Screaming Circuits: October 2006

A little more on REP

If I ask for a definition of the the difference between prototyping and rapid prototyping, would the correct respond be "it's faster - duh"?

I would probably hear that a few times, but that really isn't the correct assessment. Yes, it is faster - duh, but it's faster because it's done differently. That is the key: "different." With mechanical injection molded plastic parts, before the days of rapid prototyping, you would send the design off for tooling. At month two of the three month process, you would ask for a few parts to be shot out for proto's. The mold wouldn't be polished and a few things likely wouldn't fit quite right, but it would give you close enough.

You could also hand assemble some proto's with sheet plastic and glue, but that won't tell you much about what is in the CAD system. It also becomes pretty much impossible with smaller and more complex designs.

With rapid prototyping, you use a completely different process. It's milled out on a CNC machine, or built up with a 3D printing stereo lithography system. It is a completely different path that delivers an almost identical result. It is more expensive per part then a hard-tooled injection molded piece, but it doesn't take a quarter to three quarter million dollars for the tool to be built up first.

That is the thing about Rapid Electronic Prototyping. It gives you a near identical result, but through a completely different process. Our machines, our programming system, our patented loader modifications are all set up to do that - give a near identical result to a production line assembled board, but without the setup costs associated with a volume line and without the variabilities and difficulties of hand assembly.

Each individual board may cost more, but there aren't the associated start-up costs. In some cases, it will even cost less then the hourly wages that would be paid to a technician hand soldering it. Of course, as with mechanical prototyping, with the new smaller and more complex parts, it may not even be possible to hand build.

Read our earlier post on REP

Duane Benson
Screaming Circuits

Backwards LEDs don't work

Can you tell where the cathode goes on this pcb image? The answer depends on whether the diode isLed_ambiguous  a rectifier, an LED, a uni-directional TVS, part of a daisy-chain and a host of other considerations.

Have you ever had an LED or other diode placed backwards? By us? By someone else? By yourself? We don't like it when that happens and we're sure you don't either.

To ensure the best accuracy, we recommend extra care in marking your diodes to remove any ambiguity. Simply placing a "C" for cathode adjacent to the cathode on the board works well, as long as the diodes are clearly designated as diodes (D1, D2, etc.). You can also use an "A" for anode or the complete diode symbol in silk screen.

Relying on +, - or _ are not definitive in what they indicate and are not recommended.

PDF with our diode markation guidelines

Duane Benson
Don't smoke your LEDs

Are you an Eagle CAD user?

Eagle_corner_1 Are you an Eagle CAD user? If you use the Eagle layout editor, we've just made your job a little easier with a simple User Language Program (ULP).

Before assembling all the parts on your pc board, we program our machines with the Centroid file. This file contains the XY locations and orientation of all of the SMT parts on your board. Previously, Eagle CAD users needed to run through a number of steps in order to create this file.

With the new ULP, click the "ULP" button in the Eagle board layout toolbar, double click our ulp file and the Centroid is created for you - in just three mouse clicks!

If you don't use Eagle, and are looking for a good quality, low-cost PCB CAD system, check out the CadSoft website. You can download a free version that will do double layer boards up to 4" x 3.2" or a fully capable version for a very reasonable cost.

Duane Benson
Eagle CAD user - sometimes

« September 2006 | Main | November 2006 »