Please also note that SW is a component in MTBF, for embedded systems (SDRs) like the K3. This is also true in consumer electronics (where I'm stuck for the moment). We figure in SW failures as a statistical value, but don't really count them as a full failure unless a unit is "bricked" and becomes unusable. When you make millions of anything, this does happen and needs to be counted. The case is quite different for a unique and lower-volume product like the K3.
For consumer products that need an unanticipated power cycle or reset to start working again, we assign a fractional failure. Usually 0.2 or 0.3 depending on the Annoyance Quotient to the end-user. Many consumer products have SW MTBF in the hundreds of hours because release intervals prevent thorough debugging and bug fixing. That's one factor, but so is software complexity. From anecdotal evidence (I'm not an Elecraft insider, so it's mostly from this list), it looks like the collective MTBF of the K3 due only to software has a much better record than any consumer appliance has. At least that I've heard about. It goes without saying that the hardware platform is bulletproof. But when something fails, it would probably occur first in switches and controls, followed by connectors, and last place would likely be the electronics itself. Be glad Wayne, Lyle, Eric, the Aptos crew, and a large assemblage of Field Testers all try to break the K3 firmware in advance of most releases. Now - I wonder if Wayne or Eric have any data collected on MTBF.... I bet that's one of those trade secret thingies.... :) 73, matt W6NIA On Wed, 28 Oct 2009 08:47:44 +0000, you wrote: >Matt Palmer wrote: >> Such a number is really meaningless to an amateur and without >> information on the method in which it was computed. > >I suspect the sort of things being referencedplgu here are: > >- many pieces of modern electronics fail through old age in less hours >than their MTBF (e.g. computer disk drives), which means they are only >useful if you have many instances of the product and replace them before >they reach old age; > >- MTBF doesn't account for things like lightning strikes and shorted >antenna cables; > >- for a semi-kit, many failure modes will depend on the quality of the >assembly process and anti-static measures; > >- failure rates are very temperature dependent and somewhat dependent on >other environmental factors, like humidity. > >- ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html
