go_away

Author Topic: ROV  (Read 1292 times)

0 Members and 1 Guest are viewing this topic.

Offline MECH_ENGTopic starter

  • Jr. Member
  • **
  • Posts: 14
  • Helpful? 0
ROV
« on: November 08, 2012, 09:47:38 AM »
Hello all I have a question concerning power supplies and motor drivers. The current plan for this ROV will require an AC--->DC converter to act as the power supply. This will power the Pololu VNH3SP30 motor drivers (3) and six thrusters. Two Arduino Unos will be the brain, these will be powered by onboard batteries or batteries at the surface. Are there any concerns with using a power supply to power the motor drivers and motors over using a battery? Should I expect issues or will this function fine. This is a Senior Design project and I just want to be sure I have everthing covered. Thank you!

Offline newInRobotics

  • Supreme Robot
  • *****
  • Posts: 1,015
  • Helpful? 48
  • N.I.R.
Re: ROV
« Reply #1 on: November 08, 2012, 11:41:04 AM »
Powerwise, constant power supply is always better than battery. The only issue I see is cable length, how You ROV is going to be mobile while constantly plugged-in?
"Computer games don't affect kids; I mean if Pac-Man affected us as kids, we'd all be running around in darkened rooms, munching magic pills and listening to repetitive electronic music." - Kristian W

Offline MECH_ENGTopic starter

  • Jr. Member
  • **
  • Posts: 14
  • Helpful? 0
Re: ROV
« Reply #2 on: November 08, 2012, 12:31:56 PM »
Its designed to operate in a tank, the tether will be 50 ft long with analog inputs for controls through an ethernet line and 12 gauge wire for power.

Offline Mastermime

  • Supreme Robot
  • *****
  • Posts: 314
  • Helpful? 5
Re: ROV
« Reply #3 on: November 09, 2012, 05:13:15 PM »
Yes newInRobotics is correct.  Power supplies can supply a constant voltage for longer periods of times as opposed to batteries which can only supply a peak voltage for a small period of time before it dips.  I think the only constraint is the tether.

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: ROV
« Reply #4 on: November 10, 2012, 06:44:28 AM »
Hi,

Are there any concerns with using a power supply to power the motor drivers and motors over using a battery? Should I expect issues or will this function fine. This is a Senior Design project and I just want to be sure I have everthing covered. Thank you!
A well made power supply should work fine, but you need to install large buffer capacitors at the RO(U)V, to handle start current spikes and the effect of cable loss.

What do you have in mind for the supply (voltage/current) and how large is the loading?
(These numbers will tell the needed size of the capacitors).
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline MECH_ENGTopic starter

  • Jr. Member
  • **
  • Posts: 14
  • Helpful? 0
Re: ROV
« Reply #5 on: November 19, 2012, 06:06:31 AM »
Input voltage will be <18V and the maximum current draw at steady state should be around 20 amps (likely an overestimate). The stall current of the thrusters is 11.7 amps and the steady state current at maximum thrust is 4 amps, only two thrusters will be allowed to start at the same time.

Also, a power supply powered by a generator will be used for the power source and the ROV will be controlled with analog inputs from the surface.

Thanks for the responses, any more advice is appreciated.

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: ROV
« Reply #6 on: November 19, 2012, 01:22:03 PM »
Hi,

Also, a power supply powered by a generator will be used for the power source and the ROV will be controlled with analog inputs from the surface.
If you can find room for it, caps 0f 10'000µF on each thruster (right on the power terminals and assuming the power is controlled locally) would be my immediate response, but you could start with smaller caps, as it's hard to know beforehand, how much is needed - it will depend on how the thrusters behave.

This is something that can be done when you have the tether connected and the thruster in water (even a small barrel or a bath tub will do).
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline MECH_ENGTopic starter

  • Jr. Member
  • **
  • Posts: 14
  • Helpful? 0
Re: ROV
« Reply #7 on: November 19, 2012, 02:12:57 PM »
I now plan to use this motor driver: http://www.pololu.com/catalog/product/756

It comes with a 100 uf capacitor which, correct me if I'm wrong (likely as I'm a Mechanical Engineer and this electronics stuff is new to me lol), is designed to handle current spikes when the motors start,  would you still suggest adding 10k uf capacitors across the terminals?

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: ROV
« Reply #8 on: November 19, 2012, 05:19:01 PM »
Hi,

It comes with a 100 uf capacitor which, correct me if I'm wrong [...], is designed to handle current spikes when the motors start, 

I'm more concerned with power surges than spikes from the motor, which the motorcontroller should be made to handle.

You feed power through 50' of an (to me) unknown gauge of wire. This will have a certain resistance, resulting in voltage dips at the receiving end, hence the need for buffer caps, connected right at the supply side of the controllers.


would you still suggest adding 10k uf capacitors across the terminals?

Perhaps you can make do with a little less - it all depends on how your system plays.
Put a 'scope on the input side of a controller and let it rip (in
I now plan to use this motor driver: http://www.pololu.com/catalog/product/756

It comes with a 100 uf capacitor which, correct me if I'm wrong (likely as I'm a Mechanical Engineer and this electronics stuff is new to me lol), is designed to handle current spikes when the motors start,  would you still suggest adding 10k uf capacitors across the terminals?
water to get the proper load), then you'll see the voltage drop you have to deal with.
The formula to use is C = A*s/V
Where:
C = capacitance in Farad
A = current in ampere
s = period of the dip in secconds
V = the voltage drop you'll allow
Note, this is the formula for how the capacitance behave when charged once and then loaded with the current*time/drop, but since the voltage will probably not drop to zero due to the supply, you won't need quite as much capacitance as the formula will show.

The heavier the gauge, the less capacitance is needed, assuming a good supply.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline MECH_ENGTopic starter

  • Jr. Member
  • **
  • Posts: 14
  • Helpful? 0
Re: ROV
« Reply #9 on: November 21, 2012, 09:28:13 PM »
Thank you for the replies, greatly appreciated!

One more question however, since I'm using analog control with 50' of ethernet line, what would be the best way to provide 5 volts to the potentiometers? Should I provide the 5 V from a source at the surface or from the 5 V pin on the Arduino aboard the ROV via a strand on the ethernet line?

Again thank you!

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: ROV
« Reply #10 on: November 23, 2012, 03:49:14 PM »
Hi,

[...] since I'm using analog control with 50' of ethernet line, what would be the best way to provide 5 volts to the potentiometers? Should I provide the 5 V from a source at the surface or from the 5 V pin on the Arduino aboard the ROV via a strand on the ethernet line?
What potentiometers and what exactly do they control?
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline MECH_ENGTopic starter

  • Jr. Member
  • **
  • Posts: 14
  • Helpful? 0
Re: ROV
« Reply #11 on: November 23, 2012, 04:41:39 PM »
The potentiometers are inputs to the Arduino board for speed and direction control of the thrusters and robot arm (servo driven). We plan to use the analog 3 axis joystick shown below.


http://www.robotshop.com/3-axis-joystick-gim-04.html?utm_source=google&utm_medium=base&utm_campaign=jos

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: ROV
« Reply #12 on: November 23, 2012, 07:36:26 PM »
Hi,

The potentiometers are inputs to the Arduino board for speed and direction control of the thrusters and robot arm (servo driven). We plan to use the analog 3 axis joystick [...]
That is a poor solution.
The best would be to have eg. one of the cheap and very small Arduino compatible boards right at the stick, supply it and read it locally and then send a serial (digital) signal down the tether - a synchronous serial signal (i.e. Clock and Data) will probably be the easiest solution.

If you were to have 50' of wire between the stick and the A/D-lines to read it, you'd have nothing but trouble. Exactly how much is hard to say, as I cannot find any info on the potentiometer resistance (a bad sign).

The pots are likely between 50kOhm and 500kOhm and the A/D-C inputs are higher impedance - probably 1MOhm or more (see datasheet). This would result in fairly random values read at the receiving end, as the tether would act like an antenna and pick up anything from strong radio signals (broadcast and narrow-band), as well as any noise from switching electricity in any form, within a certain distance.

It's hard to ruin a digital signal with noise, but if it happens, sending data by a current signal (as opposed to a voltage signal) is easy and will cure any tendency to pick up noise.
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline MECH_ENGTopic starter

  • Jr. Member
  • **
  • Posts: 14
  • Helpful? 0
Re: ROV
« Reply #13 on: November 23, 2012, 07:48:38 PM »
The tether will be an ethernet line not sure if that would change your answer, if not could you point me in the right direction for digital control?

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: ROV
« Reply #14 on: November 24, 2012, 10:43:35 AM »
Hi,

The tether will be an ethernet line not sure if that would change your answer,
Still the same answer on the joystick control, but I wouldn't try to feed 18V at up to 20A through a Cat-5 cable.

The resistance of a single conductor in a Cat-5 varies from ~0.07 Ohm/m to over 1.2 Ohm/m (depending on make and quality). 50' is 15.24m and the length of the current path is twice that (two wires + and -).
30.5m will be between 2.135 Ohm to 36.6 Ohm.
2 thrusters (18V/4A each) equals a load of 2.25 Ohm.
So...
With the best Cat-5, the voltage over the thrusters will be:
   18V * 2.25 Ohm / (2.25 Ohm + 2.135 Ohm) = 9.24V

And with the worst quality cable, it will be:
   18V * 2.25 Ohm / (2.25 Ohm + 36.6 Ohm) = 1.04V

Parallel connecting conductors will help of course, but you'll still have a huge loss.
Solution... use the Cat-5 for control and strap a much heavier cable on to it.



could you point me in the right direction for digital control?
I thought I did? :)

A small controller to sit at the stick, reading it and transmitting the values down the Cat-5 to the onboard Arduino. What is puzzling you here?
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline MECH_ENGTopic starter

  • Jr. Member
  • **
  • Posts: 14
  • Helpful? 0
Re: ROV
« Reply #15 on: November 24, 2012, 12:22:47 PM »
My previous post was confusing, sorry about that.

The tether will consist of 3 parts: Cat 5 cable, 12 gauge power line, and a video line

The Cat 5 line will only be used as a connection between the analog joystick and Arduino. Will noise on ethernet line from the current in the power line be a significant concern? Also the pots in the joystick are 5kohm.




P.S.
You have been extremely helpful and have helped me prepare for questions during the final presentation of the desgin, thank you

Offline Soeren

  • Supreme Robot
  • *****
  • Posts: 4,672
  • Helpful? 227
  • Mind Reading: 0.0
Re: ROV
« Reply #16 on: November 24, 2012, 07:46:03 PM »
The Cat 5 line will only be used as a connection between the analog joystick and Arduino.
OK.


Will noise on ethernet line from the current in the power line be a significant concern? Also the pots in the joystick are 5kohm.
Whether induced noise from the motor cables will be a problem or not is hard to guess, as it depends on several factors (the main issue being how the motor controller handles things). Terminating the cable impedance correctly and making use of the twisted pairs as intended will help, but, to repeat myself, go digital and save yourself a lot of worries.
It's no skin off my back, if you want to connect the pots via a 50' cable and tinker with that for som time, but you're gonna end up with digital transmission in the end (or a complete control Chaos), so why not save yourself some time? :)
Regards,
Søren

A rather fast and fairly heavy robot with quite large wheels needs what? A lot of power?
Please remember...
Engineering is based on numbers - not adjectives

Offline MECH_ENGTopic starter

  • Jr. Member
  • **
  • Posts: 14
  • Helpful? 0
Re: ROV
« Reply #17 on: November 24, 2012, 10:40:08 PM »
Thank you for your help!

 


Get Your Ad Here