Hybrid Ballast Tank/Piston System

[Monahan Steam Models]

Latest on the hybrid setup: https://youtu.be/fjfNuDgUd8Y

[jlcrl]

Hi Nick,
I made a very tiny muscle wire servo in 2011 but I think it is not the solution for torpedo and micro submarine control. The best way for depth control would be using your laser feedback invention for a micro tank under antenna, making the device capable to sink as an aquarium micro submarine. A model maker was able to hold his contra rotating gearbox torpedo just some centimeters under water and the result is impressive on video. I explain this on torpedo post as it is not the topic here.
JL

[Monahan Steam Models]

Jean Louis,

Thank you for that tip. I had not heard of Magmaster until you mentioned it. I will definitely take a look at it. Eliminating the dynamic calibration procedure would be great. The dynamic calibration for the Mag was the only drawback I found. With the 9 DoF sensor, I was able to apply a linear quadratic equation (Kalman filter) algorithm for processing the raw data.

To represent the movements of the 9 DoF gyro in a 3D animation, I wrote a program and used VPython. Worked very well.

Looking forward to seeing more of your work with the torpedoes. I was looking at using Nitinol wire or muscle memory shape wire to create miniature actuators to provide control to the rudder and planes. So far I like what you have done with the coil and magnets for the rudder on your design. Look forward to seeing some testing videos!

Nick

[jlcrl]

Nick, do you know the MagMaster doing calibration for any sensor once ?

https://diydrones.com/profiles/blogs/advanced-hard-and-soft-iron-magnetometer-calibration-for-dummies

I have been using it successfully for magnetometer and accelerometer.
Without calibration the sphere is ovoid, once done it is a perfect sphere. 3 axis may be done together.
The program generates something like this just to be inserted into your arduino code and no more dynamic calibration is required.

double calibration_matrix[3][3] =
{
{7.037,-0.212,0.083},
{0.001,7.014,-0.106},
{-0.079,-0.207,7.066},
};

double bias[3] =
{329.548,329.062,328.391};

I made a modified version to generate the calibration code to insert, and also 2 viewers.
One is just an arduino data display but without scrolling, and the other one include a 3D viewer so you can verify what the sensor understands for example holding materially the
sensor at 10° pitch 3° azimuth you can see on the sensor side if it matches so have an idea of good calibration or not.
I can send you exe + source if you like (visual studio C#)
Jean Louis

[Monahan Steam Models]

Thank you Jean Louis,

I tested 5 different DoF sensors. I did like the 9 DoF and it was very precise and did not drift like a lot of sensors do. What I did not like about the 9 DoF sensor was the physical input needed to properly get the magnetometer to calibrate each time it was switched on.

The sensor that I decided to use for the project after testing the others, is a 6 DoF sensor. I utilize the accelerometer data as well as the gyroscope data to read pitch. I prefer using data from both as each has the opposite strengths and weaknesses for measurements. I use high/low pass filters on the data coming from each. The program I wrote also uses Euler Angle formulas to correct pitch axis with data from the roll axis.

I added some P.I.D. equations to the output data that is sent to the trim tank pumps so that the trim change responses can be smoothed once the submerged water testing begins.

Thanks for the link to you video. I enjoyed watching it.

Nick

[jlcrl]

Congratulations Nick, it 's a real improvement using the laser for distance feedback and a 9 doF.
Did you see a real difference if only the pitch is tested (accelerometer mode only) so not using the roll (gyroscope) ?
I made a test with and accelerometer only and did not observed significant movement on roll
video is there:

https://www.ladsm.fr/viewtopic.php?f=7&t=18

Jean Louis

[Monahan Steam Models]

Thank you, had a heart attack 4 weeks ago or so.
But still here and still ticking!

Cut that out Tom! You’re not allowed to have a heart attack!

Seriously, I glad you are still here and ticking!

The potting epoxy worked out great so far. I’ll post an update soon. The power just went out here 20 minutes ago so I’m pecking this message out on my phone.

Nick

[trout]

Hi Tom,

Just a short note! I hope you are OK!? I understand, hospitals are no fun! Take care my friend!

Rob
"Firemen can stand the heat"

Thank you, had a heart attack 4 weeks ago or so.
But still here and still ticking!

[rwtdiver]

Excellent.....when I go to the hospital, and it seems I have done that a little too much recently, I ask for the saline syringe - I always need to explain why, but I use it to inject epoxy into areas, like what you are doing here, once the nurses understand, they are happy to help out.
I have never used the potting epoxy. Look forward to your report on it.

Hi Tom,

Just a short note! I hope you are OK!? I understand, hospitals are no fun! Take care my friend!

Rob
"Firemen can stand the heat"

[trout]

Excellent.....when I go to the hospital, and it seems I have done that a little too much recently, I ask for the saline syringe - I always need to explain why, but I use it to inject epoxy into areas, like what you are doing here, once the nurses understand, they are happy to help out.
I have never used the potting epoxy. Look forward to your report on it.

[Monahan Steam Models]

Was able to do some work on the water proof brushless motor bulkhead fittings today.



Started out by cutting some 14 awg solid copper wire to the proper length needed, then soldered them to the bullet connector male plugs.




Sorry for the blurry photo above. Was trying to capture the cross holes that are drilled into each of the gold plated female bullet connectors. They're the dark dots that appear on the sides of the pins. These cross holes are only female plug pins. I will use these later as a vent for filling potting compound.



Male and Female plugs soldered back to back.




Keeping the alignment between the male and female connectors during soldering was important so that they would fit correctly into the white ABS printed housings.




Boom! Nailed it. They fit perfectly into the housings.




After the housing have been bonded together, I drill a cross hole through the ABS housings at a level just slightly above the vent holes shown earlier in the female bullet connectors.




The drilled hole is the same size as the blunt syringe applicator needle used to inject the flexible epoxy potting compound.




This is the potting compound mixed 1:1. I learned this particular potting compound from Sub Ed when he build his skeletal dive system!




A temporary hot glue bead was placed around the base of the male connector and ABS housing to created a seal. Then the potting compound was slowly injected into each connector assembly. The trimmed off ends of 22 awg silicon wire were used as temporary plugs for the fill holes in the housings.



While injecting the potting compound into the connector housings, I was able to monitor the level of filling by looking down into the bores of female bullet connectors to see the level of potting compound raise and lower through the vent holes. The connector on the left you can just make out the back potting compound at the bottom of the bore.


I'll let the sit undisturbed for the next 48 hours which should be enough time for the potting compound cure. Then they will be cleaned up and prepped for installing into the rear WTC end cap.


Nick

[Monahan Steam Models]

Added the main ballast tank fitting. I originally didn’t attach this to the main tank because the project was still too far away development wise. We are getting closer...

Nick

[Bob Gato]

Disclaimer:

... Rob- Maybe I misunderstood the above post-...

...Got it- all good thanks-BG

[rwtdiver]

Thank you Rob,

The tech available to the flight hobby mostly translates into the sub hobby but BG is correct, the GPS signals will not penetrate water. This boat that I’m working on did have a 9 DoF sensor on board so pitch, roll and yaw was accurate and recorded, but in the end I wound up using a simpler 6 DoF sensor and mainly use the x axis for tracking the pitch. Roll or (y axis) is used but only to correct the pitch axis with a Euler formula.

A set up like this does add some trim tweaking capabilities outside of the initial add lead and add foam to get the boat trimmed correctly but it is also a huge added complication that is not necessary for a typical build.

Really the only reason I Initially pursued the building and testing of a system like this was due to the final ballast differences between loading the forward and aft torpedo tubes with either electric or gas weapons as each of live and loaded weapons into the tubes would have a different effect in final trims.

Proven systems like the snort or SAS are rock solid and best in my opinion and that’s why my next build will be those methods. This build is an odd monkey that I’m testing the what if boundaries.

Nick

Nick & Bob,

I may have not stated what I was trying to say clear enough! I realize GPS systems can not be used in a water environment. I was only trying to say that Nicks auto GYRO pitch control stabilization system for the submarine is similar in its function as a GPS (satellite) system holding a quadcopter or VTOL aircraft level at a stationary point! I realize they are two completely different type systems! But the concept of AUTO control is what I was trying to get across!

Final word on this subject! Nick I really like your design and the great concept you have built into your sub. I have not been able to fly any RC aircraft for years! I am totally into RC submarines, and I am trying to learn and understand all I can! Keep posting your build!

Rob,
"Firemen can stand the heat"

[Monahan Steam Models]

Thank you Rob,

The tech available to the flight hobby mostly translates into the sub hobby but BG is correct, the GPS signals will not penetrate water. This boat that I’m working on did have a 9 DoF sensor on board so pitch, roll and yaw was accurate and recorded, but in the end I wound up using a simpler 6 DoF sensor and mainly use the x axis for tracking the pitch. Roll or (y axis) is used but only to correct the pitch axis with a Euler formula.

A set up like this does add some trim tweaking capabilities outside of the initial add lead and add foam to get the boat trimmed correctly but it is also a huge added complication that is not necessary for a typical build.

Really the only reason I Initially pursued the building and testing of a system like this was due to the final ballast differences between loading the forward and aft torpedo tubes with either electric or gas weapons as each of live and loaded weapons into the tubes would have a different effect in final trims.

Proven systems like the snort or SAS are rock solid and best in my opinion and that’s why my next build will be those methods. This build is an odd monkey that I’m testing the what if boundaries.

Nick