30-second Philosophy - Wait or Rev Part 2
When you sell a house, the offer they make sounds great until you realize that the big number gets a lot smaller after accounting for replacing the roof, digging out the 60-year-old heating oil tank and fixing the 80-year-old plumbing. The purchase price for a new car seems affordable until the extended warranty, destination charges and "added dealer markup" costs get thrown into the mix. And an expected delivery date for an electronic assembly seems doable until the sole-sourced high frequency transistor inventory at the distributor goes from 12,384 to zero, the two-week lead time on the microcontroller becomes 52 weeks and the six-week resupply time on the power supply chip becomes "you will never see this part again in your lifetime so don't even ask!"
The wait or revise decision I wrote about a few weeks ago requires more thought and guesswork than the numbers on the table might suggest. Guesswork is the problem for me. I hate depending on luck or unquantifiable parameters. Too bad for me and anyone else with the same predilection. This is the world we live in right now.
But you can reduce the amount of guesswork by stepping back and throwing a few logical questions in. Here are a couple of factors that you might need to consider when deciding to wait or revise your design.
- Are you having an early-stage concept validate prototype built?
With a prototype, the decision can sometimes be easier. Your primary concern is a working design that will produce a specific outcome. There is likely more than one way to design a circuit to produce a given result. That being the case, you may be better off making a few design revisions to accommodate parts that are more available. Just remember to factor in long-term availability as well as short term availability.
- Are you trying to get a number of boards built so your software designers can get to work?
Maybe your production parts are out on a lead time that will work for production, but you need a couple dozen boards to your software team now. This is a case where you might be able to make a design rev to accommodate a more available component package. Say your production design needs a tiny micro BGA, which won't be on your dock for another 120 days, but the chip also comes in a big SOIC package. The SOIC won't work in the production case, but a larger dev board would not bother your software team at all.
There is a lot to consider when making the wait or rev decision, but by breaking your objectives down, you can often answer the question quicker that you might think.
The answer my friend, is blowin' in the parts bin.
With that in mind, jump over to screamingcircuits.com and quote your prototype PCB assembly. Just keep in mind that no matter who is building your boards, you might have to answer that "wait or rev" question.