Matching impedances to ultrasonic transducers

[Kevin Gallagher]

Hello,
I'm building an ultrasonic tape measure and am having trouble obtaining a distance greater than 1 metre. I have an idea to try and get a greater distance: I think that maybe I am losing a lot of my transmitting power because the output impedance of my oscillator is not matched well enough to ultrasonic transducer and similarily at the reciever.

The oscillator is the 4047BE and in the datasheet there is no talk of output impedance. Does anybody know what it is or how i could find it. The o/p of osc is fed directly to transducer which is SCS-401T. At the reciever the rx transducer (SCM-401R) is fed directly through a capacitor and then to a 10k resistor which is the input resistor for the amplification stage.

On the datasheet for the transducer it says the impedance is approx 500 ohms for txer and 30 kohms for the rxer. Is this the same value for i/p and o/p impedance cos it just says impedance.
I would really appreciate any help anyone could offer me.
Thanks

 

[Roff]

Well, from the limited info that I could find, it looks like you should be driving it with about 50v p-p in order to get maximum output. Since the input impedance is 500 ohms, it's going to take a beefy driver.

 

[Optikon]

Hello,
I'm building an ultrasonic tape measure and am having trouble obtaining a distance greater than 1 metre. I have an idea to try and get a greater distance: I think that maybe I am losing a lot of my transmitting power because the output impedance of my oscillator is not matched well enough to ultrasonic transducer and similarily at the reciever.

The oscillator is the 4047BE and in the datasheet there is no talk of output impedance. Does anybody know what it is or how i could find it. The o/p of osc is fed directly to transducer which is SCS-401T. At the reciever the rx transducer (SCM-401R) is fed directly through a capacitor and then to a 10k resistor which is the input resistor for the amplification stage.

On the datasheet for the transducer it says the impedance is approx 500 ohms for txer and 30 kohms for the rxer. Is this the same value for i/p and o/p impedance cos it just says impedance.
I would really appreciate any help anyone could offer me.
Thanks

The 4047 bloack diagram shows a buffered output so you can assume it is very low, maybe less than 1 Ohm.

To match, determine what your Tx input impedance is, and then put the same value of resistor in series with the output of the 4047. If your line length is very short (from OSC to Tx) this will work OK. If your line length is longer (how much depends on rise time of signal) then you will have to build a transmission line interfacing the two.

Keep in mind, this method will match your OSc-out to Tx-in but will be attenuated by 50% (the matching resistor forms a divider) but it should be undistorted at the receiving end (and hence your will not be losing power output due to mismatch anymore) You may want to buffer the OSC with something that can give you higher volts so that you can keep the Tx power out the same.

 

[Roff]

Hello,
I'm building an ultrasonic tape measure and am having trouble obtaining a distance greater than 1 metre. I have an idea to try and get a greater distance: I think that maybe I am losing a lot of my transmitting power because the output impedance of my oscillator is not matched well enough to ultrasonic transducer and similarily at the reciever.

The oscillator is the 4047BE and in the datasheet there is no talk of output impedance. Does anybody know what it is or how i could find it. The o/p of osc is fed directly to transducer which is SCS-401T. At the reciever the rx transducer (SCM-401R) is fed directly through a capacitor and then to a 10k resistor which is the input resistor for the amplification stage.

On the datasheet for the transducer it says the impedance is approx 500 ohms for txer and 30 kohms for the rxer. Is this the same value for i/p and o/p impedance cos it just says impedance.
I would really appreciate any help anyone could offer me.
Thanks

The 4047 bloack diagram shows a buffered output so you can assume it is very low, maybe less than 1 ohm.

To match, determine what your Tx input impedance is, and then put the same value of resistor in series with the output of the 4047. If your line length is very short (from OSC to Tx) this will work OK. If your line length is longer (how much depends on rise time of signal) then you will have to build a transmission line interfacing the two.

Keep in mind, this method will match your OSc-out to Tx-in but will be attenuated by 50% (the matching resistor forms a divider) but it should be undistorted at the receiving end (and hence your will not be losing power output due to mismatch anymore) You may want to buffer the OSC with something that can give you higher volts so that you can keep the Tx power out the same.

According to figs 4 and 6 in TI's 4047 datasheet, the typical pulldown resistance is about 140 ohms, and the typical pullup resistance is about 400 ohms, and this is with a 15 volt supply.
Impedance matching doesn't deliver maximum power to the load, given a fixed load impedance. Zero driving source impedance will deliver the most power to the load, and also provide the best damping of resonances. Look at modern day solid-state audio amplifiers. They use feedback to minimize output impedance. Also, see https://www.electro-tech-online.com/custompdfs/2004/01/impmatch.pdf fo a discussion of impedance matching.

 

[jem]

Don't have direct experience with the Sanwa transducers; however, I have used similar ones. These transducers are highly capacitive, and require low impedance drivers. However, the real portion of the Tx is quite high, and actual power is quite low. Here is what I would suggest.

1. Tx - Drive the transducer diferentially. Use a CD4049BE or 4050BE. Parallel three of the gates together to get one input and one output. Drive one end of the Tx with the OP. Parallel the remaining three gates, and the common output drives the other end of the Tx transducer. Operate the chip at 10V to get 20V, PkPk which is what these types of transducers can take continuously. To drive the 4049 (or 4050) single ended, use a simple NPN inverter with about 2K collector resistor to drive one of the inputs. The other input is driven directly. This creates the out of phase drive needed.

2. Rx - Use a 27K resistor across the Rx transducer, and a fairly high Zin amp (OPAmp, or discrete). Gain has to be about 1000 to 2000. The 27K resistor provide enough damping to the Rx transducer to prevent all sorts of resonances which might confuse your echo detection circuit.

Also, ideally you need a time varying gain on the Rx. The longer the echo takes to return (longer distance to target), the fainter it will be. Thus the need for time-varying gain. A cheap way of achieving something similar is to have high gain, but use a clamp at the output of the Rx amp so that strong echoes get clamped. You will also probably need a blanking window following the Tx pulse so that you do not detect direct (from the Tx) pulses as return echoes.

Hope this helps,

Jem

 

[jem]

Forgot to mention that you also need coupling capacitors (0.22u would be fine) as the transducers do not like to have DC across them.

Jem