have you though about some thin aluminum foil or thin sheet of it to cover the holes.
Probably not a good idea to put conductive material right next to those leads, though.
finger nail polish can help with that. (Both sides of the wall)
I think I am going to stick with the black construction paper idea.
Think you could also uses shrink tubing, non conductive ( both sound and eletrical ) and you can get it evil looking “wet” black or verious colors. Also it lets you have a press in fit. Suspect you problem with the first position was the ping was being deflected by the nose section and not having a line of sight back to the reciver.
… (>>>>>>>>>>>>>>
---------------------- <
============) <
---------------------- <
___________( <
<
Sorry for the crude grafics. Can be fix by angling the eyesockets so the path of the ping can reflected back to the reciver. ( gads, need a good drafting program for these entrys 
)
Opps, graffics didn’t work out. Seems it doesn’t like extra spaces. Preview showed it ok. Anyway, if you angle the eye sockets inward about 15deg.s you should get a better line of sight for the pulse. Or come to think about it. you could angle the side of the nose so the can spread out on that side.
Thanks Userfriendly,
I moved the sensors forward and it seems to work ok. I would not recommend mounting sensors in any recessed areas since it does seem to pose a problem. I have played around with my idea a bit and sometimes it works and some times it does not work past 5" regardless of where it is pointing. If I move the wires around a bit, it seems to come back to life. I’m guessing the 12+ inches of wire could be causing problems. I need to cut the leads down to shorter lengths and try again.
if you use the code, /code buttons it will use a monospaced font. You can use notepad to creat the graphics. Then you can see the graphics better.
I have experience using the SRF-04 with extender tubes to narrow the pattern. I used 5/8" ID PVC pipe with a 1-5/8" total length. When I inserted the sensor into the tubing pair it didn’t work at all. I initially thought it was something electrical or a code problem and spent a long time trying this or that, but nothing would fix it. By accident I didn’t put it all the way back into the tube after checking some things and it started working. I then started experimenting with the distance the sensors were inserted into the tubing and found there was a sweet spot. It was 1-1/8" from the end of the tube. Now the application and the components are essentially the same, or close enough, for you to at least try different depths before you assume it the lenght of the wire etc. A difference of 1/8" can be all it takes. Here is the robot I was working on for the narrow beam ultrasonic sensors.
Good luck with it!
Thanks Jim,
I thought that I found my sweet spot before I epoxied them in but some how between when the epoxy was wet and when it cured, it caused the problem. It will work fine for a few seconds if I hold the wires a certain way but then goes back to only reading 5" max. I thought the same thing as you did where it might be the code, thinking the length of wires was throwing off the timing.
I will try to carefully break out the sensors and try to find another sweet spot. I think this time I will use rubber cement.
I have been re-designing another bot head to use a color camera. I really would like to navigate using cameras instead of PING sensors.
My older design caused problems with the PING sensor because there were not equal sides around the sensor to keep it working. Jim’s tube concept works, because there are sides of equal depth around the full circumference of the sensor, where my older design has one side open, and the other side shrouded by the nose bridge. I can get this to work if I glue a 1/8" tube the end of the sensors, but now my original design concept has changed cosmetically which I am not happy with.
Below is a screenshot of my new design to use a camera. This design will fully encase the entire camera where all that will show is the lens in the eye socket. The smaller eye socket is to put in an IR emitter for camera light. Cmos cameras have the ability to see IR, it just looks like white light.
looks pretty sweet mike, too bad you already designed the other thing and got it too and then found it wouldn’t work
Jim,
Somehow I missed your above post before…
That’s one serious-looking rover!
On my biped’s car/rover/vehicle-of-mayhem I’d been planning on using ultrasonic sensors, but hadn’t been sure how I could protect them without interfering with the signal.
Those covers look monster, and sound like they’d be perfect.
Thanks!
Mike,
I…want…now!
Any plans on making multiples of that one?
I’m sure that there’d be quite a few willing buyers who wouldn’t care that it’s just an untested prototype…
I could have a few made for those interested. Since I found a freelance machine guy, he will make them for a lot less.
Here is another view of my design. All I need is to buy the camera and get a caliper to get the proper dimensions for the lens and camera base.
That looks… AWESOME!!! it is very terminatorish I actually lik it better than the other 1. u will have to tell me how th ecameras work out becauseI hav been itching to play with them on my hex.
Here is a pic of the servo mounting bracket to be machined out of aluminum. This bracket mates the bot head to the servo horn. I just modled it today!
looks sweet mike 
Really professional. Im geussing this is your real job and not a hobby, correct???
No, just a hobby…
My real job deals with high performance electronic manufacturing at class III IPC-A-610D specifications. I currently program machines to test thes boards for defects. When I say programming, it is not code like IF-THEN-ELSE, rather it is a GUI application that has loads of settings and parameters for each component type. If I want to test a 1206 resistor at location x,y, I tell tha machine in the windows app. what part it is, and its value. When the machine runs the program, that the software generates, it goes to each location of the board and a servo driven probe(s) are applied to the locations of the landing specified, to look for a value specified, like 10k as an example. If the test returns a value of 1k, then the machine reports this data to the operator as a failure.
The machine we use is the APT-9401
Click link for more inforation on this machine:
texmac.com/takaya_apt_9401.html
so THATS why your so good as desiging boards