Computers are great. I like them. But they aren't as reliable as simple physical systems. As for the delta T a bimetal can operate over it is a question of mechanical design. An electronic controller may theoretically switch over a change of a milli degree but in the real world the conductivity from the heater to the mass, the rate of heating, amount of mixing in a tank, rates of cooling etc can make for some temperature differences that are enormously larger than the sensitivity of the controller.
As for upset my daughters computer lost its CMOS settings when the wrong breaker was thrown. Failed CMOS battery? Doesn't lose setup info if powered down. Doesn't lose it every time either. She also lost a satellite internet transponder due to a glitch on the powerline last year. Expensive spike that one. You will also note Windows users are warned to not use windows for critical control applications such as life support. In industry numerical controllers are widely used but critical operations have staff during operation and redundant controls if failure analysis indicates. Redundant controls and voting? That helps failure rate but it is still not perfect. I think if you size the heater so that if it fails on while unattended so that it cannot cause a fire or other danger then maybe your computer controller is an acceptable risk. Maybe include a bimetal disk cutout at 10 or 20 above highest temp ever used. Appliance clothes dryers have a backup cutout. They have to. Too many fires otherwise and those controllers are much more reliable than windoze. I think old computers are great for data logging but be damn careful when you include them in the loop for control. Always have a backup safety or you may have a situation not bargained for. Kirk --- Michael Redler <[EMAIL PROTECTED]> wrote: > OK Kirk. > > Thanks for the reply. > > I have a couple of follow-up questions: > First, Is accuracy of your temperature set points > (or other I/O) a critical issue for controlling a > process? Because, a PID or Smith compensator control > can give you far better control of your process than > a bi-metallic strip for example. > > Second, I'd like to rephrase the question: Is there > any way (in your opinion) that obsolete computers > can be put to use for the purposes of generating > alternative energy? > > Lastly: Can anything really be called "foolproof"? > If I count the number of times that I have a power > outage/lightning strike on my property, I still come > up with a good risk assessment -- especially since > set points can be stored in memory while power is > down and I'm not sure if any control system aside > from a purely mechanical one, would be effective > during those relatively rare times that a mishap > occurs. > > I've made all of my arguments (so far) without > touching on the potential of a computer like a 486 > as a platform for control and data acquisition -- > all for the price of a ....bi-metallic strip. So, > after reading your reply and agreeing with its > logic, I'm still not convinced that the idea of > using "obsolete" computers is too far off base. > > Mike > > Kirk McLoren <[EMAIL PROTECTED]> wrote: > A temperature controller that would not upset > because > of lightning or other power line disturbance and > would > not lose its settings with power supply failure may > prove superior especially if less expensive. > > Thus a bimetal strip may prove best in the long run. > > I am in favor of data logging and alarms via > computer > but fool proof and ultra reliable control usually > means no computer. > > Kirk > > --- Michael Redler wrote: > > > > > Hi everyone, > > > > For those of you who are working on your process > > control for ethanol/methanol or biodiesel > > production, I was wondering about something. > > > > I hate to send anything to the land fill or scrap > > yard that I might be able to use later and I'm > sure > > you feel the same way. This philosophy has caused > me > > to become a bit of a pack rat (and dumpster > diver). > > My basement is neatly organized with old (486 and > > P1) computers, motors, and other > > mechanical/electrical devices from dissected > > printers and other obsolete or broken down > > equipment. > > > > I was thinking of developing generic software (in > C > > or assembler) for old computers to convert them > into > > process controllers. For example, it might make an > > effective temperature controller for your still. > If > > this works, I would take the programs and put them > > on a web page for download. If it's truly generic > > (compatible with all XX86 architecture), it will > be > > a matter of copying the executable file to the > hard > > drive of a computer, adding a line to the > > autoexec.bat file and making a harness > > (serial/parallel port to solid state relays and > > sensors). > > > > Has anyone tried this already and if so, can that > > person please share their experience with me and > > those interested in this thread? > > > > Thanks! > > > > Regards, > > > > Mike > > > > > > _______________________________________________ > > Biofuel mailing list > > [EMAIL PROTECTED] > > http://wwia.org/mailman/listinfo.cgi/biofuel > > > > Biofuel at Journey to Forever: > > http://journeytoforever.org/biofuel.html > > > > Biofuel archives at Infoarchive.net (searchable): > > http://infoarchive.net/sgroup/biofuel/ > > > > > > > __________________________________ > Do you Yahoo!? > Yahoo! Mail - now with 250MB free storage. Learn > more. > http://info.mail.yahoo.com/mail_250 > _______________________________________________ > Biofuel mailing list > [EMAIL PROTECTED] > http://wwia.org/mailman/listinfo.cgi/biofuel > > Biofuel at Journey to Forever: > http://journeytoforever.org/biofuel.html > > Biofuel archives at Infoarchive.net (searchable): > http://infoarchive.net/sgroup/biofuel/ > > > > _______________________________________________ > Biofuel mailing list > [EMAIL PROTECTED] > http://wwia.org/mailman/listinfo.cgi/biofuel > > Biofuel at Journey to Forever: > http://journeytoforever.org/biofuel.html > > Biofuel archives at Infoarchive.net (searchable): > http://infoarchive.net/sgroup/biofuel/ > __________________________________ Do you Yahoo!? Yahoo! Mail - Easier than ever with enhanced search. Learn more. http://info.mail.yahoo.com/mail_250 _______________________________________________ Biofuel mailing list [EMAIL PROTECTED] http://wwia.org/mailman/listinfo.cgi/biofuel Biofuel at Journey to Forever: http://journeytoforever.org/biofuel.html Biofuel archives at Infoarchive.net (searchable): http://infoarchive.net/sgroup/biofuel/