OpenRex = Raspberry Pi + Arduino. Altium Files available for download.

[robertferanec]

We hope, you will like it

OpenRex really is an open source project and you can download all the files (Altium Schematic & PCB, Manufacturing data, ....).

We designed it for Playing, Learning and Hacking. It's based on NXP iMX6 (up to 1.2HGz CPU) + NXP LPC1345 (Microcontroller). We used "Arduino" like connector (so, you can easily use it with a breadboard) + there is also Raspberry Pi compatible header. Custom boards can be simply stacked on top of each other.

Here is the project website: http://www.imx6rex.com/open-rex/

Enjoy!
- Robert

[HAL-42b]

The PCB design course based on the board is very interesting.

[Gixy]

But not free...

[robertferanec]

But not free...

Hello Gixy. I completely understand your point. There are a lot of free things today, so why not these courses ....

The courses can not be free. It takes a lot of time and resources to create them. It may not look like that, but for example Advanced PCB Layout course took over 1 year to prepare. You need to include:
- Martin's salary during the time when he was designing boards for the courses. You need to design your own boards, because you need some material for teaching and also, you need to have something what people use for practicing. You can not use other people's / company's projects because of intellectual property. Designing the boards took approximately 1 year.
- Then, you really would like to teach about something what works perfectly and reliably. You need to build & test the boards. The cost: over 10 000 USD (to manufacture PCBs, to buy components, assemble the boards, testing them, ... )
- And, you still do not have the course. Add 3 months of work just to record the videos. I have recorded over 10 000 files which needed to be sorted out and connected together.

To create the courses is really a lot of work. If the courses are free, I would not do it.

There is another important reason why courses are not free. People will learn much more if they pay for it. If it's free, they skip a lot of things and watch only the parts they think are important or interesting - that completely misses the point why we are doing it. We do it, so people learn from it. If they pay for the course, most of the people go through every minute of the videos and try to learn as much as possible.

We do not keep the money for ourselves. We give them back to community. Everybody who signed up for the courses, supported development of new courses and also supported development of Open Source boards such iMX6 Rex module + Baseboard, OpenRex, 28Pins, ...

The other words, we could never design a board such OpenRex and make it Open Source without the help from people who signed up for the courses. Thank you.

PS: Thank you HAL-42b

[HAL-42b]

I understand the above point and I have huge respect for what you do. There are a lot of boards out there but you have something that sets you apart.

On the other hand offering the courses for free might increase the demand for the board considerably. In the end I think whatever creates most publicity would be the best return on your investment. That publicity will have a positive effect on your other products as well.

Also you need applications for your board. Not everyone is a hacker so you need to create a reason for non-hakers to buy your board. For this you need to apply your board towards the solution of some common problem so even people with no electronics and programming experience would want to buy it just for that application. Preferably you need several of these applications.

Think of it as Lego. Dumping a pile of Lego bricks in front of a child does not do very much if the child is not already familiar with the product. Putting a picture of a space ship or an excavator on the box gives kids ideas and something to aspire to, and this is what sells Lego bricks. So you need a killer app, a poster child, which only your board can do.

[robertferanec]

Think of it as Lego. Dumping a pile of Lego bricks in front of a child does not do very much if the child is not already familiar with the product. Putting a picture of a space ship or an excavator on the box gives kids ideas and something to aspire to, and this is what sells Lego bricks. So you need a killer app, a poster child, which only your board can do.

- Nice!

What do you think we should do as the next step with the board? I know we need some examples, what would you find most useful?

[HAL-42b]

There are several classes of things that have been done, Radio and SDR, Robotics, Test and Measurement, 3D printing and machining, Physics...  All of these are extraordinary geeky ant therefore have narrow audiences.

Tesla500 in Canada designed his own high speed camera, there are a few folks in Austria working on Appertus open source cinema camera as well.

There is the whole field of audio and lighting. Something that works with Linux and Jack would be useful.

Copenhagen Suborbitals are trying to launch an Arduino into space.

A wifi router with a steerable antenna might be useful to the mesh networking guys.

An open source magnetic tape management robot might be useful for all sorts of cheap online storage. The librarians and (hackers alike) around the world would love you.

[Kilrah]

- Martin's salary during the time when he was designing boards for the courses. You need to design your own boards, because you need some material for teaching and also, you need to have something what people use for practicing. You can not use other people's / company's projects because of intellectual property. Designing the boards took approximately 1 year.
- Then, you really would like to teach about something what works perfectly and reliably. You need to build & test the boards. The cost: over 10 000 USD (to manufacture PCBs, to buy components, assemble the boards, testing them, ... )

To me all of these should be factored in the price of the board, not of the course. You didn't design the board only for the course but to sell it as a product, then it also happens to benefit the creation of the course but that's a "nice extra".

- And, you still do not have the course. Add 3 months of work just to record the videos. I have recorded over 10 000 files which needed to be sorted out and connected together.

That's what the course price should reflect.

So instead of asking $250 for the course and $40 for the board, ask $50 for the board and $50 for the course (knowing this kind of board already exists and sells on its own e.g. udoo neo, so you'll likely sell many more boards to hackers/tinkerers who have no intention to ever design advanced PCBs themselves than courses).

Only an opinion of course.

[robertferanec]

Wow! Where you gets all the ideas?

I am thinking to create a special driver on the iMX6 CPU which will give possibilities to microcontroller use interfaces such HDMI monitor, Audio, big RAM & FLASH memory .... That would enable to write simple code on microcontroller, but with things what normally you can not do. For example, you could simply program the old retro games (like Atari, Commodore, Spectrum, ..... .. and it could be useful in some applications .. like oscilloscope.

[HAL-42b]

Wow! Where you gets all the ideas?

I am developing something that all of these ideas lack and sorely need

Selling generic boards is easy. Once somebody buys the board there is a definite limit in complexity that can be reached. I want people to be able to go well beyond the existing level of complexity and I am developing something that should help with that (I hope).

[robertferanec]

@Kilrah is it an Sbach plane?

I am not really sure if we can compete with big companies on selling boards. I am not even sure if people will like the board

[robertferanec]

I want people to be able to go well beyond the existing level of complexity and I am developing something that should help with that (I hope).

Are you developing HW or SW?

[HAL-42b]

Hardware. Mechanics to be specific.

[Kilrah]

@Kilrah is it an Sbach plane?

Yup, the tiny Eflite UMX one, with added custom micro FPV setup w/headtracking

[robertferanec]

Yup, the tiny Eflite UMX one, with added custom micro FPV setup w/headtracking

Awesome! I love the camera!
You must be really good in flying it. I have exactly the same plane - but I have crashed it every time (I have landed only once )

[robertferanec]

Hardware. Mechanics to be specific.

Interesting. I would say you do software. I am now curious how you are going to solve "go well beyond the existing level of complexity and I am developing something that should help with that".

[HAL-42b]

Think of how the 19" rack is ubiquitous in the industry but it never got applied to PCs because the Wintel alliance prevented it by coming up with their own proprietary form factor and bus.

This must be prevented from happening to Open Source. So we need form factors and backplane architectures which are Open Source and can't be walled in.

By defining a common base for all open source boards like yours we can make several boards work side by side on a common architecture. This would let us increase the complexity of possible projects by orders of magnitude.

It would also let us prevent planned obsolescence to a significant extent. We will be able to rely on re-usability of components once again.

I could write pages but that would hijack the thread.

[Kilrah]

Awesome! I love the camera!
You must be really good in flying it. I have exactly the same plane - but I have crashed it every time (I have landed only once )

Heh it's actually quite a nice tame plane as long as wind is calm - not a beginner thing by any stretch though.

That's what I see from the camera when flying (the recording is shitty though):

By defining a common base for all open source boards like yours we can make several boards work side by side on a common architecture.

Isn't there a thing called Ethernet that's already quite appropriate to allow cooperation of multiple boards side by side with the advantage of not imposing a common architecture?

I see little interest in defining a standard architecture in the open source world - it will probably serve a couple of projects but to me most people designing such open source boards have an idea in mind with some requirements, and they will just want to do it how they'd like it to be and precisely avoid adhering to someone else's design reqs. That leads to the current diversity of SBCs which I don't think is bad, and IMO if they had constraints to follow most of those who made some wouldn't have bothered.
Most boards already have some commonality, as they usually share either the Arduino interface or the RPi one, or both.

Think of how the 19" rack is ubiquitous in the industry but it never got applied to PCs because the Wintel alliance prevented it by coming up with their own proprietary form factor and bus.

It never got applied to PCs because it was not fit for the purpose while the AT* formats were...
Who in a home/office environment would want to have to install a rack to mount their PC in?!

[robertferanec]

That's what I see from the camera when flying (the recording is shitty though):

- Nice video, great mountains

Isn't there a thing called Ethernet that's already quite appropriate to allow cooperation of multiple boards side by side with the advantage of not imposing a common architecture?

- In many cases, Ethernet only may not be enough. Especially, if technology gets obsolete and your board is mounted in a system, ideally you would like to get a 1 to 1 replacement. This means same size, same connector placement, same pinout, same power, ....

I see little interest in defining a standard architecture in the open source world - it will probably serve a couple of projects but to me most people designing such open source boards have an idea in mind with some requirements, and they will just want to do it how they'd like it to be and precisely avoid adhering to someone else's design reqs. That leads to the current diversity of SBCs which I don't think is bad, and IMO if they had constraints to follow most of those who made some wouldn't have bothered.
Most boards already have some commonality, as they usually share either the Arduino interface or the RPi one, or both.

- That really is very well written and it's true not only for open source, it's true for commercial boards too.

By defining a common base for all open source boards like yours we can make several boards work side by side on a common architecture. This would let us increase the complexity of possible projects by orders of magnitude.

It would also let us prevent planned obsolescence to a significant extent. We will be able to rely on re-usability of components once again.

- What do you think the "common base" should be like?

To define a standard may have at least two quite hard problems to solve: how to make many people to agree on one standard and how to tell people to use it in their designs. What I see, many standards are defined by big corporations which don't want to fight between them, so they agree on same way to do it .... or the other way how standards are created is if you sell something what is wide spread and other companies intentionally design their boards the way to stay compatible.

[HAL-42b]

Isn't there a thing called Ethernet that's already quite appropriate to allow cooperation of multiple boards side by side with the advantage of not imposing a common architecture?

Ethernet is only for communication between two boxes with some meters between them. There are also simple board to board signals within the same box and different formats for chip to chip communication.

There are also other things like power and structural and cooling requirements and a plethora of other things.

In that respect my work is definitely not about exclusion of some existing standard or limiting the possibilities for the designer. On the contrary, maximum freedom must be given to the designer but how can we give maximum freedom to the designer and still have a standard? Seems like a conflict but it isn't really.

I don't like to talk airy fairy about things without being able to present anything so this would have to wait until it is ready for the public.

It never got applied to PCs because it was not fit for the purpose while the AT* formats were...
Who in a home/office environment would want to have to install a rack to mount their PC in?!

The PC/AT form factor was almost but not quite like a 4U 19" sub-chassis turned on its side. The incompatibility was most likely intentional and made wintel a whole lot of money by segmenting the market into upper and lower tiers.

[HAL-42b]

- What do you think the "common base" should be like?

To define a standard may have at least two quite hard problems to solve: how to make many people to agree on one standard and how to tell people to use it in their designs. What I see, many standards are defined by big corporations which don't want to fight between them, so they agree on same way to do it .... or the other way how standards are created is if you sell something what is wide spread and other companies intentionally design their boards the way to stay compatible.

The common base should enable maximum configurability for the designer, who can be also the end user. It should enable that during the design or later when the gadget is in use. Normally we design by imposing constraints. Learning to design by leaving maximum choice down the line was quite an experience for me.

It should be cheap and easy to manufacture without relying on a particular manufacturer. It should even be possible to produce using different manufacturing processes.

It should be robust and long lasting. It should wear in, not wear out.

It should be easily modifiable using simple tools to do exactly what the end user wants.

It should satisfy best mechanical and electronic engineering practices.

It should be compatible with the maximum number of existing standards. It should definitely not leave out any existing Open Source work.

It should be future proof. Things change fast.

It should be devoid of all aesthetic elements. Those should be left to the designer and the end user.

And you are right, it is definitely not easy to satisfy this many requirements at once. It is also hard to not impose any limitations down the line.

How will I make people use it? I won't. If it is good enough I shouldn't have to, people will use it anyway, if not it will fade away.

[Kilrah]

- That really is very well written and it's true not only for open source, it's true for commercial boards too.

Correct, the reasons are just different.

For the open source/hobbyist it is "I have an idea and I want to do it the way I think because that's where the fun/interest/learning is"
For the commercial market it's "if I don't need X I don't want it there because it would be a waste of money to include it."

how can we give maximum freedom to the designer and still have a standard? Seems like a conflict but it isn't really.

That's the key, you'd rather want a horde of standards and allow people to pick those that he needs and connect them together. Oh wait, that's what we've been doing for decades...

Ethernet is only for communication between two boxes with some meters between them. There are also simple board to board signals within the same box and different formats for chip to chip communication.

We've got widely supported standards for all of those already...

A standard encompassing everything sounds a bit like the proverbial "jack of all trades that does nothing well".

Would be interesting to see what you can come up with though.

[robertferanec]

How will I make people use it? I won't. If it is good enough I shouldn't have to, people will use it anyway, if not it will fade away.

Even a very good stuff often needs a push (good presentation, publicity, media coverage, credibility, ...). Otherwise people wont use it or simply, don't know about it.