12vdc Marine Computers

[rgleason]

I suppose one could take the wood model and expoxy it. Then use marine sealants. Might be an approach for more 1 off wood kits the user completes?

[mvandervoort]

Originally Posted by rgleason

I suppose one could take the wood model and expoxy it. Then use marine sealants. Might be an approach for more 1 off wood kits the user completes?

Sure, this approach is very well suitable for a one off. There are a lot of plywood epoxied sailing boats out there! So for low budget DIY, this is one way to go. In the current design there is one weakness, and that are the 2 mini switches. I haven't been able to find good and small IP65 ones right now. I'm sure they are available, however not found yet.


On this image you'll look at MIDC on my boat just captured this weekend. You may notice that this display doesn't have an AR surface. So lots of reflection!

With laser cutting plywood, you have to be sure that the thickness of the plywood sheets are precise. The thickness is part of the dimensional design and they must be 3 and 6 mm (+/- 5%). Low quality birch plywood generally shows big tolerances. Ideally you could pick your own exterior quality plywood and visit a Fablab with a laser cutter, or offering this as a service. Laser cutting services are generally available in the modern world, however some of them aren't capable of cutting (ply)wood. On Hackster.io, you'll find the national cutting service provider I use here in The Netherlands.

The other ideal one off DIY building approach is 3D Printing on which we have talked in earlier posts quite extensively.

So to conclude, I'll offering two shared source designs for DIY/Makers:

1. The laser cutting one for building up the display enclosure with plywood (max. €50 budget for high quality results)
2. The STL 3D model one for 3D Printing the body and lid part (min. €275 budget for high quality pro grade 3D printing services)

The 3rd one will be the vacuum / resin casting one, which might become part of a commercial release of the MIDC. Targetted budget somewhere around €150 per enclosure for a 10' display in small volumes.

[mvandervoort]

Originally Posted by rgleason

I suppose one could take the wood model and expoxy it. Then use marine sealants. Might be an approach for more 1 off wood kits the user completes?

Sure, this approach is very well suitable for a one off. There are a lot of plywood expoxied sailing boats out there! So for low budget DIY, this is one way to go. In the current design there is one weakness, and that are the 2 mini switches. I haven't been able to find good and small IP65 ones right now. I'm sure they are available, however not found yet.


On this image you'll look at MIDC on my boat just captured this weekend. You may notice that this display doesn't have a AR surface. So lots of reflection!

With laser cutting plywood, you have to be sure that the thickness of the plywood sheets are precise. The thickness is part of the dimensional design and they must be 3 and 6 mm (+/- 5%). Low quality birch plywood generally shows big tolerances. Ideally you could pick your own exterior quality plywood and visit a Fablab with a laser cutter, or offering this as a service. Laser cutting services are generally available in the modern world, however some of them aren't capable of cutting (ply)wood. On Hackster.io, you'll find the national cutting service provider I use here in The Netherlands.

The other ideal one off DIY building approach is 3D Printing on which we have talked in earlier posts quite extensively.

So to conclude, I'll offering two shared source designs for DIY/Makers:

1. The laser cutting one for building up the display enclosure with plywood (max. €50 budget for high quality results)
2. The STL 3D model one for 3D Printing the body and lid part (min. €275 budget for high quality pro grade 3D printing services)

The 3rd one will be the vacuum / resin casting one, which might become part of a commercial release of the MIDC. Targetted budget somewhere around €150 per enclosure for a 10' display in small volumes.

[Sailoog]

Here another initiative:

There are people who buy boats but there are also people who build them, why not build your own electronics too? OpenPlotter is a combination of software and hardware to be used as navigational aid on small and medium boats. It works on ARM computers like the Raspberry Pi and is open-source, low-cost and low-consumption. Its design is modular, so you just have to implement what your boat needs. Do it yourself.

http://www.cruisersforum.com/forums/...ct-152754.html

[mvandervoort]

@rgleason

3rd Q&A post on my MIDC project.

How much display heat is there expected and will it need to be dissipated?

Good question over again. The current LG display I'm using for demo & test purposes, is really a low power one. In the specs there is typical power consumpion of 2,5W. So in the current enclosure I don't expect heat dissipations problems with overheating as a consequence. With the foreseen sunlight readble screens, there is another situation. Especially when the size of the displays increases. The typical power of the expected 10' display is about 8W. Combined with the IP65/67 ambition this will give me a tough challenge. Nevertheless, I do think I have a really dead simple solution for that as well. Before I will publish about it, I would be happy to verifiy that this provisioning will work on the different screen sizes, the different cockpit mounting situations and the way cables will brought from instruments towards the cabin/underdeck. I know my own boat very well, however that's not the case for the bigger ones.

The only thing I would like to tell that this design aspect is part of my design scope becasue overheating a serious problem to prevent.

Is the enclosure for the Arduino UDOO NEO going to be sealed I hope?

Not now. I don't see a good reason for it, because that part of the MIDC can and will be mounted in the cabin. The reason that the cabin is the logical place to mount this device is the possibility to run Android on it (with a SD card change) and to connect a cabin screen alongside the cockpit screen! So with my MIDC you'll get both a dual OS and a dual screen solution. That's quite cool, isn't it?

So for now I'm saying no. However if there is a very good reason to do so, I'll rethink my current decision. It will enlarge the current engineering challenge specifically for all the connectors on UDOO. So the reason must be very, very good to change my mind

How much programming and electronic work is going to be needed?

No programming at all. At most installation and configuration. The whole idea will be targeted on people like me who can't program software. That's the reason I'm strongly looking to OpenCPN, SignalK, Open Plotter and Open Sailing Processor and willing to cooperate with them. It's the freedom of the end user to choose whatever he or she wants to use as the most beloved maritime application.
MIDC will only be the low power high performance maritime platform which might be build yourself, delivered as a DIY kit or as an end user product ready to run with commercial support.

So, rgleason; this was the last part of 3 posts with a very interesting set of questions. Hopefully this show the ambition of MIDC and what might be expected and what not. If there are any questions feel free to post them here. I helps me design a attractive and valuable product not only for me personally, but for all interoperable and open source minded boaters out there.

It's in my timeline to publish the first and integral MIDC, somewehere in autumn. In the mean time I'll publish the progress on Hackets.io, with some cross postings here.

[rgleason]

Thanks mvandervoort. Good postings to read.

[mvandervoort]

You're welcome. Thanks for the compliment and I'm happy you like the content of my postings.

[mvandervoort]

After months of waiting for both the UDOO Neo SBC board to arrive, plus time to work further on the MIDC project, I'm happy to announce good progress.

In short: the bare basics of OpenCPN, NMEA 0183 sensor, AIS receiver and a non touch display running all on Lubuntu functions.

Have a look a the Hackster project log for more information and some photo's.

Obviously there are some issues to resolve. In 2016, I'll catch up and move on to the more difficult part: sunlight readable touch display on Linux and some wireless addons. Plus a waterproof enclosure. Design is already done.

To be continued.

[sailnow2011]

MSI Cubi 018BUS - Core I5 5200U - 0 MB - 0 GB - CUBI-018BUS - White | Staples®

external adapter means you could get a 12v adapter for it.

[bfloyd4445]

Originally Posted by sailnow2011

MSI Cubi 018BUS - Core I5 5200U - 0 MB - 0 GB - CUBI-018BUS - White | Staples®

external adapter means you could get a 12v adapter for it.

pretty soon we will be able to have a do everything computer wirelessly connected to displays and devices in our pockets. The Lenovo win 10 stick at $50 or the Chromebit at $100 can almost do that now. I like the win 10 stick best. Thgese chips plug into an hdmi port and can bluetooth keyboard and mouse plus wi fi

[seandepagnier]

1. The laser cutting one for building up the display enclosure with plywood (max. €50 budget for high quality results)
2. The STL 3D model one for 3D Printing the body and lid part (min. €275 budget for high quality pro grade 3D printing services)

The 3rd one will be the vacuum / resin casting one, which might become part of a commercial release of the MIDC. Targetted budget somewhere around €150 per enclosure for a 10' display in small volumes.

I don't understand why you can't take a $20 hdmi display and replace the backlight with 10x the amount of leds all mounted on heat syncs. Then fit the screen into an aluminum case for heat. I found that the cheapest displays don't have the most efficient drivers so this wastes 1-2 extra watts which is a lot.

You can buy a 3d printer for a few hundred dollars, so I would imagine that using one to make enclosures for various small computer boards could cost no more than a few dollars.

As for the actual devices... how well do they run opencpn. In theory the $5 raspberry pi zero should be good enough once I made a few more optimizations.

[gamayun]

Hey. First time on this thread so pardon me if this has been addressed earlier. I am planning to put a computer in the nav station instead of a chart plotter that will run OpenCPN and hook into my B&G radar. Thinking about a NUC and a monitor, maybe with a touch screen. Any thoughts on which NUC and monitor? How about memory and storage capacity? Anything I'm not asking about that I should? And if I'm in the wrong place (which certainly might be given the content of the last few prior postings), just kick me off...gently.

[bcn]

A NUC will just do it fine. OCPN is not resource hungry. For Radar perhaps not the most basic one but even this should work.

[mvandervoort]

Originally Posted by boat_alexandra

I don't understand why you can't take a $20 hdmi display and replace the backlight with 10x the amount of leds all mounted on heat syncs. Then fit the screen into an aluminum case for heat. I found that the cheapest displays don't have the most efficient drivers so this wastes 1-2 extra watts which is a lot.

You can buy a 3d printer for a few hundred dollars, so I would imagine that using one to make enclosures for various small computer boards could cost no more than a few dollars.

As for the actual devices... how well do they run opencpn. In theory the $5 raspberry pi zero should be good enough once I made a few more optimizations.

You can't buy a "standard" $20 HDMI display which also:

• has touch interface which can be used in wet marine environments, even to use with your gloves on
• is sunlight readable
• has AR/AR treatment
• is waterproof
• has audio
• operates from -20 / + 70 Celcius

Regarding 3D Printing.
I wonder if you've used such a $300 in practice. If so, you know the results can't be used on a boat

I'm developing a device which also is capable of OpenCPN using vector maps and will be capable of directly connecting NMEA200 devices. This requires a CAN bus interface. No RPI has them. UDOO has.

I'll appreciate all comment which have the intention to contribute to the design. It might be a good idea to read the extensive design documentation on https://www.hackster.io/mvandervoort...troller-2b3736

[sailnow2011]

Originally Posted by bfloyd4445

pretty soon we will be able to have a do everything computer wirelessly connected to displays and devices in our pockets. The Lenovo win 10 stick at $50 or the Chromebit at $100 can almost do that now. I like the win 10 stick best. Thgese chips plug into an hdmi port and can bluetooth keyboard and mouse plus wi fi

processing power is in this box, it is not in the others.