Hi Rick,
The parabolic reflector (Squirrel Baffle) that I used has a f/D ratio
somewhere between 0.34 and 0.39 based on how I determine the focal point
location. I believe this ratio is a pretty reasonable value based on
everything I have read. The reflector is pretty shallow with a depth of
approximately 3.6 inches and the diameter measures 19.75 inches.
I looked at many squirrel baffles and could not find any that were anywhere
close to being a true parabolic curve besides the one that I used which is
also what Charles N0TT used and recommended. In reality we have no idea
how much the squirrel baffle surface deviates from a true parabolic curve
but based on how it generates a well defined focal point it's acting
reasonable (and our crude graphical overlays looked reasonable). At some
point I might generate a parabolic curve using Auto-Cad and print it out
full size to compare with the Squirrel Baffle I'm using but that's low on
my priority list at this point in time.
*The total beam angle of the ultrasonic receiver element (microphone) I
used is as follows:*
-6dB total beam angle = 85 degrees (listed on the datasheet)
-12dB total beam angle = 150 degrees (approximated from the datasheet plot)
If I do a little trig based on the placement of my microphone, it looks
like the full angle relative to the outer perimeter of the reflector is
approximately137 degrees which is within the -12 dB window but outside of
the - 6 dB window.
So as not to put someone on the spot I won't mention names but someone told
me the following which makes a lot of sense:
*"Oh, and on the pattern, you may find that the mic won't pick up much at
the fringes of the dish. That's a good thing since it increases its F/B
ratio when mounted on the dish....acts as a shield from US (ultrasound)
noise arriving behind the dish." * Therefore having the outer edges of the
dish not fully utilized sounds like a desirable feature.
I've also been told the Radar Engineers unit uses a Log amp which might be
another reason for its stellar performance, but I can tell you I was amazed
when I went to my test power pole with my unit that it was instantly
obvious what device was generating the arc and I had plenty of signal and
very narrow beamwidth.
Also reading the RFI archives I see a long time RFI investigator for a
utility company said that based on his many years on the job that
ultrasound only worked about 50% of the time so we should not be surprised
that there are cases where ultrasound does not provide fruitful information
(and he was using Radar Engineers equipment).
The Radar Engineers reflector diameter is 18" compared with the 19.75"
diameter of the Squirrel Baffle I'm using but I don't know the depth of the
Radar Engineers reflector. Charles (N0TT) and/or Jeff (W4DD) should be
able to tell us the depth of the RE250 as they both have one, then we can
compare the f/D ratio.
While the radar engineers unit is the gold standard (I have used a Radar
Engineers model 250 that belongs to one of our local utility companies) we
should not be disappointed when a home built unit costing $200 does not
compare equally with a $4000 unit (list price was $3650 for the Radar
Engineers model 251 in 2021) but it's a good goal.
Just FYI, and thanks for the thought provoking comments.
73.
Don (wd8dsb)
On Wed, Mar 25, 2026 at 8:31 PM Richard Karlquist <richard@karlquist.com>
wrote:
> I investigated building a homebrew detector like this some 20 years ago or
> so, and found that the ultrasonic transducers that I could find had a
> fairly narrow angle of reception. The only parabolic reflectors I could
> find were very "deep" and it turned out that those transducers would poorly
> utilize the dish. We would say they wouldn't "illuminate" much of the
> dish. It was obvious from a catalog page showing the Radar Engineers
> product, that it used a much "shallower" dish, which would give much better
> "illumination". It is not sufficient to say that the dish is accurately
> parabolic and then you just locate the focal point and put the transducer
> at the FP.
>
> I just wanted to give everyone a reality check. My suboptimal device did
> "work", but probably not as well as the Radar Engineer's one, and was much
> larger to boot. If I had to do it now, I would make a 3D printed shallow
> dish and as much as possible clone the Radar Engineer's dish. I would say
> that the dish in the photo looks much shallower than the one I tried to
> use, but it is hard to say just from the photo. The key figure of merit
> for a dish is the ratio of diameter to focal length. My dish was such that
> the focal point was in the plane of the rim of the dish.
>
> Another thing is that the transducers are very narrow band in terms of
> frequency, probably because they are optimized for TV remote controls, I
> guess. Since we are looking at wideband noise, I suppose it doesn't matter
> a lot, except we lose a lot of sensitivity that way.
> ---
> Rick Karlquist
> N6RK
>
>
> On 2026-03-25 13:42, Don Kirk via RFI wrote:
>
> I finally decided it was time to add the one missing tool I had from my RFI
> detection/direction finding toolbox so I undertook the building of my own
> ultrasonic parabolic dish receiver system for detecting arcing on power
> line hardware based on the W1TRC design that was in the April 2006 issue of
> QST. Unfortunately Jim (W1TRC is an SK).
>
> I know this tool has been discussed many times over the years on the RFI
> reflector but thought I would share my recent build information with the
> group as it might be helpful to those currently wanting to build one, and I
> also think I kept my build very simple.
>
> I had a lot of assistance from Charles (N0TT) and I also picked up a lot of
> great information from Jeff (W4DD) and Frank (K7SFN).
>
> Here is the link to my simple website that I created today for those
> interested in my build:
> https://sites.google.com/view/ultrasoundarcdetector/home
>
> 73,
> Don (wd8dsb)
> _______________________________________________
> RFI mailing list
> RFI@contesting.com
> http://lists.contesting.com/mailman/listinfo/rfi
>
>
_______________________________________________
RFI mailing list
RFI@contesting.com
http://lists.contesting.com/mailman/listinfo/rfi
|