Autopilot controlled by reversing polarity via relays help needed

Thread Starter

Vonasi

Joined Aug 19, 2018
34
I have an old autopilot control head that simply outputs a 12v supply to a motor and reverses the polarity to change the direction of the motor to steer in the opposite direction. I am thinking of changing the motor to a linear ram that works the same way by reversing the polarity to push or pull. I suspect that the old autopilot control head cannot cope with the higher amp loads of the linear ram and so I need to find some way to switch the linear rams direction using the reversing polarity via relays (and probably diodes I am guessing). At the moment I just cant get my head around how to achieve this. Could anyone point me in the right direction (no autopilot pun intended)?
 

Thread Starter

Vonasi

Joined Aug 19, 2018
34
The autopilot head is so old and the manufacturers wash their hands of their products after only a few years so exact specification is impossible to find out - but bearing in mind the size of the motor that it currently drives and the low amps that draws I have to assume that it is not designed to drive something that draws up to 20 amps. I feel it is best just to wire up something that works and can handle the amps and then know that nothing is overloaded.
 

Thread Starter

Vonasi

Joined Aug 19, 2018
34
I have found some documents on line that show the autopilot control head being protected by a 12A fuse so this seems to point that it would not cope with the 20A of the linear ram so I need to find a relay wiring solution if any one can assist.
 

DNA Robotics

Joined Jun 13, 2014
647
My first thoughts on this were like trim / tilt relays on an outboard that change polarity based on which relay you energize. Both motor wires are grounded in the normally off contacts. Whichever relay you energize makes that motor wire positive.

This video starts talking about 2 relays at the 5 minute mark.

 

Dodgydave

Joined Jun 22, 2012
11,285
If the Autopilot has two outputs to drive the motor, then simply use those outputs to drive dpdt relays at the rated coil voltage. Then use the contacts to feed your linear ram.
 

Thread Starter

Vonasi

Joined Aug 19, 2018
34
The autopilot control head only has one output. Two wires that it reverses the polarity on to make the motor/ram work in two directions.
 

Thread Starter

Vonasi

Joined Aug 19, 2018
34
Thanks for the video link DNA Robotics but that does not work in this case. The problem is how to only energise one relay when the polarity is one way and only energise the other relay when it is reversed.
 

Thread Starter

Vonasi

Joined Aug 19, 2018
34
Yes it does produce zero volts. It is either giving no output or 12v in either polarity to go left or right.
I think I may have come up with the solution using two diodes and two DPST relays and will try to create a wiring diagram of it to post on here for others to give their thoughts on.
 

DNA Robotics

Joined Jun 13, 2014
647
Thanks for the video link DNA Robotics but that does not work in this case. The problem is how to only energise one relay when the polarity is one way and only energise the other relay when it is reversed.
Sure it will. Ground the coil of both relays. Whichever wire goes positive will turn that relay on.

Change polarity relays.jpg
 

crutschow

Joined Mar 14, 2008
34,285
You should add four diodes, going from each motor winding to V+ and ground, to suppress the motor inductive transient that can damage/weld the relay contacts.

If you wanted to go solid-state and avoid relays, you could also do it with a four MOSFET H-bridge (below).
The MOSFETs should be sized to handle the motor current. For 20A, the MOSFET ON resistance should be no more than about 2.5mΩ, to avoid having to put them on a heatsink.
The MOSFET substrate diodes absorb the inductive transient, so additional diodes are not needed.
upload_2018-8-19_12-26-26.png
 
Last edited:

cork_ie

Joined Oct 8, 2011
428
The circuits already provided by others will allow you to drive your linear actuator.
However it is not quite that simple, A number of the larger autopilots had rudder position feedback sensors eg NECO, CEETREK, B&G, AUTOHELM ST6000 & ST7000 systems.
Most of the smaller autopilots, with built in rotary or linear actuators, did not, and the rudder stops were determined by current draw in the output circuit being monitored and stored as a calibration in memory.
If you just connect up your new linear actuator, as you suggest, without any form of angular limits; You will at a minimum do damage to your steering system and possibly apply enough force to burst the actuator mounting and force it out through the side of the boat.
You are also very likely to have difficulty with Yaw control unless you have a very accurate compass with fast response.
 

Thread Starter

Vonasi

Joined Aug 19, 2018
34
cork_ie - I am aware of all that you have commented on and I am in communication with the linear ram manufacturers regarding some of these issues. The ram I am considering is actually a combo hydraulic/electric ram (not a continuously running hydraulic pump but a reversible hydraulic pump) with overload valves built in to the hydraulic circuit so that it should not be possible to overload anything if the ram is installed correctly. If the maximum travel limit of the combo hydraulic ram is reached before the rudder hits its stops then there should be no possibilities of the ram causing damage.
As for speed of response the autopilot control head I am considering using has adjustable 'rudder response' so it should be possible to calibrate it for different reaction speeds needed.
At the moment I am just considering whether what I am thinking of doing is actually possible as the option of buying a full manufacturers kit of display, compass, computer, rudder angle sensor and ram is about 1/5th of the value of the boat whereas buying just a linear ram and a couple of relays is a third of that price. Why you can buy a powerful laptop for £250 but a simple autopilot and ram costs £2500 is beyond me but apparently all us sailors have money to burn.
 

Dodgydave

Joined Jun 22, 2012
11,285
Yes it does produce zero volts. It is either giving no output or 12v in either polarity to go left or right.
I think I may have come up with the solution using two diodes and two DPST relays and will try to create a wiring diagram of it to post on here for others to give their thoughts on.
Simple, the use the two wires to feed your relays with series diodes, so that only one relay operates in each direction, and use the contacts to reverse your motor.


reversible-dc-motor-using-2-relays.gif

Diodes as in post #13, with back emf suppresion too.
 
Last edited:

Thread Starter

Vonasi

Joined Aug 19, 2018
34
Simple, the use the two wires to feed your relays with series diodes, so that only one relay operates in each direction, and use the contacts to reverse your motor.


View attachment 158360

As in post #13.
Which was my very own solution to my own problem! However I think that DNA Robotics solution in post #11 is neater and cheaper.
Now I know it is possible I need the ram manufacturers to confirm that it is a possible use for their ram and then I need to empty lots of lockers and crawl inside and measure up.

Thank you to everyone for your help.
 

cork_ie

Joined Oct 8, 2011
428
cork_ie - I am aware of all that you have commented on and I am in communication with the linear ram manufacturers regarding some of these issues. The ram I am considering is actually a combo hydraulic/electric ram (not a continuously running hydraulic pump but a reversible hydraulic pump) with overload valves built in to the hydraulic circuit so that it should not be possible to overload anything if the ram is installed correctly. If the maximum travel limit of the combo hydraulic ram is reached before the rudder hits its stops then there should be no possibilities of the ram causing damage.
As for speed of response the autopilot control head I am considering using has adjustable 'rudder response' so it should be possible to calibrate it for different reaction speeds needed.
At the moment I am just considering whether what I am thinking of doing is actually possible as the option of buying a full manufacturers kit of display, compass, computer, rudder angle sensor and ram is about 1/5th of the value of the boat whereas buying just a linear ram and a couple of relays is a third of that price. Why you can buy a powerful laptop for £250 but a simple autopilot and ram costs £2500 is beyond me but apparently all us sailors have money to burn.
As for speed of response the autopilot control head I am considering using has adjustable 'rudder response' so it should be possible to calibrate it for different reaction speeds needed.- This is exactly the issue I am trying to alert you to.The two important factors are "Yaw rate" and "Rudder response".
"Yaw rate" is basically adjustable deadband , i.e. how far the boat can veer off course before the pilot reacts.
"Response rate" is something completely different and is related to the activation of the ram on a particular boat to avoid oversteering and subsequent need for correction, continuously correcting to port and starboard . How much this happens depends very much on the quality of the compass, This is achieved in a normal direct drive autopilot by pulsing the output as the desired course is reached, to minimise oversteer and the need for correction. If you are just actuating relays then they draw very little current in comparison and will go hard on every time. I have spent 38 years at this stuff and have seen pilots from very early crude devices that just sucked battery power to really good modern units running on software, some, (believe it or not) with built in algorithms that learn about the characteristics of your boat and perform to suit and minimise battery consumption. What make of autopilot control are you using?
 
Top