HomeRoast Digest


Topic: Changed to: Ultrasonic Crack Detector (38 msgs / 816 lines)
1) From: Michael Wade
This is a multi-part message in MIME format.
Brett:
When I first heard about the HG/DB, I thought it was a joke.  (Get outta =
here!)  I've been immersed in high tech much too long, and tend to =
forget about "appropriate technology".  As I've read more about it, I've =
come to appreciate the possibilities.  I just might try it soon.
In the meantime I keep thinking about a gadget I used way back when I =
was a Fire Service mechanic, an air leak detector that down-converted =
the ultrasonic sound generated by an air leak into the audible range.  =
The thing looked like a flashlight with a microphone where the lens =
would be and a headphone jack in the other end.  It didn't amplify =
audible sound at all, but a tiny air leak sounded like a jet engine in =
the headphones.  You could use it on a running diesel and find a leak =
you couldn't hear even with the truck shut down.
The question, of course, is whether 1st and 2nd crack generate =
significant ultrasonic energy so that the detector could downconvert and =
amplify it.  A commercial one costs almost as much as a Hot Top, but =
wouldn't it be cool to amplify only the sound of the crack!
I know there are some real power technogeeks and gearheads on this list. =
 I stand in awe of some of the threads I've followed.  How 'bout it?  =
Anybody know of an inexpensive practical way to do this? 
Michael Wade

2) From: Aaron
Ultrasonic crack detector?
A plumbers tool eh?
Does this detect plumbers crack?
Oh there are SOOoooo many directions I don't dare take this topic :)
It might amplify it but given the snap is probably going to be more 
closer to an impulse type noise it might attenuate it.  It'd  be hard to 
narrow on one frequency because well, the beans hitting the sides of the 
glass or steel deflector.. for example.. is another type of 
impact/impulse noise.
Given the amount of noise that is going on in the whole bean process, id 
think it would probably take some significant processing to isolate out 
one or two components there.
aaron

3) From: Steve Hay
On 4/21/06, Aaron  wrote:
<Snip>
Maybe, but I think it would be worth a try to see what's going on up there
in that frequency band..
--
Steven Hay
hay.steve -AT- gmail.com
Barry Paradox: Consider k to be the greatest element of the set of natural
numbers whose description require maximum of 50 words: "(k+1) is a natural
number which requires more than 50 words to describe it."

4) From: Michael Wade
I knew, the instant I posted this subject, that I was making a terrible 
mistake.
Oh well, at least it wasn't as bad (so far) as the "Suri with the fringe on 
top..."
The unit that I used a few times years ago seemed pretty simple, just 
down-converting broad-band ultrasonic sound, say 20 - 100 Khz, down to the 
audible range.  Maybe just a 10:1 ratio?  I have no idea how you could 
perform the conversion.
Michael Wade
<Snip>

5) From: Dan Bollinger
It is quite possible that it would work.  Cracks are due to the sound of 
expanding gases causing cellular ruptures. Escaping gases often emit 
ultrasonics.  You'd want to borrow a unit and try it.
Ultrasonic detectors are more commonly used to detect vacuum leaks. The old 
soapy water trick doesn't work on vacuum.
Dan
<Snip>

6) From: Steve Hay
On 4/21/06, Michael Wade  wrote:
<Snip>
e
<Snip>
It is fairly simple, actually.  Its the same technology used for AM signal
demodulation, and you essentially "tune-in" the band you want to listen to.=
.
--
Steven Hay
hay.steve -AT- gmail.com
Barry Paradox: Consider k to be the greatest element of the set of natural
numbers whose description require maximum of 50 words: "(k+1) is a natural
number which requires more than 50 words to describe it."

7) From: J.W.Bullfrog
Dan Bollinger wrote:
<Snip>
Yeah it does, it just sucks! ;-)
<Snip>

8) From: Michael Wade
I Googled ultrasonic microphone as a starting point and lo and behold I 
found a wealth of information on ultrasonic detection and frequency 
conversion under the subject of "Bat Detectors".  Yes, the little furry 
winged creatures we used to chase under the street lights on summer nights.http://www.jwaller.co.uk/batgroup/bat_detectors.aspIt's a british site, wouldn't you know.  Apparently they take bat detection 
very seriously.
Now who will be the first with "batty" jokes?  Aaron?  Are you there?
Michael Wade

9) From: Ken Mary
Build a bat detector?http://home.netcom.com/~t-rex/BatDetector.htmlhttp://njsas.org/projects/bat_detector/index.html
I just googled these links, it seems easy enough.
--

10) From: Tom Bellhouse

11) From: Aaron
I believe you can find a reliable bat detector from the Gotham City 
Public Works Department.  Might want to try the division of lamps and 
illumination.
and hanging upside down helps get the caffeine to my brain to work 
better so there!!  either that or every so often I just have to and take 
a sabBATical and rest a few days.
Aaron

12) From: raymanowen
" ...try the division of lamps and illumination." Bright Idea!
Steve has a good idea about a "superheterodyne" sound detector. It can
demodulate  a carrier, whatever the frequency. The modulation is just
spikes- not much without a real carrier.
If you have only spikes and you can't hear spikes, use them to excite a
tuned resonant "tank,"  mix with a BFO like a CW detector. An ultrasonic
whistle would be easier to mix with a BFO.
Something ought to work.
Cheers -RayO, aka Opa!
First and second DING!

13) From: Michael Wade
Ken, thanks for the reply.  I had fouind Tony Messina's site and was 
considering ordering a circuit board to build one of his frequency division 
SBD's since I'm only listening for clicks and pops anyway, but on the other 
URL you sent I found a link to a $20 "ultrasonic translator" that is a 
tunable heterodyne unit.http://xtronics.com/kits/SK-207.htmI still thiink that a wide-band scanning frequency division unit would be 
simplest to test with, but I couldn't resist the $20 price for a fully 
assembled unit.
I've ordered one and will post my results when I have some...
Michael Wade

14) From: Michael Wade
This is a multi-part message in MIME format.
Ray:
I found and ordered a heterodyne detector: =http://xtronics.com/kits/SK-207.htm(I wondered what that bright light was...)
Michael Wade

15) From: Michael Wade
Oops, that'll teach me to read the f(ine) print...  It's a kit, but still a 
great price.
mw

16) From: raymanowen
"Comes with its own speaker, illustrated instructions, attractive case,
special ultrasonic piezoelectric transducer, digital CMOS IC (integrated
circuit), Operational amplifier IC and all other parts. This is a favorite
kit."
It's really not a complex kit, Michael.
Oh, how I wish the Heathkit Co. of Benton Harbor, MI were still in business=
.
Back in the day, I'd send off for a kit with my paper route money. The
Heathkit box would arrive in a couple of weeks while I was in school, but m=
y
folks wouldn't let me in on it until I got home Friday afternoon.
It was like Christmas in August, or whenever. Those were the days.
Set up an assembly station and keep stuff in order. Better keep kids and th=
e
cat away, too.
Cheers -RayO, aka Opa!
--
"When the theme hits the bass, I dance the Jig!" - -Virgil Fox at the Might=
y
Wichita WurliTzer- 1976

17) From: Brett Mason
Ultrasonic Crack Detector....
"Ultrasonic" sounds fast....
I never knew a plumber could move that fast.
They were always slow to my house....
Brett
On 4/25/06, raymanowen  wrote:
<Snip>
e
<Snip>
ey.
<Snip>
--
Regards,
Brett Mason
HomeRoast
   Zassman

18) From: DJ Garcia
Aaahh! The old Heathkits! I did a few, mostly test instruments. They
were really well thought out. Also had a few Knight Kits from Allied
Radio, who made you figure out the resistors from the color code - Heath
had the value printed on the packing card that held the resistors :-).

19) From: Jim Wheeler
Heathkit was the Sweet Maria's of electronics kits.  My wife and I
built a number of their kits and learned a great deal from them.  The
tech support people were super to deal with.  We have a couple of
their home-brew monitor repair videos that are still interesting.
Good people are always to be cherished.  Too bad the purchasers of
Heathkit the company didn't understand that.
Sitting with a cup of home roast coffee and listening to the water
fowl on our pond this evening.
--
Jim in Skull Valley

20) From: Angelo
Same for Dynakits ...Yearning for that time when quality meant 
something more than a just word in a slogan...
A+
<Snip>

21) From: Michael Wade
This is a multi-part message in MIME format.
Oh, how I wish the Heathkit Co. of Benton Harbor, MI were still in =
business. Back in the day, I'd send off for a kit with my paper route =
money. The Heathkit box would arrive in a couple of weeks while I was in =
school, but my folks wouldn't let me in on it until I got home Friday =
afternoon. 
It was like Christmas in August, or whenever. Those were the days. 
Set up an assembly station and keep stuff in order. Better keep kids and =
the cat away, too.
Thanks, Ray,
The kids are long gone but that's good advice about the cat!
My dad was a 2-way radio tech for the State of Wyoming and had a small =
repair shop in a shed in our back yard.  He built almost all of his test =
equipment from Heath kits.  Oscilloscope, frequency generator, battery =
eliminator, multimeter, etc, etc.  I'm sure he could have assembled them =
in 1/4 the time if not for letting me "help".  I turned out to be more =
mechanically than electronically inclined but to this day when I smell =
hot rosin core solder I get a flash of memory of him.  
Michael

22) From: Michael Wade
This is a multi-part message in MIME format.
It refers to the speed with which they present the bill...
mw

23) From: raymanowen
" ...I get a flash of memory of him."
How neat, Michael- How Absolutely Neat!- ro

24) From: Aaron
mw... remember,  that bill for giving you the privelage of witnessing a 
plumbers crack first hand.
is only for the pipes.  it doesn't include repairing the holes in the 
wall.... the holes in the floor... the section of ceiling they cut 
out... cleaning the carpet... etc etc...
that's some expensive crack there...
Aaron

25) From: kofi
On 4/20/06, Michael Wade  wrote:
<Snip>
ing
<Snip>
ld
<Snip>
e
<Snip>
t
<Snip>
 A
<Snip>
 to
<Snip>
Way back in 1997 I ordered a kit that did just that, converted ultrasonic t=
o
regular audio using some cmos.
It never worked that well but the theory is sound (sic)
Better still would be using FFT to discriminate and tell you when the crack=
s
are occurring.
Not brain surgery but not that simple either.
So who's going to record a roast and doo FFT on the recording and make it
happen?
I eagerley await your results.
Kofi

26) From: Scott Marquardt
I'll do the recording. That little radiant roaster of mine? The cracks (bot=
h
first and second) are LOUD, and I can get a decent microphone very near to
the beans.
Can anyone recommend a good FFT utility? I could use any of a handful of
conventional sound editors and plot a spectrum, I guess, but I don't think
that would work all that well on just the envelope of an individual snap,
would it? Or would it?
I'll do some manipulations and see what comes of it. This'll be fun!
- Scott
On 4/26/06, kofi  wrote:
<Snip>
he
<Snip>
thing
<Snip>
a
<Snip>
l,
<Snip>
ould
<Snip>
the
<Snip>
d
<Snip>
--
Scott

27) From: Justin Marquez
On 4/26/06, kofi  wrote:
<Snip>
he
<Snip>
thing
<Snip>
a
<Snip>
l,
<Snip>
ould
<Snip>
the
<Snip>
d
<Snip>
Or, maybe you could just listen to the beans as you roast them....
Safe Journeys and Sweet Music
Justin Marquez (Snyder, TX)

28) From: Aaron
scott it might work but a few issues.
if you are going to be playing with the ultrasonic scale...  you need a 
microphone that is sensitive enough to pick it up.  the equipment that 
is sensitive enough to work it, and 'filter' out all the other noise.
now your sound editing software. youll need it sensitive to work in the 
ultrasonic range too.  just because the 'sampling rate' might be 48Khz 
doesn't mean the program itself is set up or will reliably work with 
actual frequencies near that... not to mention the quality degradation 
because you are approaching the 1:1 ratio on the freq vs sampled 
freq.... as you get less 'samples per second' to work with, the software 
will have to start 'filling in' the rest of the data with what it thinks 
should be there by gauging on what was before and after the sampled 
interval...
most sound editing software is going to cut out much anything about 22 
ish k anyways because it thinks it's white noise, hiss, static, 
harmonics, beat freqs etc etc.and human hearing don't go much above that 
anyways.
In other words, lets say I told you to make change for a dollar for 
folks BUT you can NOT use more than 5 coins per dollar to do so......  
you are automatically going to 'dump' the pennies and nickles and dimes  
from the pile of coins you have. as it is impossible to do the change 
using them with your limitations... kind of the same principle here.
you are definately working on the right path though in your idea, just 
need a bit better equipment.  and a 'snap' you could technicaly 
'capture' and analyze it.  it would be pretty much a noise burst or an 
impulse noise, kind of like a firecracker or gunfire
aaron

29) From: Scott Marquardt
OOOoooh, I forgot we were doing ULTRA. Dang. Well, I'll do something with
the normal range anyway, just for fun. Yeah, I'd get aliasing and all that.
Dang.
Hanning window, I presume? 1024 point?
On 4/26/06, Aaron  wrote:
<Snip>
--
Scott

30) From: Aaron
if not hanning a gesthalt??  and id sample a good 30 seconds of it as it 
comes into the crack stage.  that way you have a lot of 'samples' to 
pick out one really nice 'pop'  or a few of them to do an like RMS 
average of the 'typical crack' if you want...  since computers are way 
faster now than the were in 1996, and gigs and gigs of both ram and HDD 
are cheap, you can store a 30 meg file and work with it... without 
having to wait for a pot of coffee to brew between tweaks to the sample..
I was involved in a project once in the navy, we were writing software 
to pick up 'unusual noises' in machinery... many years ago... again  
back around 1996 heh imagine that...  anyways with this....  you might 
want to check the fan motor on this thing.... what is the rpm of the 
motor.  filter out that freq with a user patch.... how many blades on 
the fan itself... mult fan speed times blade number to filter out that 
harmonic.... is it gear driven, if so how many gear teeth, once you got 
all that .. you  might want to go 3 or 4 orders down from those too or 
until a certain natural dB attenuation on order harmonics.  Does the 
thing have a 'support element' that the airflow will go through,  this 
'interruption' in the air flow can cause 'wave harmonics'  so.. if it 
does,  how many 'vanes' does that have....    Now.... since the crack is 
going to be an instantaneous noise pretty much... as would the 'tink' of 
a bean hitting the glass for example or the metal deflector... are you 
going to want the software to be able to filter out a glass 'ting' from 
a metal 'bing' from a been crack... see my point...I would say a source 
triggering O scope or a spectrum analyzer might help you see this 'big 
picture' of your bean cracks and stuff but i better shut up now... once 
again I am geeking this wayyyyy too much.
Great project you have going there but it can too easily drag you way 
into the nether regions of sanity.... kind of like this whole 'coffee 
roasting' thing has already done, given we are taking about such stuff 
as this :)
Aaron

31) From: M. McCandless
I've been thinking that a notch filter might be the answer.
Determine the freq of the crack sound & filter out everything else . . .
then amplify.
McSparky
At 01:50 PM 4/26/2006 -0700, you wrote:
<Snip>

32) From: Brett Mason
So I am thinking...
Or - switch to a Plan B:
  Switch to a skillet roast where there is no sound but the gas flame and
the snaps...  This loses all Geekiness but results in ZENness...  I just di=
d
2 lb of Colombia Mesa De Los Santos in my skillet for our trip to Vancouver
tomorrow (Son is graduating from Trinity Western University on Saturday!).
Heard every crack, and was totally engaged in the process.  (You have to be
or you get black eyed beans...)
Brett
On 4/26/06, M. McCandless  wrote:
<Snip>
ing
<Snip>
ld
<Snip>
e
<Snip>
t
<Snip>
 A
<Snip>
 to
<Snip>
--
Regards,
Brett Mason
HomeRoast
   Zassman

33) From: Angelo
Justin, you take all the fun out of it  :-)
<Snip>

34) From: Scott Marquardt
Roger the averaging, though I'd have to do that post-FFT for each pop, whic=
h
is a bit of a pain.
No blower, and it's pretty quiet (the radiant thing).
I can read the freq distribution for the ambience between cracks and
subtract that from the snap's analysis.
- Scott
On 4/26/06, Aaron  wrote:
<Snip>
--
Scott

35) From: Michael Wade
This is a multi-part message in MIME format.
Holy HotTop, Batman!   Look where these guys have gone with this =
Ultrasonic thing...
Justin, this whole thing started because yours truly has a high freq =
hearing loss courtesy of too many hours on the flight line during =
run-ups.  I bought an iRoast2 and can't hear the cracks over the blower =
noise, so I started looking for a way to detect them electronically. 
If this doesn't work out, I'm going Luddite with an HG/DB setup...
Michael Wade

36) From: Spencer Thomas
On 4/21/06, Dan Bollinger  wrote:
<Snip>
I don't think I'd want that plumber in my house. :-)

37) From: Dan Bollinger
<Snip>
<Snip>
Then you'll be interested in this: http://www.crackseal.com/

38) From: Justin Marquez
On 4/27/06, Michael Wade  wrote:
<Snip>
.
<Snip>
Well, now, then it DOES make some sense!  I guess I missed the first part o=
f
the discussion.
Safe Journeys and Sweet Music
Justin Marquez (Snyder, TX)


HomeRoast Digest