Ultrasonic Range Meter
roots_of_culture
I am trying to build an ultrasonic range meter using an AU5550
(http://www.jaycar.co.nz/products_uploaded/AU5550.pdf) - 40 khz
transducer. i am using the pic16f877a to send the 40 khz pulse to the
transducer, 12v supply.
-- I realise that i need to use P/N channel hexfet power mosfet totem
pole, but i am stuck at which driver to use and how.
-- Other Driver Ideas....Circuits...
Any and all sorts of help appreciated.
Nav2u
LED => Transducer...the non visible end of transducer is grounded via a
resistor...
roots_of....
Stef Mientki
In my opinion 5V is enough,
take a look here
http://oase.uci.kun.nl/~mientki/pic/projects/rapid_prototyping/rapid_prototyping_us_ranger.html
Stef Mientki
Nav2u
roots_of_culture
>In my opinion 5V is enough,
>take a look here
Im looking for a range of about 10cm- 6m, and i didnt think that a 5V
supply would be enough to drive the transducer hard enough to get the
range i desire.
And specifically 12 v supply, because i am using a 12 v car battery.
And i was also thinking that i will need a voltage regulator after the
12V battery, which will then leave me with a voltage of around 10.5 V.
i would like to use a driver circuit because the transducer is highly
capacitive and i didnt think i would be able to get sharp rise and
falls in the pulses i wish to transmit otherwise.
In the schematic below i have used schimdt trigger chip. But i want to
use a driver chip, something apart from the schimdt trigger.
(http://navdeep2u.googlepages.com/transmitting.JPG)
Unknown
wrote:
Try these links.
http://www.interq.or.jp/japan/se-inoue/e_pic6_6.htm
http://www.robot-electronics.co.uk/htm/srf1.shtml
http://www.ele.uva.es/~jesus/sonar/sonar.html
http://www.restena.lu/convict/Jeunes/beacon/Sender.htm
http://www.hexamite.com/he240tr.htm
http://www.ee.latrobe.edu.au/~djc/UltraSonics/Ultrasonics.htm
http://www.rdrop.com/~cary/html/ultrasonic.html
Andrew.
Chris
Hi, Roots O' Culture. Your transducer can be seen as a capacitive
load. You're basically trying to do this (view in fixed font or M$
Notepad):
|
| .---------------.
| | |
| | | 5V-to-12V
| | | Level Translation/
| | | High-Current source/sink
| | | Buffer
| | | (Non-Inverting) |.--.| (Inverting)
| | | |\ || || /|
| | PIC o----o-----| >-----|| ||-----O< |---.
| | | | |/ || || \| |
| | | | |'--'| |
| | | | Transducer |
| | | | (1800pF) |
| | | '-------------------------------'
| | |
| | |
| | |
| '---------------'
|
(created by AACircuit v1.28.6 beta 04/19/05 www.tech-chat.de)
Since you're already going with a PIC, Microchip has a pair of drivers
that will do exactly what you want -- the TC4420 (non-inverting) and
the TC4429 (inverting) high speed MOSFET drivers. They have the
capability to drive up to 6 amps with low Rds(on) MOSFETs built in, and
the IC takes care of the level translation for you. Put one on each
side of the transducer, and you're good to go. They're made to drive
capacitive loads, and have crossover protection built in. They'll
smack that T/R40-16B around real good (140 volts max!?! Is that an
aluminum, steel, or titanium case?).
Unless this is a high volume application, I don't think you're going to
be able to do better than this. The ICs are available in 8-pin DIPs,
too (CPA suffix).
Here's the datasheet:
http://ww1.microchip.com/downloads/en/devicedoc/21419c.pdf
If you want to be really safe, add a 2.2 ohm non-inductive resistor (a
1 watt carbon comp is good here) in series with the transducer to limit
maximum current doesn't exceed 6A.
|
| .---------------.
| | |
| | | 5V-to-12V
| | | Level Translation/
| | | High-Current
| | | Buffer
| | | (Non-Inverting) |.--.| (Inverting)
| | | |\ || || ___ /|
| | PIC o----o-----| >-----|| ||--|___|----O< |---.
| | | | |/ || || 2.2 ohm \| |
| | | | |'--'| |
| | | | Transducer |
| | | | (1800pF) |
| | | '-------------------------------------'
| | |
| | |
| | |
| '---------------'
|
(created by AACircuit v1.28.6 beta 04/19/05 www.tech-chat.de)
On the power supply end, the TC4420 and -29 are good to 18V, so you
shouldn't need another regulator. You also want to maximize voltage.
I'd recommend using another 2.2 ohm series resistor or choke and a
large (2200uF or greater, 25WV) cap with a Transzorb for a localized
charge source and to protect against load dump.
|
|Battery ___ To XDCR Power
|B+o----|___|--o---o->
| 2.2 ohm | |
| +| /-/
| 2200uF --- ^ P6KE16A
| 25WV --- |
| | |
| | |
| === ===
| GND GND
(created by AACircuit v1.28.6 beta 04/19/05 www.tech-chat.de)
Good luck
Chris
Unknown
Chris,
Great diagrams and advice. I've been trying myself to create a simple
but "as long a range as I can get" ultrasonic remote control using the
same transducers and 16F628A PICs. The transmitter uses an AND gate to
switch a 100V N-mosfet with the transducer, 1000uH inductor and a
100ohm resistor. At 9.5V, 60% duty cycle sending two bytes over and
over at 300bps it does work up to approx 11 meters from the receiver.
I'm working again on a bandpass filter for the receiver and once I
figure that out I'll need help hopefully from the members of this
group to get an automatic gain cirucuit built. The last time I fiddled
with bandpass filters I cut the range down to two centimeters! Not
much of a remote then!
The initial link that roots_of_culture provided for the transducer PDF
is a Jaycar branch in New Zealand and assuming that he too lives here
then acquiring the drivers may be a problem for both of us.
For roots_of_culture - one company, South Island Components, is
trade/retail has only the TC4429CPA but doesn't stock the 4420 while
other retail outlets like Jaycar, Dick Smith and Altronics don't stock
them or any of the range at all. RS components deals with the trade
and requires a company account to be opened but they do have the:
TC4421CPA - 9A Single Inverting
TC4426CPA - 1.5A Dual Inverting
TC4427CPA - 1.5A Dual Non-inverting
TC4428CPA - 1.5A Combined Single Inverting and Non-invereting
TC4429CPA - 6A Single Inverting
(Note - RS sells the TC4429CPA for NZ$10.70 + GST eachwhile South
Island Components sells them for NZ$5.48 + GST)
Other than trying a few more companies I've found online in the Yellow
pages there's Farnell in Australia I suppose. Should my assumption be
correct and you do find a source for the TC4420 in NZ could you please
post it in this NG as I'd greatly appreciate that.
Regards
Andrew.
roots_of_culture
Thanks, that has solved my transmitting circuit.
Andrew,
you are right TC4420 is not available at hand locally, but Farnell (
$4.40) delivers within 3 days, so that is good.
Though in order to get going im using the TC4428CPA, it has a low
Rds(on) 40 ns.
As for the recieving circuit, i saw a design from national in the
Electronic Design (jan 2006) journal. it uses a 2 stage LMP7711 (17MHz)
as active filters (band pass), with a gain of 100 * 10, but these are
not available locally either. I am trying to find substitutes.
Unknown
wrote:
My receiver is a POS affair without automatic gain so I've set the
transmitter PIC to be able to PWM at two different rates and as yet
the bandpass filter eludes me (rank novice). Need to do more research
when I'm fresh. I'm using the Dual LM833N 8pin DIP (Min 10MHz, Max
15MHz), high slew rate. I'll have a look at the LMP7711 too.
Regards,
Andrew.
Unknown
First off it was rude of me not to thank you in my last post for the
Farnell source confirmation - Thanks.
Secondly where did you get that Electronic Design Jan 2006 publication
from? I've looked on the net, found its publisher (Penton
Publications) and searched the CHCH city libraries to no avail. The
Electronic Design website is only a front page that constantly reloads
itself. The Europe one wasn't helpfull either.
Regards,
Andrew.
Mebart
Chris
HI, Andrew. The sun never sets on Farnell (I'm in the States -- it's
Newark here). Unfortunately, the Farnell website is being updated.
But looking at the Newark site, it seems they have the same issue.
You may want to try the TC4426 (dual inverting) and TC4427 (dual
non-inverting). You'll have to leave half of each chip dormant, unless
you have something else to do with them. To my recollection, you can't
parallel outputs safely.
The datasheets for the TC4426CPA (8-pin DIP, commercial temp range) and
TC4427CPA are available at the Microchip website:
http://ww1.microchip.com/downloads/en/devicedoc/21421d.pdf
The datasheet says they're rated for capacitive loads, and after all,
this isn't continuous use. You're only snapping the transducer a
couple dozen times, eight to twelve times a second. You really
shouldn't need a series resistor with the transducer unless you're
really cautious, especially considering the Rds(on) is typically 7 to 8
ohms. But by rounding the edges of the transition with the additional
Rds(on), you're losing some power.
http://ww1.microchip.com/downloads/en/devicedoc/21422d.pdf
Good luck
Chris
roots_of_culture
was away... Found the journal in the Manukau Library- looked in
national website, below is the link
http://www.national.com/signalpath/files/sp_designer104.pdf
hope it helps
Unknown
Jonathan Kirwan
wrote:
><snip>
>-- Other Driver Ideas....Circuits...
>Any and all sorts of help appreciated.
I only today noticed that Radio Shack is selling a completed
ultrasonic ranger as part of the VEX robotics stuff they sell.
Everything is "half off" until June, I'm told. The ranger is normally
$29.99 but is now selling for $14.99. It uses 5V supply and 3cm to
300cm ranging with 40kHz. Paperwork says it uses a 10us pulse; but
with a 25us single cycle width I doubt this spec. I might believe
100us, though, or 4 cycles. (Their minimum distance of 3cm
corresponds to a trip time of 175us.) Comes with screws for mounting
and a nice enough enclosure.
Just to let you know...
Jon
Jonathan Kirwan
>I'm doing a project on ultrasonic ranging too and the desired distance
>is 50m. It seems rather far and I wonder if TC4420/4429 could help? And
>can you explain how these things work cos I'm really confused about how
>6A would help driving my transducer. :-/
Hi, Martin. I'm no expert on this stuff, but am generally interested.
But it seems to me that 50m really is a far distance. Two problems
pop into mind -- (1) energy spreads out (ignoring dissipation of
energy at 40kHz in air) at D^2 and returns with another D^2 factor, so
this suggests to me a D^4 hurdle to overcome and that would mean a lot
of initial energy (read as very high driving voltage) to get there;
and (2) reasonable resolution over that kind of range seems vaguely
problematic, too. I also don't know what the attenuation of the usual
40kHz is in (wet) air or if there are better or worse bands nearby you
might prefer. I wonder if there is a page documenting it over
frequency...
Have you found any existing ultrasonic ranger that works at 50m?
Jon
Ban
> Hi Jon,
> I think 50m is way too far too actually. There's this paper on
> ultrasonic ceramic transducers on
> www.senscomp.com/specs/piezo%20application%20note.pdf and it explains
> all the losses and such.
> For a 100V pp driver with a transducer's SPL of 108dB and the
> receiver's sensitivity of -75db, according to my calculations, the
> voltage at the receiver at 50m is about 0.2uV :)) With the help of a
> cone I think this could get up a bit but still... Do you think I can
> do anything with this?
Can I say my 2 cents:
atmospheric absorption at 40kHz is a whopping 150dB/100m (20C 70%H) at nasty
10kHz it would be only 18dB. So for larger distances go down with your
frequency.
Also your calculation seems off. 100m means -80dB (inverse square law). Then
there is absorption/diffusion at the reflector, let that be another 20dB.
This makes it -100dB. Your receiver needs 94-75= +19dB SPL, you have only
108 - 250 = -142dB available. Forget it.
You can concentrate your beam with an old satelite dish (not the offset
ones) by mounting the transceiver in there. This would maybe improve the
ratio by 30dB. But then any temperature gradients or air movement would
carry your beam away.
Figures are from "Sound System Engineering" by Davis&Davis, ISBN
0-240-80305-1
--
ciao Ban
Apricale, Italy
Ban
> Thanks Ban. I want to point out that you were thinking 100m instead of
> 50m. :D I'm using that application note from senscomp (link in my
> previous post) and the calculations seem right. SPL reduction at 50m
> is 20*log(30/5000) ~ -45dB.
> There's one thing that was confusing though. According to senscomp's
> AN and http://www.earthworksaudio.com/tech/hf_sound.pdf air
> absorption is about 15dB for 50m. However, according to numerous
> atmospheric absorption (based on ISO 9613-1) calculators, this value
> should be about 1.2 - 1.3dB/m (about the same with your value
> 150dB/100m) for a 40kHz signal at 20 degrees C and 70%.
> I suppose ISO 9613 must be right. So 50m seems a bridge too far here.
That guy is an idiot, for 1kHz it is 0.7dB/100m, it is also not the frequncy
sqared, but has two plateaus, where the nitrogen(0.6dB/100m) and
oxygen(300dB/100m) levels are, which are then constant for all higher freq
in this gas alone.
Rich Grise
> Hey Chris,
> I'm doing a project on ultrasonic ranging too and the desired distance
> is 50m. It seems rather far and I wonder if TC4420/4429 could help? And
> can you explain how these things work cos I'm really confused about how
> 6A would help driving my transducer. :-/
> Thanks
> Martin
I wouldn't even bother to try ultrasonics beyond about maybe 3M (10 feet),
for a whole bunch of reasons. But, laser rangefinders for maybe 10' and
farther are available off-the-shelf:
http://www.google.com/search?q=laser+rangefinder
So I'd look at the minimum range on the lasers, and do the ultrasonics
for short-range sensors. ;-) (also pretty much off-the-shelf.)
Who's Chris?
Good Luck!
Rich
Jonathan Kirwan
wrote:
>On Thu, 06 Apr 2006 08:20:09 -0700, Martin wrote:
>
>> Hey Chris,
>> I'm doing a project on ultrasonic ranging too and the desired distance
>> is 50m. It seems rather far and I wonder if TC4420/4429 could help? And
>> can you explain how these things work cos I'm really confused about how
>> 6A would help driving my transducer. :-/
>> Thanks
>> Martin
>
>I wouldn't even bother to try ultrasonics beyond about maybe 3M (10 feet),
>for a whole bunch of reasons.
><snip>
Just as a note: Radio Shack has an off-the-shelf ultrasonic unit that
claims 3cm to 300cm (3m) range. Cost is under US$15 until some time
in June (it's on sale -- half price.) It's a unit for the Vex robot
kit thing.
Jon