The AD592 ( analog.com ) should be relatively noise free over reasonable distances. 1.0 uA/degree K. Bias it with something between 4 and 20 V and use a current to voltage converter near the computer. The time constant is pretty good if in flowing air and even better with a small Al heat sink.
HTH Dave On Sat, 2011-07-23 at 11:53 -0400, gene heskett wrote: > On Saturday, July 23, 2011 11:33:03 AM Przemek Klosowski did opine: > > > On Fri, Jul 22, 2011 at 3:15 PM, Igor Chudov <ichu...@gmail.com> wrote: > > > El cheapo ethernet routers cost $9.99: > > > > > > http://www.newegg.com/Product/Product.aspx?Item=N82E16833166034 > > > > > > I believe that a cheap Ethernet thermometer would cost roughly as > > > much, if someone wanted to produce it. > > > > Yeah but this Rosewill box is really a plastic box with five > > connectors surrounding a dedicated ethernet switch chip---a mass > > market item that's cheap because they make zillions. Anything else has > > volumes that are multiple orders of magnitude smaller. > > > > To get a good price, the only way is to judo the strength of the mass > > market---repurpose an OpenWRT router or a cast-off PC. Even then, > > however, you need the Ethernet cable AND the power line---unless you > > jigged the power (low-voltage DC, of course) over the unused pairs in > > the Ethernet. > > > > This gives me an idea--I think it might work to hook one of those > > two-wire or I2C Dallas/Maxim temp sensors over a long 'ethernet' > > cable, and bitbang them. Does anyone know what are the practical > > limitations on I2C/two-wire? A twisted-pair ethernet cable should > > help... > > Unforch, I2C is not really suited for long distances, the cables > capacitance limit is 400 pf, it is not a "terminated" transmission line > design by any stretch. No cat5 type cable has any advantage in its twisted > pair, differential receiver design. Neither do the std 4 wire phone > cables. 40 feet and I2C is Dead in the Water. > > That isn't saying that a 3 wire circuit, with active pullups that also > serve as terminators for echo & ringing control, couldn't be made to work > at 100's of yards, but the thing is going to need about 5 watts of power > available at both ends of the cable for driver and term power. If flat > ribbon cable was used, which has an impedance of about 110 ohms, then the > chips designed for active terms on a scsi bus could be 'borrowed', but for > other cable types, like 4 wire round or flat telco and cat5 twisted pair > styles, the term match will not be as close as they will range down to the > 60 ohms area. One might be reduced to looking at first one end, then the > other of the circuit and adjusting the terminating R for minimum ringing > and echo's as seen on a 100 mhz scope. Not practical for a just plug it in > and its supposed to work, even for folks who have no clue what 'VSWR' > stands for. > > The lesson is to stick with properly terminated twisted pair cabling if any > distance is involved. > > Cheers, gene ------------------------------------------------------------------------------ Storage Efficiency Calculator This modeling tool is based on patent-pending intellectual property that has been used successfully in hundreds of IBM storage optimization engage- ments, worldwide. Store less, Store more with what you own, Move data to the right place. Try It Now! http://www.accelacomm.com/jaw/sfnl/114/51427378/ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users