Advice for using/hacking ST16 for Fathom ROV
- Thread starter Biggs Darklighter
- Start date
Depending on what you are using the steady grip for this may not matter for you, but the steady grip uses a different communication protocol to the camera than the typhoon H. There is a PPM like protocol that steadygrip uses that can only tilt the camera. To use both the tilt and pan functions you have to use the serial mavlink messages.
Wow really?! In fact, yes I do! The source for the android app is right here: Fathom App Github
If you were able to get the Fathom App working on a stock ST16, I would be eternally thankful!
I would modify the throttle stick inside for ratchet action instead of centered. Should be easy enough to do. The app is setup this way.
Last edited:
Yeah speaking of connectors... More poking and prodding has determined that the topside access point/router isn't powering up at all. No power is getting to it. Closer inspection found damaged/broken wires on the tether plug. I am trying to reverse engineer it and sort what goes where. HOPEFULLY I will get it powered up and connected shortly!
i have received a api scripting reference very similar to the one used in the ST series. and obfuscated ST10+ source. I was able to de-obfuscate and after a few weeks got a basic understanding. the fine young gentleman who i received this from is not exactly in the right ti get me proprietary material, so i dont thin he wants it to get out in the wild. lol that being said, im on vacation right now, i get back the 21st. shoot me a message then
ok, i get back from vacation on the 21st. when i get back ill see what i can do
The root directory of the ST16 has a lot of *.rc files, which give an overall view of the boot sequence and used tty's.
Access there is simple. Developer options -> USB debugging thick selected.
adb shell su gives a root prompt.
Access there is simple. Developer options -> USB debugging thick selected.
adb shell su gives a root prompt.
Sounds awesome, thank you! Cool, I feel better that you are not tackling it now actually. This gives me time to sort out the connection problem. If I have to open the Fathom up and dig it out, then just changing the electronics out might be the easiest path to take.
LOL sometimes it feels that way... I paid a lot of money for the privilege of rescuing this from going into the trash bin too! Actually, I suspect it may already have been IN the trash bin... I am glad I did though because it isn't easy to find these. I can't wait to get it in the water! From the videos, these look to be some of the most agile ROV's out there. But, I have to repair it first and that is proving DIFFICULT...
I managed to clean most of the corrosion off the AP. But I am fairly certain it is fried. Close inspection of the tether plug under magnification "probably" got most of the wires back into the right spots. Except the power wire.
THE POWER WIRE WAS NEVER SOLDERED!!!! NOT EVEN TINNED!!!
It was twisted, stuck in the spot, and heat shrunk. And three open pins on the plug, none of those tinned either. I "THINK" I have the rest of the wires sorted for the actual ethernet connection. I am going to try and figure out if I can connect to the ROV directly by plugging that cable into my laptop. I have an address to try, but I suspect the AP being removed changes that? Networking is not my strong point.
I am trying to avoid opening ROV itself up. That would potentially make this easy. But it is also buoyancy foam potted/sealed, and possibly glued too during assembly.
Whenever I plug the tether in, the drone powers up. So I suspect that 2 of those 3 open spots are jumpered on the ROV end of the tether. If this is true, then the third spot is the power wire. Obviously, I can also power the AP from the topside. And leaving that wire off for now should be safer anyways.
I think for the moment my strategy is to ohm out the tether cable. That might tell me something at least... I can verify my theory on the "power switch" and see how the lines run through the main tether. If it is wired up like an ethernet cable, with numbers matching the norm...
Last edited:
h-elsner
Premium Pilot
- Joined
- Mar 23, 2016
- Messages
- 2,310
- Reaction score
- 2,043
- Location
- Bavaria / Germany
- Website
- h-elsner.mooo.com
Modern network interfaces (1000Base-T or Auto_MDI-X) can be connected directly. It detects by themselves what is Rx and what Tx.
Older network interfaces need crossover cables if no switch or hub is used.
Wiring pictures here: Medium-dependent interface - Wikipedia
Older network interfaces need crossover cables if no switch or hub is used.
Wiring pictures here: Medium-dependent interface - Wikipedia
Okay, progress! Ohming our the cable conformed my theory. What an elegant solution! Instead of a switch through the hull, using the cable plug itself as a switch. So if you look closely, there is a tiny jumper sticking out of the black plug. The cable seems to be wired straight through 1-1, 2-2, etc. This pattern is reinforced by 9 and 10 on the ROV side being the switch (I checked the cable FIRST before randomly stuffing jumpers into the ROV lol). The AP side had those, and a third spot, empty. The third pin was bent, like somebody may have tried to crimp it. I thought this might be the power wire initially too. I had two other wires come off. But I think I have them in the right spots. One broke off with the heat shrink intact, the other left half the heat shrink on the plug, the other half on the wire.
I am assuming that this AP and the connection on the raspberry pi would fit the modern wiring/connection scheme. The tether appears to go straight through (verified quickly). BUT that doesn't really tell me about the wiring between the network cable end, and the plug for the tether. I don't have a way to easily ohm that, and it is a bit of a jumble to trace. Still, I think I put the puzzle pieces together.
I think I can find something at Home Depot, like an outlet plug end or similar, that will give me a breakout to make testing that last connector cable simple. But not critical, if the puzzle works.
Speaking of the AP, I unsoldered the network plug and am cleaning that up now. There was more corrosion underneath the plug. It is probably a lost cause, but still worth a shot. I am going to finish fixing that and try it first. If no-go, I will power up my Toughbook and plug in directly bypassing the AP, and see what happens...
No luck. Still no lights from the AP. I found a video here: Fathom App 101 (facebook) Towards the end they give a manual method to access the ROV with an address of: 192.168.254.10/manager I followed the network settings described in the OpenROV manual over here: OpenROV Software I don't know networking well enough to know if I have simply done something wrong? Or if not being connected to the AP and just plugging the "ethernet cable" from the ROV directly into my laptop, changes that address...
EDIT: It looks like that address shouldn't change, and I should be able to connect directly...
I really want to verify communications with the drone directly... When I plug that cable in directly with the ROV on though, nothing happens. Can somebody tell me if I am doing something obviously wrong?
I am getting closer to just opening this up...
EDIT: I opened it! The nose cap isn't glued on or anything crazy. But the camera/light board are connected and not a lot of room to work with. The camera cable should have been about 2" longer. A plug for the lights would have been nice.
This wasn't meant to be user "serviceable" though. As I suspected, the whole body cavity is foam filled. I can't get to anything inside to inspect or investigate further. YET! I am at a good stopping point for now. I need to figure out exactly what type of foam they used (I saw it somewhere), and then figure out if I can safely melt it out with acetone/gasoline or something.
Besides, I don't know if there is an actual issue INSIDE this yet, or if it is just me not knowing what I am doing...
EDIT: It looks like that address shouldn't change, and I should be able to connect directly...
I really want to verify communications with the drone directly... When I plug that cable in directly with the ROV on though, nothing happens. Can somebody tell me if I am doing something obviously wrong?
I am getting closer to just opening this up...
EDIT: I opened it! The nose cap isn't glued on or anything crazy. But the camera/light board are connected and not a lot of room to work with. The camera cable should have been about 2" longer. A plug for the lights would have been nice.
This wasn't meant to be user "serviceable" though. As I suspected, the whole body cavity is foam filled. I can't get to anything inside to inspect or investigate further. YET! I am at a good stopping point for now. I need to figure out exactly what type of foam they used (I saw it somewhere), and then figure out if I can safely melt it out with acetone/gasoline or something.
Besides, I don't know if there is an actual issue INSIDE this yet, or if it is just me not knowing what I am doing...
Last edited:
CRAP I so can't win... They used "624/625 potting foam". Google doesn't know what this is exactly, but come up with a 624 foam and a 625 foam. The 624 doesn't sound right, but the 625 I am sure is the correct stuff. It is used to make buoys by the Coast Guard...
So yeah it is a polyurethane foam and the only way to remove it once set, is mechanically. That won't go well I am sure...
So yeah it is a polyurethane foam and the only way to remove it once set, is mechanically. That won't go well I am sure...
So reading and re-reading the assembly instructions, the surprising primary function of the foam, is waterproofing the drone body. Bonus, it also adds a LOT of strength to the structure.
No measurable progress, but I found an interesting tidbit of information. I was able to verify the Pi version in the Fathom. It is a Raspberry Pi 3-Model B. This is the same/only Pi that ArduSub runs on. I can access the SD card easily with the nose off... Not sure it is as simple as, or even safe, to just swap cards though. But of course the card doesn't want to just pull out. I broke the edge trying to remove it. Some of that poly foam glued it in. Hopefully the card isn't damaged...
With the nose off, I can see the Pi lights. A red one that stays on, and a green one that goes on and off.
Alright guys I am really stuck. I need a strategy to verify the internal and external ethernet connections and circuits on the ROV. I would like to access 192.168.254.10/manager This would at least tell me that I can get a connection. If I plug the cable in directly to either a Mac or PC, shouldn't it at least realize I plugged a cable in???
Would a tool like this allow me to test and maybe ping the raspberry?
Here is the deal, I need to verify the tether and connectors. I can unplug and ohm the tether (did so quickly, it seemed all good), I can also just test the tether to AP plug adapter. But if there is no failures there, then the problem is inside the ROV and at best I will have a nightmare getting it dug out. More likely, I will destroy the pi or robot shield, or both...
PS: The File Repository does contain schematics. I figured out how to open them online at least. Does somebody have a way to convert those to pdf files I can print out and study them??? I tried a couple online converters and got gibberish.
No measurable progress, but I found an interesting tidbit of information. I was able to verify the Pi version in the Fathom. It is a Raspberry Pi 3-Model B. This is the same/only Pi that ArduSub runs on. I can access the SD card easily with the nose off... Not sure it is as simple as, or even safe, to just swap cards though. But of course the card doesn't want to just pull out. I broke the edge trying to remove it. Some of that poly foam glued it in. Hopefully the card isn't damaged...
With the nose off, I can see the Pi lights. A red one that stays on, and a green one that goes on and off.
Alright guys I am really stuck. I need a strategy to verify the internal and external ethernet connections and circuits on the ROV. I would like to access 192.168.254.10/manager This would at least tell me that I can get a connection. If I plug the cable in directly to either a Mac or PC, shouldn't it at least realize I plugged a cable in???
Would a tool like this allow me to test and maybe ping the raspberry?
NF-8601S
If I understand how it works (not really) it looks like I can at least plug the far end in, and trace a break in a single wire at a time. If this can help me determine if the cable between the tether and the AP plug is good... Sure, I could just ohm this, but it is going to be a fiddly PAIN to get a thin enough probe on the network plug side. If this will let me verify that, and test back to the Pi (even though I can't unplug or access the cable there) then I will grab one.Here is the deal, I need to verify the tether and connectors. I can unplug and ohm the tether (did so quickly, it seemed all good), I can also just test the tether to AP plug adapter. But if there is no failures there, then the problem is inside the ROV and at best I will have a nightmare getting it dug out. More likely, I will destroy the pi or robot shield, or both...
PS: The File Repository does contain schematics. I figured out how to open them online at least. Does somebody have a way to convert those to pdf files I can print out and study them??? I tried a couple online converters and got gibberish.
Last edited:
Thank you that would be super helpful! There were a couple in the electronics folder of the file repository I linked above. I will go dig up direct links tonight when on my computer.
The new AP arrived yesterday but I was too tired to try it out. I’m going to give it a shot tonight. If I can get online, SWEET! If not, I’m ordering that networking tester above. Probably still going to order it anyways, as it looks like it will let me test wires without access to the other end. Along with a bunch of other useful functions.
I did finally manage to get the SD card out of the Pi after about half an hour of careful wiggling side to side. Can’t see anything on it, but looks like I need special software to see the Pi file format. It seems like it is a Linux district, so I will try one of my Linux boxes later I want to backup the OS on it. Hopefully the contents are still intact.
I plan on grabbing another Pi. Then I can power it up with the Fathom card, monitor, mouse, and keyboard. Go exploring. Not sure what I will learn, but I want to compare the Fathom OS with the ArduSub one.
Tonight is the big night. Hopefully it all just powers up and works!
The new AP arrived yesterday but I was too tired to try it out. I’m going to give it a shot tonight. If I can get online, SWEET! If not, I’m ordering that networking tester above. Probably still going to order it anyways, as it looks like it will let me test wires without access to the other end. Along with a bunch of other useful functions.
I did finally manage to get the SD card out of the Pi after about half an hour of careful wiggling side to side. Can’t see anything on it, but looks like I need special software to see the Pi file format. It seems like it is a Linux district, so I will try one of my Linux boxes later I want to backup the OS on it. Hopefully the contents are still intact.
I plan on grabbing another Pi. Then I can power it up with the Fathom card, monitor, mouse, and keyboard. Go exploring. Not sure what I will learn, but I want to compare the Fathom OS with the ArduSub one.
Tonight is the big night. Hopefully it all just powers up and works!
h-elsner
Premium Pilot
- Joined
- Mar 23, 2016
- Messages
- 2,310
- Reaction score
- 2,043
- Location
- Bavaria / Germany
- Website
- h-elsner.mooo.com
About the network stuff:
Sorry, again I must ask questions to understand whats going on.
- From the OpenROV manual I understand that the thetering is a 2-wire connection with a top end that has network for control and USB for 5V power. Is this right? With top-end I mean this thing:
- You want to connect the network directly to the computer. Is it a Windows computer or a Raspberry Pi?
- To access the OpenROV offers a web interface https on Port 8080? IP address of the webserver is 192.168.254.1
We can check if your network setup is correct by following command at Windows terminal:
ipconfig -all
You can send the output of the command to a file and post it here.
ipconfig -all >filename
filename should be a file name like C:\temp\network.txt or so somewhere in your computer.
To open the terminal hit win key+R and type cmd, go ahead with the black window...
Network cable should a normal patch cable, Cat5 is enough.
br HE
Sorry, again I must ask questions to understand whats going on.
- From the OpenROV manual I understand that the thetering is a 2-wire connection with a top end that has network for control and USB for 5V power. Is this right? With top-end I mean this thing:
- You want to connect the network directly to the computer. Is it a Windows computer or a Raspberry Pi?
- To access the OpenROV offers a web interface https on Port 8080? IP address of the webserver is 192.168.254.1
We can check if your network setup is correct by following command at Windows terminal:
ipconfig -all
You can send the output of the command to a file and post it here.
ipconfig -all >filename
filename should be a file name like C:\temp\network.txt or so somewhere in your computer.
To open the terminal hit win key+R and type cmd, go ahead with the black window...
Network cable should a normal patch cable, Cat5 is enough.
br HE
I tried three different sources to open the .sch and .brd files from the Electronics folder in the GitHub project for the Fathom One. All say the files are not formatted properly.
