where expertise comes together - since 1996

Web's Largest Portal of Nondestructive Testing (NDT)
Open Access Database (Conference Proceedings, Articles, News), Exhibition, Forum, Network

All Forum Boards
Technical Discussions >
crosstalk about Phased array transducer
Career Discussions
Job Offers
Job Seeks
Classified Ads
About NDT.net
Articles & News

872 views
02:50 Oct-18-2003

Peter

Engineering
China,
Joined Oct 2003
3
crosstalk about Phased array transducer

I'm developing phased array transducers by use of piozoelectric ceramic. And a couple of samples have been finished. However there is very high crosstalk between adjacent elements, which is about -28dB. As you know, crosstalk is very essential in phased array transducers. Kindly sirs, could you like to provide some ideas tp solve my questions.

Thanks in advance!


 
06:14 Oct-18-2003
bill
Re: crosstalk about Phased array transducer ----------- Start Original Message -----------
: I'm developing phased array transducers by use of piozoelectric ceramic. And a couple of samples have been finished. However there is very high crosstalk between adjacent elements, which is about -28dB. As you know, crosstalk is very essential in phased array transducers. Kindly sirs, could you like to provide some ideas tp solve my questions.
: Thanks in advance!
------------ End Original Message ------------

I suggest you contact Auto UT Ltd., they seem to be the absolute experts in Automatic ultrasonics and all connected issues.



 
00:34 Oct-18-2003
Joerg Schulze-Clewing
Re: crosstalk about Phased array transducer Peter, one thing that is crucial in minimizing crosstalk is to make sure each element is matched in its characteristic impedance by the electronics and cable connected to it. This is easier said than done as there may be long coax cables involved. Maybe if you share more details about the connections and the receiver/pulse setup this forum could be of more help.

Then there is the backing material. After dicing, is it allowed to fill the gaps? Is it nicely degassed so there are no bubbles entrapped? Do you have enough backing behind the elements? Backing material and its acoustical properties make a big difference.

Then, the biggest culprit I found in my consulting work: Grounding. This has to be a nice conductive stretch. Bonding magnet wires to ground works in theory but it results in a not very good crosstalk performance.

Also important is the way you are dicing. If the array remains on a thin contact carrier before applying the backing material, did you dice through it?

Measuring crosstalk at frequencies above 10 MHz is not very easy. Unless you are an RF engineer it might be a good idea to have one around when doing that.

Regards, Joerg.

----------- Start Original Message -----------
: I'm developing phased array transducers by use of piozoelectric ceramic. And a couple of samples have been finished. However there is very high crosstalk between adjacent elements, which is about -28dB. As you know, crosstalk is very essential in phased array transducers. Kindly sirs, could you like to provide some ideas tp solve my questions.
: Thanks in advance!
------------ End Original Message ------------




 
02:51 Oct-19-2003

Peter

Engineering
China,
Joined Oct 2003
3
Re: crosstalk about Phased array transducer Hi,Joerg Schulze-Clewing,

Thanks a lot for your kindness. I thiink you are profession and experience. thank you very much again.
I will remember you in orderto study from you.
You let me share your four tips in all. i.e. characteristic impedance, backing fill and its properties, Grounding, and dicing way. About characteristic impedance, whether it means the characteristic impedance of cable, or 50 OHm, or 75 Ohm, determined by the system? I seldom do with it, maybe this results in relatively high crosstalk. As for backing fill, I ususually select flexible exopy and alumina powder as babacking materials with a impedance of about 5Mrayls and more than 12mm in thickness. But I use the mixture of flexible epoxy and micro-tiny glass ball to fill in the gap between elements, and I make sure no bubble entrapped. As for your third suggestion, Grounding, we bond magnet wire down the both sides of ceramic about 0.7mm in width. We also cut through the elements into the backing about 0.0005inch. Maybe I should do more labor in characteristic impedance. Another thing, It's also a big problem about heat and elevating temperature in phased array transducers. I would like to get your suggestion. I usually add an melt part such as aluminum to scatter heat. Looking forward to hearing from you. if possible, please write me email: liyongchuan@asisz.com. Would you like to share you email address eith me. Thank you very much
----------- Start Original Message -----------
: Peter, one thing that is crucial in minimizing crosstalk is to make sure each element is matched in its characteristic impedance by the electronics and cable connected to it. This is easier said than done as there may be long coax cables involved. Maybe if you share more details about the connections and the receiver/pulse setup this forum could be of more help.
: Then there is the backing material. After dicing, is it allowed to fill the gaps? Is it nicely degassed so there are no bubbles entrapped? Do you have enough backing behind the elements? Backing material and its acoustical properties make a big difference.
: Then, the biggest culprit I found in my consulting work: Grounding. This has to be a nice conductive stretch. Bonding magnet wires to ground works in theory but it results in a not very good crosstalk performance.
: Also important is the way you are dicing. If the array remains on a thin contact carrier before applying the backing material, did you dice through it?
: Measuring crosstalk at frequencies above 10 MHz is not very easy. Unless you are an RF engineer it might be a good idea to have one around when doing that.
: Regards, Joerg.
: : I'm developing phased array transducers by use of piozoelectric ceramic. And a couple of samples have been finished. However there is very high crosstalk between adjacent elements, which is about -28dB. As you know, crosstalk is very essential in phased array transducers. Kindly sirs, could you like to provide some ideas tp solve my questions.
: : Thanks in advance!
------------ End Original Message ------------




 
02:54 Oct-19-2003

peter

Engineering
China,
Joined Oct 2003
3
Re: crosstalk about Phased array transducer Thank Bill very much for your kindness. I will follow your direction. Thanks !
-
---------- Start Original Message -----------
: : I'm developing phased array transducers by use of piozoelectric ceramic. And a couple of samples have been finished. However there is very high crosstalk between adjacent elements, which is about -28dB. As you know, crosstalk is very essential in phased array transducers. Kindly sirs, could you like to provide some ideas tp solve my questions.
: : Thanks in advance!
: I suggest you contact Auto UT Ltd., they seem to be the absolute experts in Automatic ultrasonics and all connected issues.
------------ End Original Message ------------




 
02:29 Oct-20-2003
Joerg Schulze-Clewing
Re: crosstalk about Phased array transducer Hi Peter,

Although I am not sure which frequency you are using right now, in general the impedance of the transducer will be very different from the system and the coax. That requires matching at both sides but at least at the system side. There are three commonly used methods to figure that out. The first one would be to model it in SPICE but that is tricky because it is not easy at all to model the transducer element correctly. One reason for that is, for example, the acoustical characteristics of the material in front of the transducer, the backing etc.

Another method would be a network impedance analyzer, and the most complete method would be a true pulse echo test on a phantom (or water bath if you have to) with system, transducer, cables and all. Mostly you end up having to compensate a bit for an overly capacitive component.

As to heat I am a bit surprised. Usually transducers do not get hot unless you are doing a heavy duty Doppler scan. The only time I have seen one exceed a temperature too hot to the touch was in a lab setting but we were exceeding the FDA limits on acoustic output in that test. It could also be that you are using pulse trains (bursts) at a fundamental frequency that is too far off the transducer center frequency. That could cause heat dissipation.

Ground: I don't understand how your wire runs but if it goes from one end of the transducer to the other and only those two ends are grounded that would not work too well. You need really short paths to ground for each element, no more than a few (very few) millimeters.

A word of encouragement: 28 dB of crosstalk is pretty good for a first attempt in making an array. Anyway, I'll send you my email address. The reason I don't post it here is that every time I do that there is a surge in new junk email, probably from junk mail programs that automatically scan such forums. Good luck with your experiments.

Regards, Joerg.


 


© NDT.net - The Web's Largest Portal of Nondestructive Testing (NDT) ISSN 1435-4934

Open Access Database, |Conference Proceedings| |Articles| |News| |Exhibition| |Forum| |Professional Network|