The Serial Bit Rate Converter runs an ARM7 CPU which is complete overkill for the simple job of relaying serial data. I’ve often thought I should redesign it to run a cheaper and less powerful CPU like a PIC or AVR. I originally designed it with an ARM7 simply because I like them and enjoy programming them.
Another reason I never redesigned it for AVR is that the economics are not that great. It would probably take $5 off the price but cost me a couple of weeks of time to redesign the entire thing.
This decision has had an unexpected benefit recently. A customer asked me if I could add an extra function to search-and-replace byte sequences in the data stream. Now I’m really glad to have 50 MIPS on board because a smaller CPU might not have had the grunt to do the job.
So having a way-too-powerful CPU has produced a long-term benefit to enable me to quickly sieze a market opportunity.
I haven’t written very much about my goals here. My general long-term goal is to produce a range of for-sale products which might be either hardware, software or a combination of both, but my short-term goal is to simply become faster at designing and implementing stuff and that means:
- Being able to learn new technologies quickly
- Picking and choosing which design techniques to use to make a good match with my experience and strengths
- Remaining focused and managing my time well
I was quite pleased to get the Volume Box product from concept to production in only six weeks – especially since most of that time was spent waiting for suppliers and subcontractors, I probably only spent 10 days of my own time on it.
But of course I’ll have to try to be even quicker next time.
I already know what my next product will be but I am up to my eyeballs in contract work at the moment which, while it does pay the bills, takes up all my available time. So I will have to hold off until I get the contract jobs out of the way, then I can go hard on the next product and see how quick I can do it.