On a product that will be produced by the boat load and where every penny counts it could be 95% out of 100%. You might sit there for a while agonizing about whether that 3c FET could possibly be replaced with a 1c BJT plus one 0.8c resistor. For a one-off or low volume industrial product it can be under 5%. This is for mainly analog/mixed designs which is my turf.
And yeah, I've spent hours finding a good source for a single custom inductor on a switcher. But those hours paid off nicely for the client, within just a few months of production.
Not just substantial work. Engineers need a good dose of clairvoyant skills. They must be able to sniff out whether a part might fall from grace withing then next 10-20 years and be obsoleted. They must also have a good smeller for vaporware.
What typically happens with many of my designs are the following: I look at various components and find that a good solution would be using component A B and C from company X with components D and E from company Y. I then find that I cannot get component B in smallish quantities or the price is prohibetive or whatever. Then often one have to chuck all the components from the design and start from scratch with a different set of components. I hence would say that for a fairly experienced designer - " Searching fro the right parts" takes more than 90% of the time. For very simple low quantity designs where cost is not an issue, "Searching for the right parts" would take less than 5% of the time, but in my case this is very seldom the case.
Considering that component selection is a very intricate and necessary part of the design/development process itself it is a little hard to extract that sort of detail out of the development logs. Not only have you the time spent on browising the net or catalogues but also the time you spend reading the datasheets of likely candidates, phoning or emailing the prospective suppliers for price and delivery quotations and the time spent in review meetings defending the selections you make. Whether or not this is more or less than 50% of the development effort will be hard to tell for most of us I should think.
I am guessing that you are more concentrating on hardware components in asking this question but, given the nature of this newsgroup, you should also consider the impact of selection of software components too (extraction from previous projects, library searches, bought in software components etc.). I wouldn't be surprised if the total hardware and software component selection was greater than 50%.
Paul E. Bennett ....................
Mmm, indeed. This is why - at some level - contract work is easier than day-job stuff. If you're building five of something, the NREs dwarf the BOM cost. So you can just read the datasheets, buy the right parts for the desired functionality, and add whatever glue is required in between to get the beast operational. If you're building five million of something, considerably more time is spent trying to find that magical single chip that contains all the peripherals you need and nothing else.
What's a development log? Seriously, what is it, how is it organized, and what is done with it?
What's a review meeting? (smilie on that one)
The real annoyance is vendors who make it hard to get price and availability, or to even locate a distributor, or who want a purchase order for two samples of a $2 transformer (actual recent case!) If I consider what my time is worth, and what time-to-market is worth, I often just bail on a part that's a lot of trouble to get the info on.
A design history. In some markets (like mine) you must keep this tidy at all times. IOW I must be able to go back to my just finished laser design x years from now. While I probably won't remember the exact schematic next month I must be able to find it and trace back as to why I have chosen this particular architecture. Same for specialty chips on there etc.
In Europe they call that "crusted up" sales strategies from yesteryear. Gets the boot from me as well. If pricing info is kept secret that typically forecasts trouble down the road. Not good.
Regarding time spent on component research: Is it OK when the client refuses to pay for (or gets bent when I ask him to pay for) component research?
I told one guy I was not going to answer his "how about this one" emails anymore (after a dozen or more of them) until he got serious about entering an arrangement, and he told me that I was being shallow.
To the OP: I agree that component selection is important, and is part of "engineering" (as differentiated from "design")
I learned my lesson early when I gave the barn away by being frank with a potential customer. He quizzed me about the topic and I never got the job. It seems he was just looking for ideas. I solved his problem for him and didn't get a cent.
I've even been on interviews where the so called hiring engineer asks how I would approach a job. Eventually you learn that some outfits fish for ideas with bogus job interviews.
I rarely have that happen but it does happen. Typically I catch it early on but often will still give them ideas if the investment of time is reasonable. It's like with car dealers that always see a certain number of people who test drive a vehicle but then vanish and buy it on the Internet.
Thing is, you might give away an idea and they then try it on their own or hire it out to a cheaper consultant. On complicated projects such potential clients (and I see them all as potential clients) often call back with some contrition in their hearts because this cheaper path didn't pan out. The one that didn't call back, BTW, went belly-up.
Look at it from the client's side: How can they gauge your capabilities other than by a detailed proposal? Most consultants like me cannot give out references, or at least not a lot of them, because often my agreements would be very specific about not divulging the identity of my client. Almost down to specifying which gun would be used in the execution. So, I have to at least let the potential client know how I'd begin to tackle a given problem. It has to make sense to them and they must get the feeling that this guy can really fix it for them.
Ha-Ha-Ha. There are only few things in this world that are cheaper then the good ideas. Good ideas are dozen per dime. You haven't got a job just because you think of yourself as a hell of important person. Nobody needs a power screwdriver with the habits like that :)