By Fran Ridge & Jim Klotz
August 10, 2018

August 10, 2018

One of the things we have been needing in the field is a way of testing a device that might have developed a bad chip or software problem caused possibly by illegal shutdowns, power outages, and internet/router issues.

The earlier Node 100 device had made a direct correlation on May 12th so we knew it was working well and we were reluctant to pull it out of the field. But the chip would not allow automatic or remotely-pushed s/w upgrade and we opted to replace the entire device. Then the node had been "quiet" since the device was replaced earlier, so we wanted to make sure it was running properly.

Using a permanent magnet of undetermined strength a test was conducted by waving it very briefly approximately two feet away from the MADAR-III DataProbe. Nothing showed up on the 60 line per hour readout. But when the same magnet was waved in the same manner at one foot, results were obtained. In order to see an update one must hit the resubmit button, and the data came right up and was very surprising.

1) Node 100 uses an older "H" chip but the s/w version had been updated to 1.13RBU. The sensor immediately picked up the test "anomaly".

2) The device returned to normal "status" mode on its own as it was designed to do.

3) The AlertStart began at 00:30:21 and went to AlertEnd at 00:30:27, which means the test anomaly lasted only 6 seconds. We imagine that 5 of those seconds were the sensor settling back to normal.

4) As prescribed by the software the dataline plots went from normal (status) at one per minute to two per second, which is 120X's faster in the "alert" mode.

5) After the alert was over and AlertEnd was loogged, the data-rate of 120X's per second was maintained for six minutes but under "status" before dropping to the normal rate of once per minute. This is the way it is supposed to work, but I had thought the period was supposed to be 5 minutes. The original White Paper had suggested 15 mins but that was WAY too long at the 7200 lines per hour rate. I now think that 3 minutes is more than enough if the anomaly re-occurs because the device will go back into AlertStart anyway.

6) The milligaus reading had been "1" up until the anomaly when it was 15 just a second before it went to 72. I thought that was strange. When it went to status again just 6 seconds later it resumed the low number "0", which is close to the "1" previously.  The background or "ambient" reading varies between other devices and locations so there is no significance particularly to the number itself.

7) The "limits" setting for Node 100 at the time was set at 50. This is better than 60 but I have seen some devices set at much lower levels that have not been reporting anomalies. This doesn't mean they are not working. Anomalies should be somewhat rare and rarer in some areas than others. But the sure way of checking the devices is by waving the magnet as we did.
8) During this test the compass heading varied about 4 degrees. This suggests extreme sensitivity and would explain why MADAR-III is getting more "hits" than the older MADARs. Some ops have even reported minor earthquake activity.

What we learned from this simple test is that
1) Node 100 is working. It had been quiet since 5/21 and had been replaced in July. The sensativity was not determined in this test but further tests will be conducted. The device had been activated, went to alert mode at 120X's a second, returned to normal status, collected data for 6 minutes and resumed the normal data rate of once per minute.

2) An automatic alert email was sent to the operator as prescribed and a copy was sent to me to document a false target. Perfect. If we ever get the cell-phone alert text message we will have a real detector instead of a datalogger, but we have a major tool in MADAR right now that nobody else has or is even close to getting.

3) Of great importance to me was the fact that the device did not lock up as other devices had been doing with the older chips and out-dated software which couldn't be upgraded from a remote source. The only way to get them out of lock was to have the op reboot and for us to reset (raise) the limits to prevent another situation, always hoping that we did not hurt the sensitivity. That will be determined by the data from future anomalies and sightings of interest. We may find that 60 is not that bad.

Jim was unable to conduct a test with a hand-held compass at 2' and 1'. He couldn't locate a compass. I had a compass and found a magnet so I attempted to duplicate Jim;'s test, but without trying it on Node 6.

I tried a powerful but small donut magnet and put the magnet two foot away with the handheld compass pointing to the right on my desk, which means N was right and W was to the rear.. Nothing happened. But when I moved it to about 12", the needle moved about 3 degrees! And, of course, if I put the magnet on the right I had to move it much closer because the lines of force were parallel. This tells me that supermagnets have a closer tight field and, although the intensity normally drops off at the inverse cube of the distance, maybe even more so with concentrated SMs. In otherwords, this magnetic flux density is more powerful at the source and less powerful at distance, which explains why a 2' distance didn't show up in either test.

This test analysis report is subject to updates.