most breadboards. Many of these firm will do a low-cost assembly as well and supply many of the parts. You only need to supply them the parts they don't normally carry. Or, you can have them assemble the PCB short, then install the special parts yourself when you get the board.

2. You can quite accurately simulate many complex circuits for free using LTSPICEXVII which is free from Analog Devices Inc (formerly by Linear Technology Corp). Most IC manufacturers provide no cost SPICE models of their parts. And even if you plan on doing a board layout, it's often a good idea to simulate at least part of your circuit ahead of time to work out any problems.

3. There are special breadboarding boards you can get that have several SMT parts patterns already on the board, such as QFP for ICs or SOIC, and various resistor and cap sizes like 0805, 0603, and 0402. The drawback is that the layout won't be optimized. This is fine for low frequency circuits, but not so good for high frequency.

  • in reply to sokol07

    I don't think any one solution is the answer. But as time has gone by, it's harder and harder to get the parts I need in thru hole for traditional breadboarding..

  • I am developing a control circuit for a device at the moment. At the very early prototype version we are basing on some ready-made modules (by Polulu, ST Microelectronics, etc). I don't imagine developing the application for the microcontroller (and control app for PC) without breadboarding and than soldering things together onto universal boards.

    It is totally not optimized (however, I don't bother so much because the fastest protocol I have here is I2C), the wires are looking like spaghetti (which was dropped from a plate. At 2 floors height.) but it works, I can test it, run it and even (if packed in a nice box) show to an investor as a working, fully-functional model of the solution which will be optimized in future and the whole "electronics box" will become a single PCB.

    On the photo below there is the back (down?) side of the main control board (of course, that's work in progress, missing some elements, which are waiting to be placed there):

    image

    However, I discovered that among few breadboards I have only one is usable - the other two have got many loose contacts but also some contacts so stiff that it is hard to put the element inside. I also have many concerns about connection quality with this boards... I think that I trust the spaghetti I soldered on my own onto a universal board far more that the Chinese breadboards and jumper wires.

    Speaking of THT/SMD elements - all passive elements I have are THTs, however, this are only single pieces as at this stage most of the crucial elements are already placed on the ready-made boards. One of the sensors I use is Amphenol pressure sensor which I wasn't able to find in THT package and the solution was this little thing:

    image

    And from the solutions you proposed - solution 2 doesn't solve all my problems as I'm mainly prototyping not the electric part but the microcontroller app, its communication, interfaces, etc. I know that the simulations can be very precise, even more precise that the parts manufacturing sometimes and that's why I prefer to perform the tests "on a living organism" if only it is possible (luckily, in my case - it is). Solution 3 is, as pictured above, in fact something I am using now but, honestly, solution 3 is just a first step developing and soon I will be moving to the solution 1 - PCB prototyping and manufacturing is so fast and cheap that when I will have all necessary components I will just prepare the board, have it made and, if something won't be ideal, tune its design later.

  • I use them for generic experimenting - because the turnaround time to change something is short (minutes),
    sometimes as "holder" for a development board,
    and most often as patch bay: When I want to use I2C, SPI, CAN, ... between two devices, and in the mean time look at those signals with a scope and a protocol analyser.