Hi,

is there a theorical maximal lentgh of an ultrasound probe cable (btw
beamformer and transducer), due to signal loss ?
Why are these cables so short ?
Is it possible to lengthen them by reamplification systems ? (Please
give an example of existing product)

Been a few years since I worked on ultrasound. Is this medical or
crack detection etc? You say beamforming so I?m assuming medical.
Presumably also phased-array rather than the older rotating transducer
type. Whilst we were all keen to examine our internals with the
abdominal head, no one ever volunteered to test the anal probe.

Anyway, usually there will be electronics in the head, especially at
higher frequencies. The signal may be relayed by simple line drivers
in which case there is a limit to the length of the cable otherwise
the fine detail will be degraded. Re-amplification would not restore.
If it incorporates matched impedance drivers, there is no good reason
for short length other than too long a cable gets in the way and is
annoying to store. So much depends upon the actual electronics. If I
were designing one now, I?d probably go for a high speed serial
interface similar to firewire or USB so the cable could be thin and
flexible. This may be how the latest generation are made.

If the head is either completely or largely passive, the cable
characteristics will be factored into the phase requirements and will
not be amenable to lengthening.

So basically, the answer hinges upon what electronics are present in
the head, and the driver signal. Do you have any more info?

Best

It is medical ultrasound phased-array used with a 5MHz linear probe.
Your answer really helped me to understand what the problem is.
Do you have an idea of how much the signal can be degraded in a such
passive head, in function of the cable length ? (lower, greater than
10dB per m for instance ? )
With a bibliographic or scientific reference, this answer will be good for me :)

This is a wee bit tricky with my out of date experience. Perhaps
someone else can add to the comments.

In the meantime though, I had a look at some head pictures and the
cables were much thinner than previously. Simple heads with 64 or 128
transducers *could* be raw analogue parallel but the cable would be
quite thick. High speed serial hasn?t been around for too long but the
more complex 2D (and 3D) must do a lot of processing in the head.

You might look up ? "phased array" medical ultrasound ? which came up
with this and thousands of others :-

http://www.ifi.uio.no/~ftp/publications/pub/preprints/SHolm-3.pdf

Other searches such as ? "phased array" medical ultrasound "transducer
configuration" ? brought up lots of good stuff too. It?ll take a lot
of sifting though.

What I would really like to know is what is in your head :-) ideally
the semiconductors. Any idea on the number of elements, number of
wires (which will be similar to the plug if it?s not hard wired).
Thickness of cable and is head one axis or two? If you can open the
cable connector how many wires are co-ax?

If the interface is analogue with an impedance matched cable drive,
the attenuation even at double the length will be negligible but there
may be problems with phase distortion and loss of fine detail. Just
depends on what is going up and down the cable. The gain control (or
AGC) could compensate for attenuation and if there is edge
enhancement, the fine detail could be compensated too. If it?s
digitised parallel, you?d probably be OK with double the length, like
printer cables, though because of the higher speed it might screw it
up. Depends again on whether it?s a matched impedance and what sort of
encoding. Of course if you are lengthening the cable you?d probably
have to be careful to use similar type which, will be difficult as it
won?t be off-the-shelf.

Do you have the make and specs so I can check out the maker?s pages,
or a link to the maker? The more info you can give, the better.

Must admit I?m curious (and interested) in how your requirements. Did
you acquire the machine and you want to use it for other purposes?

Best