Almost Breaking News

2012-01-01T20:15:00.000-08:00

I'm beginning the new year with a software project! Gasp...

I've created a script using AutoHotKey that receives data from a PicAXE microprocessor and sends it to an openoffice.org spreadsheet. I added a few movement controls so that the PicAXE can maneuver around the spreadsheet well enough to create rows of data, for multiple inputs. For example, it would be a snap to recreate my "live data" page on the website, for a total expenditure under $20, assuming an available PC. All that cost money is the little PicAXE IC and possibly a serial to USB converter.

The script will open the spreadsheet, if necessary, and gives the user the ability to set the COM port properties. It's basically an RS232 receiver with a few special commands for tabbing around the spreadsheet. It isn't particularly fast, due to the lag of the spreadsheet program, but it seems to handle three channels sampling about once per second, maybe a little faster. The program seems to work flawlessly right now, so I just need to come up with a project for it.

Ending Americium Experiment

2011-08-23T17:43:00.000-07:00

I'm not seeing any effect on the decay rate of americium from solar activity, so I'm ending that experiment. I've been told that an alpha emitter like americium is the wrong isotope to use, anyway. Carbon-14 looks to be an obvious choice, but the source is fairly expensive and the ion chamber would need to be rebuilt to accommodate pressurization, since beta particles have a lot more penetrating power. Realistically, it probably isn't going to happen!

X6.9 flare produces negative result

2011-08-09T12:54:00.000-07:00

There was a really big flare today (X6.9) and I don't see anything out of the ordinary on the decay rate plot. Conveniently, the plot is right near the end of the week so there's over six days preceding the flare. I don't see anything, and my sensitivity is pretty high; the reported 0.3% variation caused by similar flares would be six divisions on the chart. I just see flicker noise.

Another negative result

2011-07-29T22:07:00.001-07:00

A good-sized flare just occurred (M9.1). But there's no sign of anything on the chart about 24 hours ago. There's a nice record of the previous three days, so this is a good run. The data is wandering only about 0.05%, so an effect of even half the should be obvious. That's about 10% of the reported change supposedly associated with a large flare.

Negative Result, sort-of

2011-06-08T14:44:00.000-07:00

There was a good-sized flare yesterday, but nothing is evident on the isotope decay rate plot. The plot went off-scale about two days earlier, presumably due to excessive lab temperature. It took me about a day to get it back into good working order, and the flare occurred about 32 hours later. There is no sign of a dip in the plot during that window. It would have been nice to have the earlier hours, but they were lost. Unless something was lost during that earlier period, this looks like a negative result. I'm pretty sure the off-scale move wasn't due to the phenomenon! : ) (wrong direction, for one thing)

2011-04-18T15:42:00.001-07:00

I have a version of the spreadsheet that captures all eight channels of the Dataq DI-148U and creates 8 individual charts. The spreadsheet takes data every 100 seconds for 24 hours, then starts over at the beginning. The user needs to start the macro about a minute before 0:00 UTC for best results. If anyone is interested in this spreadsheet, send me an email and I'll send a copy.

2011-04-14T16:20:00.000-07:00

A little more progress capturing data with a VB macro in Excel: http://www.techlib.com/science/livedata.htm (second plot) . Also, I positioned a portion of an Argo screen right below the spreadsheet plots, and it looks like I can grab that, too, with a single screen capture.

2011-04-13T16:59:00.000-07:00

I've found a way to make Excel plots with the inexpensive Dataq I/O devices. I have the DI-148U, but the technique will work with any of their products with some modifications. The experiments are the second plot at http://www.techlib.com/science/livedata.htm

2011-03-18T17:00:00.000-07:00

I've added a live Geiger counter to the web page. I just added a dust collector to enhance the sensitivity, so the reading might climb over the next few hours.

2011-03-06T10:06:00.001-08:00

Here's a page where I'll put all my live data plots:
http://www.techlib.com/science/livedata.htm

2011-03-04T18:43:00.000-08:00

I discovered a simple way to put live data on a web page. See http://www.techlib.com/solarflare.htm for an explanation. I see a daily sine wave superimposed on the data that I can't understand. It doesn't seem to be temperature or pressure and it's pretty much a sine wave. All I can think of is that the sun rises and sets. That would just be weird! What other daily variations are there?

2011-02-16T15:43:00.001-08:00

Since it looks slightly possible that there is a 24 hour effect, I'm changing the data taker to make 1 week plots with 4 hours per division, starting at 0:00 UTC today. I've added some filtering, too, since the action seems to be quite slow.

2011-02-15T15:34:00.000-08:00

Finally, an X class flare! The graph is suggestive, as usual: http://www.techlib.com/solarflare.htm . So, now I'm not sure that this isn't a real phenomenon. That's a pretty big dip, relative to typical charts.

2011-02-13T11:08:00.001-08:00

I will leave the flare experiment running, in case there's a X class flare. It's entirely possible that the effect only shows up at much higher intensities and M class flares, even big ones, aren't all that threatening.

2011-02-13T10:37:00.000-08:00

The solar flare theory is in trouble at my house: http://www.techlib.com/solarflare.htm . We just had a M 6.6 flare and the data simply shows no effect, certainly nothing on the order described. There are all sorts of reasons my experiment might be invalid, but my results don't require a reprint of all the physics books, so I'm officially a "flare denier."

2011-02-05T13:31:00.000-08:00

I have a better circuit for discharging an ionization chamber through the gate of the measuring JFET: http://www.techlib.com/newprojects.htm

I think this is a neat trick, if I do say so myself! : ) I realize it's probably as old as the hills.

2011-01-27T18:09:00.000-08:00

Another M-class flare today, but nothing significant on the plot. (see http://www.techlib.com/solarflare.htm)

2010-12-21T16:10:00.000-08:00

My decay rate plot was very flat for the last 24 hours, staying within 0.05% for the entire time, and the x-ray flux is nearly flat (see http://www.swpc.noaa.gov/ftpdir/plots/xray/20101222_xray.gif). I know that negative results aren't particularly useful, but I'm fascinated by the possibility that the noisier plots are caused by more solar "undulations." If I had a big grant, I'd try a cross-correlation experiment!

2010-12-18T07:24:00.000-08:00

I moved the experiment to a different room last weekend and it's had about a week to stabilize at the new location. This last week I saw three dips that seemed to correspond to three flares. In fact, I "predicted" the last two within minutes of their occurrence, simply by adding six hours to the minimum point of dips in the plots. These flares were small and I think this is just more correlating of independent, random events, but who knows! I'll keep it running in wait of some bigger events. Here's the new home for the experiment, off in a corner:

2010-11-08T19:01:00.000-08:00

Yesterday and today yielded flat plots, hovering over 2.5, near 3.0; nothing like the earlier dip on Saturday. As an exercise, I'll consider 2.5 and lower to be the "alert" level and see if I can "predict" a flare. I might read some tea leaves in the mean time. Actually, tea sounds great right now...

2010-11-06T10:43:00.000-07:00

There was a big flare today and an intriguing dip in the ion chamber output about 15 hours earlier. The funny part is that I was sitting at my computer speculating that there would be a big flare when I received an email alert and my SID receiver went nuts. But I would associate the two events if they happened anywhere within 24 hours of each other, so it's going to take a lot more to convince me. It wasn't the straight line I was expecting, I must say.

2010-11-05T13:48:00.000-07:00

The latest graph shows two day's worth of data along with the times of several flares, including two M-class flares. I don't see any obvious correlation.

2010-11-01T17:16:00.000-07:00

Today's plot was uneventful, but it looks like the system is working well. See http://www.techlib.com/newprojects.htm

2010-10-29T15:26:00.000-07:00

I opened the Keyboard Wedgie and increased the front-end gain from 25 to 300. Then I added precision resistors on each side of the offset potentiometer so that its total range is only 150 millivolts, instead of 2.5 volts (derived from a precision voltage reference IC). So, the two most obvious contributions to possible temperature coefficient should be greatly reduced. The PIC now sees a large 480 millivolt change for only 0.1% signal change.

2010-10-28T12:29:00.000-07:00

There might be a slight temp-co in the data taker; I'm using lots of gain and a big DC offset. I'm going to see if I can cram that part into the temperature chamber, too. The weather is starting to change in Austin, and it's conceivable that the coolest time in the basement is right at that apparent peak, early in the morning.

2010-10-27T18:03:00.000-07:00

I can't explain today's plot yet: http://www.techlib.com/newprojects.htm
The barometric pressure doesn't fit, nor does anything else I can think of.

2010-10-25T16:23:00.000-07:00

See http://www.techlib.com/newprojects.htm

That's as good as I was hoping to achieve. Let's see if it produces that same bump, then dip, seven hours before a flare!

2010-10-24T11:04:00.001-07:00

I think I've got it working well. I discovered a little trick for making incubation chambers a little more stable, on average. Simply put a small fan inside to stir the air a little. The downside is that the temperature will change more on a short-term basis, but the average temperature will be much more immune to ambient temperature changes. My ion chamber is in a large block of styrofoam, so minute-to-minute temperature changes are averaged nicely. The temperature is holding steady, well below 0.1 degree of wander. The output is still settling so I'll start some serious monitoring on Monday.

2010-10-21T09:08:00.000-07:00

The good news is that the barometric pressure has little to no effect on the reading. That's based on changing the pressure in a glove box by an inch of mercury and observing no change in the graph for hours. The bad news is that the sealed box converts temperature changes into pressure changes, so the temperature coefficient is pretty high. I'll look into just how manageable it is tonight.

2010-10-19T21:14:00.000-07:00

The plot looks pretty good (http://www.techlib.com/newprojects.htm). I'm going to characterize the pressure and temperature sensitivities to know exactly what I'm dealing with.

2010-10-18T15:52:00.001-07:00

Yes, it was a leak, I fear. Here's today's plot:

You can see the chamber settle, starting at midnight, then draw a fairly straight line for the whole day, despite plenty of solar activity. The sensitivity is good, with 0.1% of span being five counts (one small division). So, I just need another M class flare. It's pretty odd (and unlucky) that barometric pressure changes painted three similar graphs in a row that appeared to predict flares! One thing I noticed is that the curve settled much faster this time, and that's consistent with there being a slow leak before. I bet I get nothing but straight lines from now on. If I get another characteristic bump, followed by a flare, that will be convincing.

2010-10-18T08:56:00.000-07:00

I took a quick peak at the data this morning and the graph is settling more quickly than in the past. It should be flat by noon. The leak theory is looking pretty likely. Interestingly, there was a flare last night, about an hour after midnight, right on schedule.

We'll see if the thing can still predict flares. Who knows, maybe the barometric pressure predicts flares. Now, that would be a hard one to explain! : )

2010-10-17T17:19:00.000-07:00

I still think I'm gazing at N-rays. I actually suspect that there's a slow leak and I'm seeing delayed pressure changes. I've taken the box out of the oven and I'm applying copious amounts of sealant on every crack and screw. I'll let it cure for a while and put it back in tonight. It will drift all night and most of the day tomorrow. The graph was trending upward rather quickly when I made this decision. Occam's Razor tells me there's a leak! But watch for a huge flare overnight. : )

2010-10-17T12:53:00.001-07:00

This time I made the prediction to my wife several hours in advance. I'm getting my courage up. : ) You can see the slight bump, followed by the sweeping drop in the measured radioactivity. Right on schedule, the red plot shows a flare as detected by a GOES satellite. It happened about seven hours after that peak.

2010-10-16T19:04:00.000-07:00

I don't know, guys; the chart is returning to "normal" despite barometric pressure changes. I'm beginning to suspect that this was a legitimate catch. See http://www.techlib.com/newprojects.htm for the latest pictures and plots.

2010-10-16T13:25:00.000-07:00

We're having a class-M flare as I type. Here's the 60 kHz SID receiver output:

Ignore the graph after the steep decline on the right side of the huge pulse; that's yesterday's data. This thing pegged my receiver! Now, here's the intriguing part. I've been puzzling over why my decay rate chamber started drifting this morning, without any external cause that I can identify:
I wasn't careful making the graph, since I'm still in trouble-shooting mode, so the time scale is a bit off. There was an apparent peak at about 8:10 this morning, followed by a dip that spans six or more hours. The flare was at about 2:10 PM, or near the end of the plot immediately above. Did I just predict the flare? The one thing that has me optimistic, is the long, flat plot over night. I'm still concerned that I'm seeing barometric pressure, but this curve doesn't really look like the pressure curve for today.

2010-10-12T20:31:00.000-07:00

Another day, more issues. I suspect it will be a couple of days before I get everything ironed out for good, stable runs.

2010-10-10T20:22:00.001-07:00

I had to re-open it today; I forgot the desiccant pack. I also improved the seals since I might have been seeing a barometric pressure effect. It was small and will be easy to compensate, if it turns out to be real. Despite these issues, I think I'll be able to see changes on the order of 0.05% without much trouble. Who knows, maybe something is lurking at that level with this short resolving time. Why do I feel like I'm trying to detect N-rays? : )

It will be late tomorrow before it's completely stable, judging by previous tests.

Here are some better photos of the project:
http://www.techlib.com/newprojects.htm

2010-10-09T21:11:00.000-07:00

Here's the finished experiment. I poked holes in the aluminum foil to balance the pressure. The outer box will be sealed. The case is supposedly sealed, but I added Plumber's Goop around the edges. It would be nice to eliminate the barometer from the setup. The circuit uses an AD210AN isolation amplifier with a carefully biased (for zero temp-co) JFET front-end. I increased the chamber voltage to 50 volts and I'm getting 100 pA. I switched the polarity so that I'm collecting electrons on the sense wire, but it doesn't seem to make much difference. A 10 gigohm resistor converts the 100 pA to just under 1 volt which is perfect for the data taker. The whole thing is back in the temperature chamber. Already, the minute-to-minute variations seem smaller, perhaps under 0.1% of span. That should make it possible to see variations well below 0.1%. That's a little premature since I haven't see a full day of the wander.

2010-10-08T12:51:00.000-07:00

Here's the start of the new chamber. It has two elements in the bottom with the collection wire directly above. The far end of the wire is secured with styrene and the feedthru is Teflon. This chamber will have a cover and be mounted in an airtight metal case that will house the electronics. I lined the inside with aluminum foil to prevent any charge from accumulating on the surface of the plastic liner. The stained areas are silver paint dabbed on to make sure there is good electrical conductivity.

2010-10-03T22:07:00.001-07:00

I think I need to make a sealed ionization chamber. The thing is a smoke detector, after all, so it might be reacting to something in the air, including humidity. Maybe I'll build a new one this week. That will also give me a chance to add another element or two and to calibrate it better.

2010-10-03T10:22:00.000-07:00

For the last six hours (about 11:00 AM to 6 PM UTC) the plot has been dropping. The temperature chamber is steady and the barometric pressure doesn't fit the shape. There is still humidity that hasn't been characterized. I really need a sealed chamber. Anyway, if a bunch of solar activity happens later today, don't say I didn't warn you. : ) I don't think this is humidity, by the way; we started doing laundry a few hours ago, and the humidity takes a big step up. I don't see that in the data at all, and the downward slope began much earlier. I suspect there's something else I need to control.

2010-10-03T09:51:00.000-07:00

I should mention that the scales in the graphs have a large offset. To make a long story short, a count of 10 is about 0.3% of full scale.

2010-09-30T19:48:00.000-07:00

The old incubation temperature chamber was wandering a couple of degrees, so I decided to add a more precise temperature controller:

I installed a solid-state relay inside the chamber, simply in series with the line cord, and I built the new controller in its own box (above). I had to add a little heat sink to the RTD to couple it to the air better, and mount it close to the heater coils, but, other than that, it was quite straightforward. I hand-picked the resistors to give a range of about 25 to 225 degrees Celsius. The chamber is holding better than one degree, and that's outside the foam insulation surrounding the ion chamber, so the temperature variations should disappear.

2010-09-26T10:13:00.000-07:00

The charts for the last few days have be totally unremarkable. The above sample is from earlier today and runs for about nine hours. I believe the earlier "signal" had something to do with the power outage, perhaps an early glitch preceding the outage. Or maybe this thing also predicts power failures! : )

The sensitivity is about 0.3% of full-scale per division. I would think that even a 0.1% shift would be apparent and the articles talked about flares and seasonal variations on the order of 0.3 to 0.4%. I determined the sensitivity by covering the americium with some duct tape and noting the drop in the reading. My estimation is probably a little low.

2010-09-22T15:56:00.000-07:00

Take a look at that! The above chart is the decay rate of the americium in my ion chamber. The dip starts right at 4:10 UTC on 9/22/10. It turns out that there was a lot of activity leading up to this time. There was a power outage about an hour later, hence the drop at the end. I don't think that affected the earlier readings. Here's the previous day's data from NOAA:

4940 +     1811   1814      1816  G14  5   XRA  1-8A      B3.7    5.9E-05   1109
4950 +     1934   1945      1949  G14  5   XRA  1-8A      C2.1    8.5E-04   1109
4960 +     2059   2103      2105  G14  5   XRA  1-8A      C1.2    2.7E-04   1109
4970       2219   2225      2227  G15  5   XRA  1-8A      C1.0    2.7E-04   1109
4980 +     2338   2343      2345  G14  5   XRA  1-8A      B8.2    2.1E-04   1109

And today:

4990       0306   ////      0434  LEA  C   RSP  066-164   CTM/1                
5040       0557   0608      0700  G14  5   XRA  1-8A      B9.0    8.8E-04   1109

So there's 10 hours of unusual activity preceding the dip and a radio signal during the
dip, plus a final flare at the end, to mull over. I'm still a skeptic, but a few more of
these, and I'll begin to wonder.

2010-09-20T17:09:00.000-07:00

Here's a nice SID caught with the 60 kHz receiver. That corresponds to a C2.1 X-ray flare.

I have been "secretly" plotting the output current of an ionization chamber with a smoke detector element inside (americium), in an effort to see the strange effect I mentioned earlier. The ion chamber is in an insulated box inside a temperature chamber and an electronic barometer is used to remove the effects of barometric pressure. Normally, the chamber draws a straight, noisy line. Here's todays plot:

Notice that glitch about 40% into the plot (about 8:00, UTC). Each division represents about 0.1% of the full-scale current. (There's another glitch later on, so maybe another flare is coming.) Glitches like that are rare on this plot, but it seems like a reach to associate that with a flare that occurred at 19:40. But, one paper associated a similar leading anomaly with a flare several hours later. The idea is that something happens within the sun that creates a lot of neutrinos or something else and those escape at, or near, the speed of light, whereas it takes hours for the mechanical effects to reach the surface of the sun. I'm going to need a really big flare and a corresponding big glitch on this plot to be convinced. Well, maybe not even then.

2010-09-03T15:13:00.001-07:00

While trying to find out how intense a solar flare would need to be to detect an increase in radiation in an ion chamber on the ground, I stumbled upon an amazing report: http://www.astrobio.net/pressrelease/3599/a-strange-solar-flare-connection. That sounds like a non-linear effect. If that can happen here, it must be a huge factor in or on the sun itself, where the neutrino or the "whatever-it-is" flux must be much greater. This might be big...I'm tempted to put my pitchblende in front of my ion chamber and plot the current vs time!

2010-08-28T13:59:00.000-07:00

I've set up a large-area ion chamber to see if I can correlate increased background radiation with a SID. I seem to be seeing an increase in background during the day, so maybe there's hope. I have no idea how much, if any, of the X-ray flux makes it to the ground. There might also be a lagging response from protons, etc. Now, I just need the sun to wake up again.

2010-08-15T15:32:00.000-07:00

Here's one more capture of a SID today, using the Oddball SID Receiver. This was reported as a C5.4 X-ray burst. The time scale is local time due to a "cockpit error." Add five for UTC.

That's pretty darn good, if I do say so myself! : )
I forgot to start the program last night, hence the dead trace at the beginning.

2010-08-14T18:17:00.000-07:00

Today, there were two fairly low magnitude X-ray bursts:

This is a graph from the Sid Seizer operating at 60 kHz.

2010-08-13T22:11:00.000-07:00

I bought a few of those cheap LED flashlights because the machined aluminum cases look like a great project housings. Here's a start of a new C-Beeper probe:

I removed the switch on the rear and replaced it with perforated aluminum for the speaker. That's not a bad housing for $3, plus I have nine bright LEDs, to boot.

Soil Moisture Meter Flops

2010-08-10T20:56:00.000-07:00

It looks like determining soil moisture by measuring the humidity in a volume of trapped air in contact with the soil isn't going to work very well. The humidity tracks nicely as water is added to the soil, but the humidity stays high well after the soil dries. It looks like ordinary soil isn't a very good desiccant.

Possible Soil Moisture Meter

2010-08-08T17:03:00.000-07:00

I am conducting some experiments for a new way to measure soil moisture. The idea is to measure the humidity in a confined pocket of air directly over the layer of soil where the moisture is to be measured. There's a little fan secured in the bottom of the jar, along with an air humidity sensor, and the jar is face-down in the soil. The fan speeds up the process. There's a little filter paper over the mouth to keep the dirt out, but the inverted jar will not fill with water, due to the trapped air. In this implementation, the jar remains mostly above the soil, but it could be buried.

The larger fan is for drying the soil to test measurement retrace. The scale is a quick way to tell how much water is actually in the soil. Preliminary results look promising. The humidity runs on the high side; the only time I saw low humidity was when I started with baked soil.

M1.0 Solar Flare Detected

2010-08-07T19:00:00.000-07:00

There was a M1.0 solar flare today that I recorded with the Oddball SID Receiver (http://www.techlib.com/area_50/sidseizer.htm#Oddball)

The SID is visible at about 6:30 PM (UTC). The peaks at both ends are due to night time propagation and should be ignored. This demonstrates how nicely WWVB works for SID detection, at least at my location.