Test and Measurement Forum

Influence of screen savers on PicoLog ; date-time formatting

Forum for discussing PicoLog

Influence of screen savers on PicoLog ; date-time formatting

Postby RockDoctor » Thu Oct 06, 2011 9:13 am

I found time to get my DrDAQ wired up and logging data last night, but came across a surprise on examining the log data this morning. It appears (though this may be a coincidence) that the effectiveness of data logging decreased when the screensaver became active.

My setup, with (hopefully relevant) background information :
I'm at work, in a remote location where "mains" power is provided by one company's generators (in a ropy state - the radiator leaks several tons of difficult-to-obtain fresh water per day!) ; another company runs analytical equipment from that power supply ; I QC that second company's data ; sometimes the data goes insanely wrong, and they blame the first company's power supply dropping frequency from a nominal 50Hz to under 20Hz, but their power supply monitor is pretty ropy too (everything is ; cheapskate company).
Piggy-in-the-middle is not a comfortable place, so out of my own pocket I got a DrDAQ on my leave (I'd planned to get one one day ; this job is just my trigger), built some signal conditioning and tried to set it up last night to log the mains frequency, to resolve who is speaking "*****".
Is it working ? ... attached waveform from PicoScope : ... oh dammit, how do you attach a file in a PHPbb ... (is it a waveform or a settings file ... still new to the software. Well, the scope screenshot
test-1.png
Scope screenshot
tells me that I've got a signal that should be usable.

I left the device running overnight, but when I got back (24x7 cover work ; sleep when you can, and be thankful for it), saw that while I'd been getting a decent-looking signal when I went off to my tent, through a lot of the night I'd not been getting good data (spreadsheet chart, attached).
PL-output.png
Spreadsheet chart
.

What prompted this message was my initial impression that the bad data coincides largely with the time I was off-shift, and the laptop's screensaver (plain-old screen blank with password ; no dancing penguins) had cut in. But now I'm not so sure - the spreadsheet chart suggests that the data actually improved about the time I was swearing at my alarm clock. So I may actually have a real signal, though not the one I expected.

So, two questions : does PicoLog get upset by screensavers cutting in?
Is there a (relatively) simple way to make PicoLog log times in a human-readable form, e.g. ISO 8601 YYYY-MM-DD HH:MM:SS.sss?

(I've just been told that I'm leaving the continent tomorrow, so I'm going to have to pack up now.)
Attachments
20111005-0003-mains display.psdata
PicoScope settings file?
(36.65 KiB) Downloaded 152 times
--
It's a signature, dammit!
RockDoctor
Newbie
 
Posts: 0
Joined: Tue Sep 20, 2011 1:37 pm

Re: Influence of screen savers on PicoLog ; date-time format

Postby Hitesh » Fri Oct 07, 2011 4:06 pm

Hi RockDoctor,

The screensaver itself shouldn't affect the logging, although if your laptop goes to sleep then that would affect the USB connection to the device. How often are you sampling - the maximum number of samples per channel is 1 million.

With regards to time and date, within the PLW Spreadsheet window (click the blue rectangle icon in the PicoLog Recorder window) click the 'tick' icon (View Options) and select the Time format as 'Date/time'. This gives you the format DD-MM-YYYY HH:MM:SS.

PicoLog_DateTime.png
Screenshot of date/time feature.
PicoLog_DateTime.png (30.94 KiB) Viewed 2014 times


Can I ask which continent you were posting from? :)

Best wishes,
Hitesh

Technical Specialist
Pico Technology
Hitesh
Site Admin
Site Admin
 
Posts: 2058
Joined: Tue May 31, 2011 3:43 pm
Location: St. Neots, Cambridgeshire

Re: Influence of screen savers on PicoLog ; date-time format

Postby RockDoctor » Mon Oct 10, 2011 12:47 pm

Hi Hitesh,
Found the set-up options for the spreadsheet view and ... yes, there's a date-time format that I can hack about as I see fit. ... seems to be doing appropriate, usable data output.

I got stood down on Friday, back to the UK for a couple of weeks after just a week and a bit in Tanzania. Spent the morning doing a yo-yo impersonation between different parts of the Minerals Ministry in Dar es Salaam, getting the paperwork for exporting samples stamped in 17 different colours of wax and ink - "Welcome to Africa! Badai, badai!"

The sampling I'd tried to set up for was to read the frequencies once per second, and write those readings to file every second - observation suggested that some of the power fluctuations last for 5-10 seconds, before the genny's controller either does it's stuff, and boosts the injection volume/cylinder sufficiently to bring the prime mover's power generation into line with the draw on the electrical generator, and to bring the frequency back into the acceptable window. (To my tiny surprise, despite the rig and site being rigged for UK/ European 250V/50Hz, the gennys are running at around 61Hz, suggesting they're set up for chunky induction motors designed for US-standard 60Hz power supplies. Big deal - it's excursions down to <25Hz that are implicated in the data quality problem.).

1-sample/second would give me 86-odd thousand data lines / day and around 12 days before the file needs to be re-set / system re-started / whatever. That's a manageable file size (I can pre-process the data with other tools to find the "interesting" events if necessary, or hack a million lines of raw data in spreadsheets myself without undue concern), and a manageable amount of attention needed (I have other things to do in my copious free time on site. Like what is actually my job.), viz looking through the data on a weekly basis and determining if the date-time of bad-power events correlates with the insane data from the 3rd-party. At which point I can tell my clients where the problem is, or tell someone "we don't accept that excuse - fix your system or you don't get paid", then have sufficient data to back our position up.

Oh well, back to the UK now - clutching DrDAQ in hand, signal conditioner "bodge" in hold baggage (do I look stupid enough to try to take that sort of thing into an aircraft cabin?), and Linux laptop to play with when I get back home. Oh, and a 150V DC power supply to build for my friend's ex-WD Geiger counter - for which I wanted the 'scope in any case.

I'm fairly sure that getting PicoScope/ PicoLog working on the Linux laptop is going to be "fun", for quite small values of "fun".
Oh, I need to configure a signature block.
--
It's a signature, dammit!
RockDoctor
Newbie
 
Posts: 0
Joined: Tue Sep 20, 2011 1:37 pm

Re: Influence of screen savers on PicoLog ; date-time format

Postby RockDoctor » Mon Oct 10, 2011 12:51 pm

Well, sig block fixed.
Flight details now. And expenses.
--
It's a signature, dammit!
RockDoctor
Newbie
 
Posts: 0
Joined: Tue Sep 20, 2011 1:37 pm


Return to PicoLog

Who is online

Users browsing this forum: No registered users and 0 guests