Author Topic: The very long array project.  (Read 6240 times)

0 Members and 1 Guest are viewing this topic.

Offline Rickard

  • Hero Member
  • *****
  • Joined: May 2009
  • location: Luleå Sweden
  • Posts: 500
  • Unit(s): 999, 981, M37, LowBird-1, LowBird-2
The very long array project.
« on: April 28, 2013, 02:26:39 PM »
Hi,
 
It's a little bit early to announce success, but my project with building a super-long sidescan array has at least reached the point when the transducer configuration is accepted by the unit.
 
The new array is made of four SI elements from two XHS 9 SI 160 T transducer. The elements were actually retrieved from one of my Doubleducers. The lenght of the array is 440 mm. The idea is to use the array at 262 kHz in a hull mounted application for long range, deep water scanning with high definition. I don't expect the result to be good at 455 kHz because the nearfield is so long with this combination of array length and frequency.
 
I still have to wait for the ice to vanish so I can't do testing on water. In the meantime I can present an image of a skinned transducer. This is what is left after a night in acetone. Removing the ABS housing makes it easier to cut out the elements.
 
Rickard


Offline Rüdiger

  • Sr. Member
  • ****
  • Joined: Feb 2011
  • location: Hjältevad Sweden
  • Posts: 427
  • He who seeks finds
  • Unit(s): 898 CX SI ;EAGLE 500CDF
  • Software: DrDepth BT+SS
  • Accessories: Towfish and Towcam
Re: The very long array project.
« Reply #1 on: April 29, 2013, 12:17:01 AM »
Hi Richard

I hope your unit survived your experimenting pleasure ;) I am really looking forward to the first pictures.

mvh
Rüdiger
Sorry
English is not my first language !
http://de.groups.yahoo.com/group/SOB-International/

Offline abraquelebout

  • volunteer rescue diver, surfer, publicist
  • Full Member
  • ***
  • Joined: Apr 2010
  • location: france
  • Posts: 241
  • Unit(s): 898, 1198, helix G2
  • Software: deep,,drdepht,reef.
  • Accessories: towfish lowbird /LSS2-hdsi, 60mcable
Re: The very long array project.
« Reply #2 on: April 29, 2013, 12:37:03 AM »
hi rickard
nice project, waiting for the first picture
for my part, i cut all lss2 cable except the left and right array and ground,
connect them to the cable, the pink and orange wire in lls2 sim's to be the positve
and the other, the green and blue négative, i make connect with 5 wire in lss2 and 3 in hdsi
verry good result at 800 and 455 khz
i have 2 hdsi array and wait crash other to make a verry long one
the acetone make a good result to take out the ABS, interresting, the other material is not strong
can you give the formulation to aclculate fréquences for the array lenght.
with 3 array at 455 / 800 what is the using fréquence ?

Offline Rickard

  • Hero Member
  • *****
  • Joined: May 2009
  • location: Luleå Sweden
  • Posts: 500
  • Unit(s): 999, 981, M37, LowBird-1, LowBird-2
Re: The very long array project.
« Reply #3 on: April 29, 2013, 12:28:22 PM »
Thanks Rüdiger,
 
Abra, I think the LSS-2 will prove to be the optimal configuration in the end. But it's interesting to test where the limit goes, therefore I was tempted to try with this four element experiment. There are several approximate expressions for calculating beam width. An easy one is presented by EdgeTech: http://www.edgetech.com/docs/app_note_beamwidth.pdf
 
A three element array (340 mm) can be operated at 262, 455 or 800 kHz with these specifications:
 
262 kHz: One-way beam width at -3dB - 0.95 degrees, Nearfield - 5 m.
455 kHz:     "           "        "          "   -  0.55 degrees, Nearfield - 9 m.
800 kHz:     "           "        "          "   -  0.31 degrees, Nearfield - 16 m.
 
As you see the nearfield is extended in proportion to frequency. A problem with a three element design is how to connect the elements, a parallel or serial configuration? None of them will match the impedance in the original transducer. This can be solved with four elements because you can combine serial and parallell circuits to keep the impedance at the same level as in one element. This is what I have done.
 
I tried with a test from the beach in water this evening, but conditions were awkward. Fast running stream, shallow and short distance. There was an image, but is was rather blurred which was expected so the test was not conclusive. I know the system works at a basic level, though.
 
Here is an image from my kitchen.
 
 
 
« Last Edit: April 29, 2013, 12:51:25 PM by Rickard »

Offline Gimp

  • Jr. Member
  • **
  • Joined: Apr 2013
  • location: Northern, WI
  • Posts: 74
  • Unit(s): 998 HD SI
  • Accessories: i-Pilot Link
Re: The very long array project.
« Reply #4 on: April 29, 2013, 12:36:39 PM »
Here is an image from the new array.
If that's an image From the new array, it is clearer than any image I've ever seen from a SI unit ;) The resolution on the place mats is incredible  ;D

Offline Rickard

  • Hero Member
  • *****
  • Joined: May 2009
  • location: Luleå Sweden
  • Posts: 500
  • Unit(s): 999, 981, M37, LowBird-1, LowBird-2
Re: The very long array project.
« Reply #5 on: May 01, 2013, 07:02:30 AM »
Hi,
 
I found an ice-free location where I could do some sector scanning from shore in pretty good conditions. The place is a 100 m wide and 9 m deep channel with smooth bottom. The test showed the array works well at 455 kHz, but there is an issue with noise and pulsating something at 262 kHz. It could have to do with too much element mass in the array, or maybe the combination of depth at the transducer, one foot, and the max range SI setting at 120 m (360 feet) puts too much stress on the units control algorithms. Further testings on water will tell more. The pulses show like spokes that radiate at even intervals in the image. But I can say for sure there are no bugs like failing elements or mixed up pooling directions. In fact, I'm very surprised and pleased with performance at 455 kHz.
 
As expected, the image is quite blurred within the first 20 m from the transducer. Sector scanning at 120 m means the transducer must be swept very slowly because ping rate is just 6 per second. It's difficult to keep a slow constant speed. The images are from the same spot and show 90 degree sectors at 120 m and 60 m at 262 and 455 kHz. The 455 kHz images are darker with more detail.
 
 
« Last Edit: May 01, 2013, 07:09:19 AM by Rickard »

Offline Klems

  • Newbie
  • *
  • Joined: Jan 2013
  • location: San Diego, CA
  • Posts: 5
Re: The very long array project.
« Reply #6 on: May 02, 2013, 04:36:42 PM »
Great work and research on tis extra long transducer Rickard. I love seeing your experiments. It's great to see your progress in this area for the rest of us tinkerers at home. Can't wait until the ice thaws in your area.

Offline Rickard

  • Hero Member
  • *****
  • Joined: May 2009
  • location: Luleå Sweden
  • Posts: 500
  • Unit(s): 999, 981, M37, LowBird-1, LowBird-2
Re: The very long array project.
« Reply #7 on: May 05, 2013, 08:38:37 AM »
Issues,
 
I noticed water had penetrated between the aluminum u-bar and the epoxy. This doesn't seem to affect the electrical properties, but the epoxy prism isn't firmly fixed by the aluminum bar and seems to have a tendency to be squeezed out in the middle. Thermal expansion is not the same in epoxy and aluminum and may have caused the epoxy prism to get out of position. So I pressed the epoxy prism out from the u-bar to inspect the situation. It was obvious my epoxy (standard laminating epoxy for marine use) does not stick well to aluminum. The epoxy prism seems strong, but is rather easy to bend (with risk for breaking an element). I have to find another housing for the array. Perhaps a wider aluminum u-bar and some sort of flexible and water resistant fill. Or maybe no fill at all and just a few supports that makes a 'floating' mount?
 
Rickard

Offline Rickard

  • Hero Member
  • *****
  • Joined: May 2009
  • location: Luleå Sweden
  • Posts: 500
  • Unit(s): 999, 981, M37, LowBird-1, LowBird-2
Re: The very long array project.
« Reply #8 on: May 07, 2013, 12:05:02 PM »
More issues,
 
The image looked less bright last time out and I found out the third element doesn't work. I believe one soldering has failed. It was probably damaged when the epoxy prism was forced out from the aluminum bar. But the fourth element does respond even though it's connected in serie with the failing element - strange.

Added: The third element seems ok after all. It was bypassed by a short circuit that is fixed now. The noise issue persists and the bar has become slightly bent so it's convex. A convex array means the main beam is widened. Any deviation from straight should be in the opposite direction. I think the noise is caused by the lack of shielding of wires and elements. In some way I will add a copper shield to the array. This shield should be isolated from water and connected to ground like in the standard HB transducers. The idea with making all connections outside the epoxy prism also means there are long unshielded wires which will pick up lots of noise.
 
Added: Noise is eliminated completly just by connecting the aluminum bar to ground. The bar need not be isolated from water. There is a new issue with weak signal after some minutes of use. This is probably caused by water ingress between the bar and the epoxy prism. It could also be water ingress in the cable.
 
Added: Testing on water showed the signal is weak although no water had penetrated into the array. No noise. Perhaps the aluminum bar and any shield must be isolated from water, after all?
 
July 23, added: At least one element is broken. I will take a time-out on this project and decide how to proceed with this later.

Rickard
« Last Edit: July 24, 2013, 12:48:12 AM by Rickard »


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo

 

Related Topics

  Subject / Started by Replies Last post
151 Replies
65564 Views
Last post August 07, 2013, 02:42:22 PM
by Sea-Rover
5 Replies
4034 Views
Last post January 06, 2010, 04:04:26 PM
by bob
0 Replies
1970 Views
Last post June 11, 2011, 06:09:23 PM
by edwardpic
2 Replies
2594 Views
Last post August 23, 2012, 03:51:01 PM
by skipt
31 Replies
18364 Views
Last post January 10, 2018, 07:19:29 AM
by ropesfish
2 Replies
1289 Views
Last post January 12, 2014, 10:49:10 PM
by slabbacks
1 Replies
888 Views
Last post June 27, 2014, 09:20:45 AM
by hays47
12 Replies
4004 Views
Last post May 18, 2016, 08:46:36 AM
by sonarbear


anything
SimplePortal 2.3.3 © 2008-2010, SimplePortal