Stepping Out, Spencer Saraband, 2389

[Island Time 1,220]

36 minutes ago, aardvarkash10 said:

Follow up on the auto tiller. It's now fitted and operational.  Nothing exciting technology wise, just a standard TP32.  However, there was earlier discussion around the tiller pin, so here it is.  Works great.

 

 

 

just remember that the TP32 max thrust is 85 Kg. The pin must support that, and the longer it is, the bigger the lever....

[aardvarkash10 950]

15 minutes ago, Island Time said:

just remember that the TP32 max thrust is 85 Kg. The pin must support that, and the longer it is, the bigger the lever....

The ram travel is not as long as the free swing of the tiller.  The pin is located about 1/3 of the distance from the tiller handgrip to the rudder stock, so 85kg is about the same as a 30kg effort at the grip - thats a lot of load!  Any time its in use is likely to be benign rather than "sporting" conditions, so its unlikely to be struggling to manage a wildly weather-helmed event.

I haven't done the math, but I in the event that it does exert maximum load, 85kg in shear would probably remove the head of the pin before anything else. 

Unless the tiller and rudder bind, I don't forsee an issue. 

 

 

[Island Time 1,220]

The correct position for mounting is 460mm forward of the rudder post center (pivot point).

The issues are power, motion range and speed. Mounting the pin 460mm forward of the pivot gives the best performance of these 3 requirements.

85kg on the standard pin is fine, but with one of extended length the torque on the tiller can be quote a bit... and, set up correctly + with the right sail combination, they can steer pretty well.

The single best improvement you can make is to add a decent compass - like a precision 9 - the difference in steering accuracy is HUGE. This is due to the fact that the little fluxgate in the tiller pilot is simply too small to give accuraqte headings in a seaway.  

[aardvarkash10 950]

Been a while between drinks.  

I posted in the weekend thread about a trip to Coromandel that started with the raw water pump impeller self destructing.

Here's the replacement.  Great price, ordered 03/03, arrived a week later.

Looks fine, if you hear nothing more you can assume it's lasting ok.

The tiller pilot has been working fine now for a year.  No issues with the mounting and leverage, but tracking is a bit random.  It's going to get connected to the OpenCPN plotter this winter so that should resolve this minor issue.

[TimeOut 2]

Just came across this thread as I have been running a Raspberry Pi, OpenCPN based chartplotter solution for the last year on my Young 88 and have a couple of issues so thought I'd reach out and share my experience. 

Firstly great job on your build aardvarkash10, you say you're not a geek but you've done a great job adding all those hardware components like UPS and power converters. I am a geek in that software and Linux is familiear teritory. In saying that my aim was to keep the build as simple as possible and do as little coding as possible which I've pretty much achieved. 

My build consists of a Raspberry Pi4, Openplotter/Opencpn, 32GB Micro SD (and yes I have 2, 1 is a backup). I use a Pican- CAN-Bus HAT (Hardware Attached Top) connected to the Pi which supplies power from and connectivity to my Garmin NMEA2k devices (speed,depth and wind) over a single cable. A GPS dongle (GlobalSat bU-353-S4) connects to the Pi via USB. My monitor was bought directly from Sihovision based in China, it's a 10” touchscreen,  IP67 waterproof and sits on an extension arm so I can move it from the cabin into the companion way. This was the expensive bit at about $800 and the motherboard died in the first 12 months! Fortunately under warranty, they sent a new board and I replaced it. I use a wireless keyboard to do any config and I power down the pi gracefully before turning it off at the instrument panel. 

The issues I have are

- touchscreen chart navigation is a little sluggish (I use ENC charts from LINZ), there's a 2 second delay when resizing or moving a chart, not a big deal i can live with this but it's annoying

- the monitor turns off if I have the engine running, so I think this is a power overload issue though Sihovision say it should handle anywhere from 9v to 32v, I'm thinking of adding a regulator - if that's the right thing?

- the GPS stops working randomly, it can go for hours without an issue and then die every 15 mins or so - unplugging the USB restarts it...this is a pain, i've moved the receiver as far away from the pi as possible thinking it might be interference but same issue, more research required...

Anyone have similar challenges?

Cheers

 

 

 

 

[aardvarkash10 950]

Hi TimeOut thanks for dropping in.

The build has taught me  lot, but Im a bit of an old dog and new tricks are getting tougher.

The build is essentially as described earlier.  I am considering V2.01 which would be based on a 4B but with 8gb, not the 2gb onboard the present CPU.  I'd also go full into the new (as in, not released yet) McArthur HAT to have everything under one roof so to speak.  It's likely this will happen in a different yacht (Mrs Aardvark wants more elbow room) so not soon.

I made a conscious decision to run very simple so no touch screen, and the keyboard and mouse are USB connected, not wifi.  The screen is huge - a 22 or 24 inch Thomson 12-volt unit with built in DVD player and a range of inputs.  It's on an articulated arm and swings almost into the companionway but easy to see from the helm anyway.  And we can play DVDs and Netflix on it on crappy days.

The only thing I will add to this build is an interface to the TP32 auto helm so we can have more robust route following.

We still have screen delay similar to yours - I am pretty sure that 8gb onboard the 4B will improve that.  I think its openCPN re-rendering the map more than it is video performance or touch response lag.  Yeah, we get the vector map updates from LINZ.  It's a bit of a faff but bearable.

My build is sensitive to voltage spiking.  Even with the UPS, the CPU will sometimes shut down when the engine is cranked over (sudden voltage drop and inductive load combined with cheap voltage droppers).  I haven't had the monitor shut down issue you describe.  IMO you can never have a power supply too stiff, so I would vote for regulation in your case.

I haven't had gps issues EXCEPT as a symptom of sd card corruption.  Like you, I have two 32gb cards onboard, one in the card slot and running, the other is in a USB adapter and has a full backup image.  I do the back up semi religiously (as in, I pray I got it right) every couple of days.  If I corrupt the working card, its simple to swap them over, and then format the corrupted card and run the backups on that one until the next hiccup. 

We run super cheap wifi by running a cheap cellphone up the pennant halyard with its hotspot on.  That's plenty for us in most locations and allows Internet access for the Pi if needed.

Nice to have an option for help with any coding - beware, I WILL track you down!

Cheers

Ashton

[Brien 22]

I commissioned (fitted) a Raymarine element 9HV recently which dropped its position randomly but quite frequently.

We resolved by reducing the resolution of its position output. It was a setting well hidden and I am not sure if it looked at less satellites or used a less intensive algorithm but this worked and may give you a lead.  

Had earlier tried to reducing the plot to 5 second separation and problem remained, after reducing resolution it now also accepts 1 second plot separation happily. 

Brien

[TimeOut 2]

Ashton,

I agree the McArthur HAT looks good as it solves the pi shutdown challenge and handles N2k connectivity and the Pi5 looks good too, better performance and it takes the NVME (SSD) hat - definately looks worth investigating. 

We raced round rangi and rakino yesterday (sunday) in the RYC closing regatta, very challenging conditions, a broach and and knockdown in those savage squalls and our no2 gave up and disintegrated heading back up the harbour to westhaven, we crossed the finish line on main only and got first on line!

Got a nice graph from a fellow competitor showing peak of 40 and avg 23knots, which made me think I should start capturing data off the pi - next project. And no problems with the GPS over the entire 6 hours.

Sure thing on the coding and I'm familiar with Linux if that's ever any help.

Jeremy

 

[CarpeDiem 430]

22 hours ago, TimeOut said:

Ashton,

I agree the McArthur HAT looks good as it solves the pi shutdown challenge and handles N2k connectivity and the Pi5 looks good too, better performance and it takes the NVME (SSD) hat - definately looks worth investigating. 

We raced round rangi and rakino yesterday (sunday) in the RYC closing regatta, very challenging conditions, a broach and and knockdown in those savage squalls and our no2 gave up and disintegrated heading back up the harbour to westhaven, we crossed the finish line on main only and got first on line!

Got a nice graph from a fellow competitor showing peak of 40 and avg 23knots, which made me think I should start capturing data off the pi - next project. And no problems with the GPS over the entire 6 hours.

Sure thing on the coding and I'm familiar with Linux if that's ever any help.

Jeremy

 

I have a very nice little unit from YachtD that captures everything on the N2K network and records it to a sdcard.

Unfortunately I now have billions of datapoints collected over 5+ years that I have been unable to process because the amount of information is over welming - but one day I plan on learning/using some big data processing tool to overcome that

In the meantime I keep recording everything whenever the network is powered on...

[aardvarkash10 950]

Loads of data, no information.

Sounds like my workplace