I only meant "analog" in the either off or on sense. I was starting to get the impression that a servo was necessary if say, I pushed a stick half forward, telling a motor to spin at half speed. I need to find better words for on/off and on/in-between/off haha.

Update: (sorry for the very quick, gamechanging update). The overall design is just about finalized. 4 bilge pumps. 1 each on the front and back for rising/diving. 2 left and right of center for forward propulsion/turning. Most likely the 12V, 500gph ones (model 375196) found on this page: http://www.westmarine.com/webapp/wc...chanintel_google&ci_src=14110944&ci_sku=34144 . Each bilge pump will be turning a propeller. If it makes it easier, I am fine with two speeds: full speed ahead or not moving.

Back to my servo question, if I plan only to have each pump either on or off, are servos still necessary? Is an ESC still necessary? The sub must have a minimum speed of 1.5"/sec; no maximum speed requirement.

 

The problem is that if you get an off the shelf RC controller, the receiver outputs will be of the PPM type (1.5mS Center). You'll need to use a PIC or AVR to decode what that signal is supposed to do. Some offer 4 PPM I/O plus 2 digital I/O, but they are harder to find.

Look through Tower Hobbies for a transmitter/receiver pack that will match your needs. Once that is decided, the rest is easy, a lot of it is available from Tower Hobbies in fact!

The 500gph pump draws 1.9 Amps@12V. If you do much in the way of moving water, your batteries will be dead fast.

 

The submarine must have a minimum forward speed of 1.5" per second. There is no maximum speed. I am fine with 2 speeds: going forward and not going forward.

I assume that it can go slower than that, but that its maximum speed must be at least 1.5"/sec. If that assumption is incorrect, please let us know.

I was starting to get the impression that a servo was necessary if say, I pushed a stick half forward, telling a motor to spin at half speed. I need to find better words for on/off and on/in-between/off haha.

The words are fine. Just use on and/or off as you have. That's clearer than getting confused with digital vs. analog.

As for what a servo does, that is not quite right. The type of model servo we have been talking about controls position. If you have not seen one, go to a hobby shop. It will be more than happy to show you. An ESC controls the speed of an electric motor. Now of course, a servo can control the throttle position on a liquid fuel engine, which in turn, controls the power/speed. But you are not using such engines here.

Update: (sorry for the very quick, gamechanging update). The overall design is just about finalized. 4 bilge pumps. 1 each on the front and back for rising/diving. 2 left and right of center for forward propulsion/turning. Most likely the 12V, 500gph ones (model 375196) found on this page: http://www.westmarine.com/webapp/wc...chanintel_google&ci_src=14110944&ci_sku=34144 . Each bilge pump will be turning a propeller. If it makes it easier, I am fine with two speeds: full speed ahead or not moving.

I don't understand what you mean by the pump turning a propeller. Are you sure the intent is not to just rely on the jet effect from the pump?

Those pumps are huge and expensive. How big a sub are you making (overall dimensions)? Can you fill us in on why the decision to use a bilge pump rather than an open or shrouded propeller? How do the bilge pumps work? That is, are they centrifugal pumps or gear pumps?

Back to my servo question, if I plan only to have each pump either on or off, are servos still necessary? Is an ESC still necessary? The sub must have a minimum speed of 1.5"/sec; no maximum speed requirement.

A servo would be a poor way to turn a pump on or off. Some people have used servos to actuate switches for that purpose, but doing it electronically is easier. There are many DIY ESC designs on the Internet. You will also find designs for simple on/off controls based on the 1.5 mS center PWM. Note, simple 4-channel transmitters will usually have proportional control on each channel. When other channels are added, they are not necessarily proportional. Some may just be on/off (i.e., > or < 1.5 mS).

Finally, back to the diving function. Some subs (like the ones I linked to appear) depend on forward motion and dive planes to submerge. They may not be able to stop under water. That is, they are neutral buoyancy. Other subs work more like real subs or blimps. They take on or discharge water to rise or sink. What method will you be using? If you plan on the latter, then you need to plan on keeping compressed air. A bilge pump filling or emptying an rubber balloon inside an airtight cylinder might work, but then you need to consider how much head pressure it can operate against. Maybe a gear pump or piston pump would be better.

Just using a compressed air cylinder and control valve would probably be much simpler, but would limit the number of dive cycles. How many dive cycles must you be able to complete? Off the shelf, RC control valves are available for that purpose.

John

 

The maximum speed must be a minimum of 1.5" per second. Sorry for the unclear wording.
The bilge pumps will be turning propellers using the jet effect.
Dimensions: 8"<W<10", 8"<H<10", 32"<L<36".
We must be able to make two dive/rise cycles.
We will be using neutral buoyancy.

The propellers in the front and the back will be pointed "down" and must be able to spin CW and CCW for diving and rising. The two propellers on the side only need to be able to spin to propel forward.

About the transmitter, I need to be able to tell the front/back ones fwd/reverse, so that's 4 channels, and then the side ones fwd, for 2 more. I need a 6 channel transmitter? Can I use this for the side on/off pumps? http://www.dimensionengineering.com/picoswitch.htm

I really wish I could tell you the rationale of these decisions, but the majority of these were made without me. The good news, though, is that now I can finally declare, with some confidence, the details of the design instead of being as vague as I have been. I think these structure decisions are near-final unless I have major issues with them.

Again, I thank you all for your patience as I take you on this crazy ride of everchanging designs and requirements (but looking forward to some stability). Please keep up the questions; I'll do my best to answer them.

 

If all 4 motors on the front and rear are all pushing down or up at the same time an speed then you will only need 1 channel for this. You will only need one channel for the side ones as well. This would make it a 2 channel transmitter/receiver. If each of the motors are to be controlled individually then you are right, you will need a 6 channel unit.

 

We will be using neutral buoyancy.

No part of the boat will ever be above water, right?

The propellers in the front and the back will be pointed "down" and must be able to spin CW and CCW for diving and rising. The two propellers on the side only need to be able to spin to propel forward.

About the transmitter, I need to be able to tell the front/back ones fwd/reverse, so that's 4 channels,

No, one channel can tell one motor fwd or reverse. If the motors are wired together so they operate the same, then you need only one channel for the dive control. I would not recommend doing that. With independent control, you need two channels total for dive function. If you go the two-channel route, I would not wire the motors in a logical sense completely independently. I would connect them to the radio/mixers so they worked together, but a trim function could be used to fine tune the angle of descent or ascent. It is similar to how twin-engine electric aircraft are "wired."

and then the side ones fwd, for 2 more. I need a 6 channel transmitter?

No. Since you want to operate the side motors independently for forward and for turning control, you need two channels for them. The advantage is that you can also have independent reverse, which can be useful in maneuvering any boat.

So far, we have defined a maximum of 4 channels or just 3 channels, if you use only one channel for dive control. As you look for transmitters, if you find one that has computer "mixers" it will be helpful. If it were a model aircraft, the turning, forward, and reverse mixer would be called a V-tail mixer. That will allow you to use one control stick for turning and forward/reverse. I would put that on the right stick. I would put the dive function on the left stick

Can I use this for the side on/off pumps? http://www.dimensionengineering.com/picoswitch.htm

That relay is only rated at 1 A; the bilge pumps you specify are 1.9 A. It might work, but you would be over stressing the relay. It will not give you reverse, and it is just on/off. For half that price ($10) you can get a fully functional, 20A ESC from Hobby King. Then, you will have proportional control and reverse, even if you don't use it. Those cheap ESCs are everywhere. Hobby King is just one option for a supplier. It is in Hong Kong. Many guys in our local club, including myself, have ordered from it and gotten good service.

In sum, my recommendation is a minimum 4-channel transmitter with mixing and 4 ESC's. Since you have mentioned sonar and other functions, more channels may be needed. Good news, the cost is not directly proportional to the number of channels.

John

 

1 channel for each side pump. Can you explain further why you don't recommend keeping the dive pumps on a single channel/control? The sonar will be a set and forget; it will require no interaction besides providing power. Mixers take multiple signals and multiply them into a single output?

These are the 20A ESCs you mentioned?
http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=14871

4-channel transmitter, and 4 ESCs (above). I'll start checking out transmitters.

On a side note, John, you're clearly very knowledgeable in electronics as well hobby matters. You mentioned you're in some sort of RC club? Is that club just model planes?

-Tim

 

1 channel for each side pump. Can you explain further why you don't recommend keeping the dive pumps on a single channel/control? The sonar will be a set and forget; it will require no interaction besides providing power. Mixers take multiple signals and multiply them into a single output?

Forward and aft thrust motors will never be exactly balanced about the center of gravity/center of thrust. Thus, if you want to go up and down with any semblance of a level attitude, you need independent control.

Mixers take signal inputs and combine them into appropriate signals for multiple servos/esc's.

These are the 20A ESCs you mentioned?
http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=14871

Yes

4-channel transmitter, and 4 ESCs (above). I'll start checking out transmitters.

I don't know your budget, but I suspect you will find a 6 to 9 channel TX/Rx setup for almost the same price. Get something decent. If you have to decide between a Tx with more channels than the receiver or visa versa, go for the Tx with more channels. Many "four channel" receivers are electronically the same as receivers with more channels. The manufacturers just don't bring the additional channels out to the plugs. There are many descriptions on the Internet about how to access the additional channels. Just think about it. Why would somebody make a chip for a 4 channel receiver when a chip for a 6 to 8 channel receiver doesn't cost any more? So, they use the 6 or 8 channel chip and only make contacts to 4 channels.

On a side note, John, you're clearly very knowledgeable in electronics as well hobby matters. You mentioned you're in some sort of RC club? Is that club just model planes?

-Tim

Yes. Planes, helicopters, parachutes -- anything that flies. One friend has a small business making RC parachutists.

I have been at it a long time. My first RC airplane was in 1952 -- a J-3 Cub with rubber-band powered escapements instead of servos. Only one channel then. The pilot had to do the coding with a key.

John

 

How about RC control for fore diving motors & internal tilt sensor for aft motors; saves on telemetry to know attitude.

 

Good idea.

Also, we have not talked about the possibility that it might rotate on its longitudinal axis, i.e., capsize. Will it have a keel? With all the heavy bilge pumps and batteries on board, it might be hard to get the CG for that axis adjusted so it is stable.

John

 

An air pocket/balloon at the top and some bathtub experiments.

Is this going to be video operated or just look-through the water operated?

You are going to want some bright LEDS to ensure you dont loose her.

 

We have a pretty healthy budget; we'll be able to spring for something nice in terms of transmitters/receivers. Do you have any brand-loyalty-based suggestions in terms of brands?

I think I'll stick with RC controls of forward/aft dive/rise pumps, for simplicity's sake. Very creative suggestion though.

In terms of CG, we're going to place our components on a metal plate in the middle, and have movable weights to help balance the unit and "center" the CG. I don't think we'll have a keel. We're meeting later today, I'll have a much better grasp on the structure details soon.

We will be operating in a swimming pool; we'll be able to see the submarine its entire trip.

 

Bernard's suggestion, which I strongly endorse, would not change the number of dive/rise pumps. It would just make operating that function a lot easier. Remember, your sub at neutral buoyancy will not be very stable in attitude as it goes up and down. The "tilt" sensor is also known as a one-axis gyro. (At least that is what I think he meant.) They are off the self items and are used in many models, most notably helicopters and some airplanes where a wing leveler is needed. Basically, the gyro would be set to control the thrust on one of the pumps. You would control the other for determining whether to rise of sink. That would potentially free up one channel, but you might want to keep that channel as a trim.

If you build your sub so the weight is concentrated in the middle along the main axis, you could paint it brown and enter it in a log rolling contest too.

John

 

Thanks John, I forgot about the Memsic 2125 Dual-Axis Accelerometer [28017] available from the Shack, great tilt sensor.
My origional sensor was a mercury switch with contacts on each end of curved glass tube with a center common contact.

 

After looking at the datasheet for the sensor Bernard mentioned, here's what I understand: I wire the forward pump (chosen arbitrarily for this explanation) to be controlled by an ESC and the receiver. The aft pump is wired (independently of the forward pump) to a tilt sensor/accelerometer. When the tilt threshold along the main axis (longitudinal?) is reached, the sensor tells the aft pump to spin the propeller either way to correct for the tilt. Thus, the forward and aft pumps are working in unison during dive/rise, with the added benefit of keeping the longitudinal axis level at all times.

The benefits are great, but it will add a huge layer of complexity. I'll have to program/code appropriately, test it on a breadboard, design a PCB schematic, have it manufactured, and finally, solder everything together.

I'm not saying this isn't a good idea, or that I won't consider it. But for now, I'd like to have the most basic of concepts down. Once I reach a completed first design, I will reconsider this idea. Our lab has breadboard serial hookups for programming, but I will not be able to access these during winter break (quickly approaching).

As for balance/capsizing/etc., the mechanical team will be in charge of placing my components, and balancing the unit appropriately.

Switching gears, what do you think of the following transmitter (and included receiver): Futaba 72MHz, 6-channel; with included 8-channel receiver? http://www.rcplanet.com/Futaba_6EXP_6-Channel_FM_p/FUTK62**.htm

 

Switching gears, what do you think of the following transmitter (and included receiver): Futaba 72MHz, 6-channel; with included 8-channel receiver? http://www.rcplanet.com/Futaba_6EXP_6-Channel_FM_p/FUTK62**.htm

It's illegal for this purpose. I sincerely hope you will not proceed in that direction.

As for the tilt sensor, have you looked up single-axis gyros (or more), as I suggested? Same principle, but no programming involved. Can you clarify what exactly is your role on the team?

John

 

Very well, I will look for another transmitter/receiver. I don't want to do anything illegal during this project; if you can tell it is illegal, then people who read my report will know it's illegal. It is illegal because it would be using aerial frequencies for surface purposes, yes?

My role in the actual project is the successful powering of the unit and the transmission of control signals. I am also project manager, handling the brunt of the report writing, meeting scheduling, documentation, etc. I hear your balance/structure concerns, and while I can't address them directly, I relay them to the mechanical team. I am not an electrical engineering major, but rather an engineering science (or engineering management) major (with electrical specialization). The curriculum substitutes engineering courses with business courses, with hopes to produce business-knowledgeable engineering managers/system engineers.

Continuing my search for programming-less tilt sensors, I have come upon these: http://www.meas-spec.com/position-s...s/single-axis-tilt-sensors-inclinometers.aspx
The G-series says "switch" for output and outputs a 10mA "alarm" signal when 10° of tilt is detected. No programming needed. Can I connect this to an ESC and have the alarm signal trigger the ESC, activating the bilge pump?

 

Continuing my search for programming-less tilt sensors, I have come upon these: http://www.meas-spec.com/position-s...s/single-axis-tilt-sensors-inclinometers.aspx
The G-series says "switch" for output and outputs a 10mA "alarm" signal when 10° of tilt is detected. No programming needed. Can I connect this to an ESC and have the alarm signal trigger the ESC, activating the bilge pump?

You might. But why that complexity? Moreover, it is just one or off control.

Here is one of dozens of gryros available for models to stabilize them.

http://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=9098

Prices range from less than $10 to more than $200. Some include ESCs, receivers, servos, etc. (so called 5 in ones). The idea is that the tilt sensor/gryo provides a servo-control type of signal proportional to the displacement. You can control that gain. In your project, the servo signal runs the ESC. In airplanes, the signal usually controls a servo.

You seem to think that motors/controls will be full on or off. It is possible to control an RC model in that way, but it is so much easier to control one, if the controls are proportional. That is, you can control the propulsion from off to full power in almost a continuous fashion (anywhere from 512 to 2048 steps is common).

You really would benefit from visiting a model boat club. If that is not possible, then visit a model car or airplane club. You will see what I mean.

John

 

I'll echo what jpanhalt stated as far as attitude control.

Using a sensor that alerts at 10° in a simple control system would cause the motors to reverse until the sensor alerted to -10°, the system would continually be fighting itself, or "Doing the Disco".

ESCs aren't On/Off, they give, Reverse, off, 512 or more steps of speed to full ON, in addition to Off. Using this with a standard accelerometer interface would only 'nudge' the motors a bit to get back to near "level".

This is instead of waiting until the sub is 10 degrees pitched, then spinning a propeller at full speed, only to reverse it a few seconds later.

 

Okay, I see the flaws with my previously suggested basic control system.

From what I can understand from the page and the below links, the Pass-Con unit linked in jpanhalt's post allows a piezo gyro to be turned on/off from the transmitter.
http://www.gwsus.com/images/note%20book/e_pc3-1.jpg (instructions)
http://www.gwsus.com/images/note book/e_pc3-2.jpg (wiring diagram)

This is the piezo gyro that corresponds with the Pass-Con. A piezo gyro measures angular movement.
http://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=9094

I really can't wrap my mind around what's happening with these 2 devices. The pass-con enables the transmitter to shut off/turn on a tilt sensor? I know I'm missing something big here; someone please explain this to me.

I stumbled upon a (pricey) helicopter gyro and servo package at Tower Hobbies. Reading the product info page and instruction manual makes more sense to me. When the gyro detects drifting (caused by wind, etc.), it sends correction signals to the attached servo, attached to the rudder.
http://www3.towerhobbies.com/cgi-bin/wti0001p?Q=1&I=LXFZP2&P=3#tech (product info)
http://manuals.hobbico.com/fut/futm0807-manual.pdf (manual)

Side note about tx/rx: 72 MHz reserved for aerial models, understood. Hunting for 27 or 75MHz radio systems.