The Coins-in-a-bag thread has been moving from anecdote to anecdote for the past week or so. Let's take this to the next level---sloppy field testing!
This morning, I procured two fresh, new 1 gallon size Baggies brand plastic
food
storage bags, a single Baggies brand sandwich storage bag, and $3.06 (6 cents,
7
nickels, 4 dimes and 9 quarters) in clean US currency.
I cut a whip antenna to 25 cm length (banana/BNC adapter included), yielding a
1/4
wave antenna for roughly 300 MHz. I then used this whip as a monopole antenna
connected first to my Tektronix TDS 640A oscilloscope (1 GHz real-time BW) and
then
to my HP-8562A spectrum analyzer.
I placed the coins into one of the 1 gallon bags, and inflated it with a
swooping
motion through the air (my breath has high humidity). I held the bag only by
the
neck, trapping the air and causing the bag walls to bulge, approximating a 10"
diameter sphere. I then vigorously shook the bag, at about 12" in front of the
whip
antenna, while observing the displays.
++++++++++++++++++++++++++
TEST RESULTS
Oscilloscope, 5uS/div, 1 V/div
Test 1: 2.46 Vpeak, with an exponential decay lasting over 45 microseconds and
a TC
of about 25 microseconds. The risetime is about 136 nanoseconds. A view of the
acquired waveform and an expanded risetime view is attached as Coins.zip.
Test 2: 2.64 Vpeak, similar decays. This has an interesting double discharge at
the
leading edge. An initial discharge triggers the scope, then another discharge
hits
about 500 nanoseconds later. A view of the acquired waveform and an expanded
risetime view is attached as Coins.zip.
Note: It only takes a few shakes to get a decently high discharge.
And now, on to the spectrum analyzer. Similar test distance and conditions.
Using a
1 MHz resolution bandwidth, a 1 MHz video bandwidth, 0 Hz span width, 0 dB of
RF
attenuation and peak hold trace.
Test 3: 10 MHz
31 dBuV noise floor
59 dBuV transient noise peaks
Test 4: 100 MHz
31 dBuV noise floor
62 dBuV transient noise peaks
Test 5: 500 MHz
31 dBuV noise floor
70 dBuV transient noise peaks
Test 6: 1 GHz
31 dBuV noise floor
66 dBuV transient noise peaks
Test 7: 5 GHz
31 dBuV noise floor
50 dBuV transient noise peaks
Test 8: 10 GHz
35 dBuV noise floor
40 dBuV transient noise peaks
Test 9: 15 GHz
43 dBuV noise floor
47 dBuV transient noise peaks
Test 10: 500 MHz
31 dBuV noise floor
70 dBuV transient noise peaks
Note: I removed all coins except the 6 cents.
Test 11: 500 MHz
31 dBuV noise floor
70 dBuV transient noise peaks
Note: I removed all coins except the 7 nickels.
Test 12: 500 MHz
31 dBuV noise floor
70 dBuV transient noise peaks
Note: I removed all coins except the 9 quarters.
Test 13: 500 MHz
31 dBuV noise floor
70 dBuV transient noise peaks
Note: All coins back in the bag.
Test 14: 500 MHz
31 dBuV noise floor
61 dBuV transient noise peaks
Note: All coins back in the small sandwich size bag. (Maybe less
vigorous
shaking.)
Test 15: 1 GHz
31 dBuV noise floor
53 dBuV transient noise peaks
Note: All coins back into the gallon bag. I exhaled into the bag
several
times, causing visible condensation on coins and bag walls. The observed
transients
were far fewer numerically, and of lower magnitude.
What this seems to show is:
1. There are measurable emissions present across the spectrum from 10 MHz to 15
GHz.
2. The emissions clearly stand out from the ambient noise level of an open
industrial area.
3. It is easy to create several volts into a high impedance load.
4. The generated levels do not depend on having a mix of different coins.
5. One type of coin works about as well as any other.
6. High humidity within the bag inhibits noise generation.
7. It's time for lunch.
Regards,
Ed
--------------------------
Ed Price
[email protected]
Electromagnetic Compatibility Lab
Cubic Defense Systems
San Diego, CA. USA
619-505-2780
List-Post: [email protected]
Date: 03/18/1999
Time: 14:59:35
--------------------------
<<attachment: Coins.zip>>
