I forgot to mention one thing about what I need: I don't care how much about the latency between when the power meter computes a result, and when it's available to my Linux box. But I *do* care a lot about the precision of the time sampling in the report.
For example, it would be fine if I could only get an update from the power meter every 10 seconds, but that update contained a time-stamped log that reported power usage (either incremental or cummulative) in 0.1 or 0.5 -second intervals. On Wed, Oct 19, 2011 at 8:40 AM, Christian Convey <[email protected]> wrote: > Thanks for the ideas. > > You're right - I *am* looking for total energy expenditure (Joules, > Watt-hours, etc.) I have no specific desire to do that integration in > my own code. > > As far as sampling frequency, here's the deal: I'm working on a > system to estimate the power draw associated with a run (or set of > runs) of an arbitrary program on my Linux computer. Some of the > program runs will be brief, but I'm not sure how brief we're talking. > I'm no statistics expert, but I believe my power-draw estimates will > get pretty inaccurate unless the sampling rate is markedly higher than > the program's running time. I *can* limit my studies to programs that > run for a long time (so as to reduce the relative error stemming from > a slow-sampling power meter), but that would put an unwelcome > limitation on my research. > > > On Wed, Oct 19, 2011 at 8:24 AM, Charles Lepple <[email protected]> wrote: >> On Oct 18, 2011, at 10:18 PM, Christian Convey wrote: >> >>> I'd like to write a Linux app that will know, in approximate >>> real-time, how much power is being drawn by a computer plugged into an >>> outlet. >> >> Can you narrow down your definition of "approximate real-time"? >> >> The default poll interval of NUT drivers is 2 seconds, but usbhid-ups has >> longer intervals for some values since retrieving values over low-speed USB >> takes a non-trivial amount of time. >> >> If you are ultimately trying to integrate power over time to get energy >> consumption, you probably want a device that does this measurement for you >> (a watt-hour meter, basically). Then, the exact update rate doesn't matter >> as much - if you poll less frequently, the energy per interval will simply >> be proportionally larger. You can still divide out the time to get average >> power, but of course, that won't show surges as effectively. >> >>> Is NUT well-suited to this usage? >>> >>> And if so, can anyone recommend relatively cheap monitoring hardware >>> (ideally less than $200 US) that will meet my needs? >> >> I haven't gone shopping for an UPS in a while, but I'd think you would need >> to spend at least USD $300-400 to get an UPS with decent current-monitoring >> hardware. (The voltage-monitoring side usually is good enough even in >> low-end UPSes.) >> >> Do you need backup power as well as monitoring capability? What about power >> control? >> >> If you just want monitoring, you might consider one of the Watts Up >> products: >> >> https://www.wattsupmeters.com/secure/products.php >> >> (Disclaimer: I haven't personally tried any of them, and I can't vouch for >> any of their software, either.) >> >> You could also look at power distribution units (PDUs) but I have a feeling >> that the monitoring capability is also not available in the under-$200 price >> range. >> > _______________________________________________ Nut-upsuser mailing list [email protected] http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/nut-upsuser
