Cockpit Pilot SPX-5 and Hydraulic Pump Possible ?

[ghaffy]

Hello,

On our 25' displacement type boat (6 knots) there is a Raymarine SPX-5 Sport Pilot installed.
The boat is equipped with a hydraulic steering system.
I installed it last year, which was straight forward&easy, it works quite well, basically everything is OK.

I am sure you know the SPX-5 Sport Pilot. There is the the course computer with integrated gyro or so, there is a fluxgate compass, there is the control unit, and there is an electric motor that turns the wheel.
Additionally I do have a rudder position transducer that is not necessary but improves accuracy a lot, especially on slower boats.

Now, I'd like to replace the electric drive unit, which turns the wheel, by a hydraulic pump, as on a "normal" bigger autopilot system.

The Raymarine manuals say that the bigger (and of course more expensive) SPX-10 corepack is necessary when using a hydraulic pump.

I found out that the main restriction of the SPX-5 is that the output current of the corepack is limited to 10 Amps (for the electric helm drive).

I have an hydraulic pump that would be absolutely sufficent for the size of our boat and the size of the rudder. It is a 100W pump, thus would draw nominally some 8.5 Amps. The manufacturer told me, the draw under normal conditions is as low as 4 to 5 Amps.
So from that point of view the SPX-5 corepack should be able to handle this.

Of course, I also asked Raymarine. They said, I need the bigger corepack, because the SPX-5 software is not designed to handle a hydraulic drive.

Now, I wonder what this could mean: To my understanding the computer gets its information from the gyro, the fluxgate, and the rudder indicator, plus the GPS when turned on. How the hell should it know that the rudder is turned by an electric drive or by a hydraulic pump?

I am familiar with the needs of the hydraulic installation (check valve and so forth).
Apart from the higher output for the drive the "bigger" corepacks provide a "clutch" output, which is needed for disengaging a linear drive or in case a mechanic steering is installed, which is not in my case. I have a hydraulic steering, I don't need the clutch signal.

Am I overlooking anything important, or might I try to replace the electric helm drive by a hydraulic pump?
What are your opinions?

Regards
Andreas

[evm1024]

Well let's just dig this thread ouut....



The SPX-5 uses the same mosfets that the spx-10 does. This implies that the H-bridge on the spx-5 has the same potential drive capabilities as the spx-10. Differences are the installed fuse, the PCB trace size and heatsink size.

Another difference is that the spx-10 has a solenoid output that enables the pump oil circuit.

Looking at the schematics of the spx-5 and spx-10 I see that they have the same pinouts in the CPU with the exception that the spc-5 pulls up the outputs for the solenoid drive and solenoid over current (and other unused functions I'm sure).

The question is: Does the spx-5 software send a solenoid signal when in auto? When I get time I'll have to measure this.

I am curious to find out if one could drive a pump. Even if you had to manually activate the solenoid.

Any others have insight to this?

Regards

[ghaffy]

Originally Posted by evm1024

...
Any others have insight to this?

...

Yes.
Meanwhile, I did it.

I put the whole story on our blog Changing The Wheel Pilot Into A "Hydro-Pilot"

To cut it short: I removed the electric helm drive, added a check valve (because DIDO's helm pump had none), installed the electric hydraulic pump, connected the pump to the pins where the electric helm drive was connected to before (A+B at the course computer).

That was all.
No solenoid.

In our case neither a new auto-learn sequence nor the change of any of the other parameters (counter rudder etc.) was necessary.
The Amp draw of the pump and the electric helm drive are obviously quite similar.

The accuracy of the AP is still the same, absolutely no change in this respect.
However, we got rid of the noise, and the steering wheel may be turned again with one finger.

So we are happy (after 100 hrs / approx. 600 nm).

BR
Andreas

[evm1024]

Super. I did see your blog but did not make the connection that it was you. Outstanding achievement! As my grandfather used to say: "Damned fool did not know it was impossible so he went ahead and did it anyway".

The ram that I have does need a solenoid so that is something that I'll have to work on.

Just looking over the schematics briefly it appears that the spx-5 and spx-10 are basically the same. Having the same H-bridge mosfets they "should" have the same drive amperage. Needing a fuse change in the spx-5 to to realize.

[ghaffy]

Hi,
now as we are going more deeply into this matter, I wonder what would be the difference if I chose a different drive type in the setup.

The SPX5 manual says "choose drive type 3" without explaining what this means or whether / what other options exist.
So I looked up "drive types" in the SPX-10 manual. There one will find the following list:


Assuming that the software of the SPX-5 and the SPX-10 is similar, this list would suggest, that one chooses drive type #4 when an auto-reversing hydro pump is used instead of the electric helm pump (which is the case in my installation).

Certainly, I shall give it a try next time I am on the boat. But any ideas what the change in terms of behavior of the computer may be?
I could think of things like differences in initial current or some sort of a different voltage curve in order to accelerate either in "pump-mode" or "drive-mode"?

At the moment - as mentioned above - I have not changed any parameters, so I run the pump as a type-3-device instead of a type-4-device, which obviously would be more appropriate (if a SPX-10 computer was used). Still works.
Although one never should change a running system, I still wonder what will be the difference changing the drive type ...

BR Andreas

[evm1024]

I would assume that the code base for the SPX-5, SPX-10 and SPX-30 is the same. But there would be (compile options) some way to remove code that does not apply to the target pilot.

I am hoping that the spx5 code does have the clutch routines but that the clutch signal is just pulled up and the clutch driver circuit left off the circuit board. (I can build a clutch driver....)

I think that the differences in drive type takes into account the need for a clutch, the max current required by the drive and perhaps some slop in the drive system (Hysteresis, loose chain?).

In the wheel pilots (spx-5) the clutch is manual so there is not a clutch signal coming out of the corepack.

This works well with your system because (If I remember correctly) you have a hydraulic steering system. You turn the wheel and the pump attached to the wheel moves the hydraulic ram. The pilot drives the pilot pump which is attached in parallel to the steering pump and the ram moves the rudder. Thus no clutch is needed.

Type 3 drives (linear and rotary) need a clutch and are more limited in drive thrust. Being a motor/gear type system. The motors are reversing.

Type 4 drives are hydraulic with a reversing pump. They do not need a clutch for hydraulic steering and do need a clutch for mechanical steering. The pumps are of differing sizes matched to the ram oil displacement.

Type 5 drives are not reversing. They run constantly and use a hydraulic solenoids to divert oil to the the ram for in or out motion.

I notice that the H-Bridge circuit has both a pulse width modulation as well as direction inputs. Thus the corepack can select direction to move the rudder as well as (using PWM) the speed/force with which the rudder will move. I assume this is all part of the tuning parameters in the setup pages.

Also, the different corepacks have H-bridge current monitoring. This could be used to shutdown in a stall case or as feedback to allow modulation of the force of the ram.

Sorry to ramble a bit here.

I will at some point take a look at the SPX-5 I have and see if there is an actual clutch signal coming off the CPU. The clutch drive comes out of the CPU on pin 14 (TP793) and on the SPX-10/30 there is an additional test point (TP50) for the clutch drive.

Regards, evm

[ghaffy]

evm,

just came from the boat.
In the setup / dealer calibration of the SPX5 the option "drive type" is shown, however not accessible, i.e. "drive type 3" may not be altered.
But as mentioned before, everything is working fine anyway (at least in my case).

So I wonder whether youu'll get a solenoid signal, as this (to my understanding) would require to define the drive type as #5 (constantly running pump where the rudder movement is controlled by solenoid valves).
But as the manual is a bit unclear in this respect it might even be the case that with slightly different software versions the drive type may be changed. Otherwise the statement in the manual "drive type 3 must be chosen" hardly makes any sense.

I am curious how things develop in your case ...

BR Andreas

[evm1024]

Hi,

I'm thinking of using my SPX-5 with a reversing pump and a bypass valve. The SPX-5 may not have a bypass valve signal available. In that case one could manually engage the bypass valve much like the need to use the lever on the wheel drive.

I do expect to buy a SPX-30 at some point. One reason to look at having an SPX-5 driving a pump is to have a low cost backup for a more robust pilot.

Plus of course the DIY factor....

I expect that in software there is a bit of code that set what parameters are changable and what their value range is. It would be interesting to know if there was a jumper that set the pilot type for the software.

It has been a long time since I reversed engineered other peoples code. The pilot uses an ARM processor by the way ( STR912FA ).

[Cap Erict3]

I am taking the boat out today for sea trials on my wheel pilot to pump conversion. I found the wheel pilot noisy and undependable.

The pump is rated at 19 amps deadheaded so rather than risk overloading my course computer, I used directional diodes to feed external relays.

Easy to change out a plug in relay or fuse.

The dock tests worked well enough that I have already pulled the wheel pilot.

I used an Octopus piston pump so it is not whisper quiet. I tied the pump into my hard piped system with 2 3/8 and 1 1/4" 18" hoses made for me at Napa. It sounds about like my fresh water pump.


The pump did not require check valves and can closed off using built-in needle valves. The flow rate can be adjusted down but for now is set to full.

I have rudder feedback and changed the rudder limit in the software from 19 to 15 degrees.

Full rudder is around 25.


Sent from my iPad using Tapatalk

[Cap Erict3]

As a follow up, the steering performed fine. It was nice not to have the noise and a system that worked as expected.

The wind was piping up pretty good, so I did not go the the self calibration. It did not oversteer as the WP was prone to do.

Not having the wheel turning will take some getting used to but I will strive to persevere.



Sent from my iPad using Tapatalk

[ghaffy]

Originally Posted by Cap Erict3

As a follow up, the steering performed fine. It was nice not to have the noise and a system that worked as expected.

The wind was piping up pretty good, so I did not go the the self calibration. It did not oversteer as the WP was prone to do.

Not having the wheel turning will take some getting used to but I will strive to persevere.

Ha! Good to hear that meanwhile I am not the only one who has exchanged the drive by a pump.
If I understood Raymarine's latest ads correctly, they cheated here and offer the SPX5 as a "pump version", too.
Not the first time that CF turns out to be a development platform for manufacturers ...

The relays sound to be a very reasonable improvement.

[olaf hart]

So would the SPX-5 wheelpilot drive a type 1 linear drive direct, or would it read the power load as overload?
Is there an electronic version of the diode / relay option?
Would adding a rudder position indicator help?

[ghaffy]

Originally Posted by olaf hart

So would the SPX-5 wheelpilot drive a type 1 linear drive direct, or would it read the power load as overload?
Is there an electronic version of the diode / relay option?
Would adding a rudder position indicator help?

A collegue here was working on getting an SPX-5 course computer (built before 2014) and a linear drive work together. However, he did not succeed.

A linear drive type needs a clutch signal to disengage the drive when not in use. Otherwise you may not turn the wheel. When using a hydro pump this is done by the check valves, and the original drive may be turned anyway if not in use.

Not so a linear drive, he told me.

So from that point of view the SPX-5 may not be used together with a linear drive. There are no connectors for the clutch.
This is what I've been told.

[Neptune's Gear]

Using relays is fine, but fairly short term. About 1 mill cycles is common. Try a mosfet - a much more robust system with no moving parts or contacts. Think of it like a solid state relay... cheap and effective... and Olaf, you could use mosfets to control the type 1 linear drive, otherwise the AP computer will overload...
The clutch is just a switch, can be auto or manually controlled, depending on the sophistication of your circuit (with the mosfets).
Matt

[Cap Erict3]

Originally Posted by olaf hart

So would the SPX-5 wheelpilot drive a type 1 linear drive direct, or would it read the power load as overload?
Is there an electronic version of the diode / relay option?
Would adding a rudder position indicator help?

I preferred not to risk the course computer by increasing the load so did not try the pump directly to it.

The diodes were used to close one of 2 relays depending on direction signal from SPX5. There are solid state relays that will work. I was driving a borrowed truck and intent on finding all parts in one trip in a small town. I bought spare relays and diodes.

I added rudder feedback on my initial install. I consider it a "must have". You don't have to buy Raymarine. I didn't.


Sent from my iPad using Tapatalk