Finally, I'm back on my feet!
For those who didn't read it:
I've catched a flu and wasn't really able to do a lot of work or write lenghty posts last week, but now I've fully recovered!
So MORE than time to do a news post, as some important things have been going on.
1. Prototype Units
Thanks to the flu, I was only able to build one prototype so far, and that one has been sent to zmatt (the one who fixed the TILER for rotation).
He has already received it and is already poking around on it for a few days.
With good success! Thanks to some help from him, Nikolaus could boot from the internal MicroSD card as well, which was one of the last tests we were still missing.
Right now, he's working on improving U-Boot even further (with patches specifically for our OMAP5) as well as getting a good stable LPAE kernel running.
aTc is helping him as good as possible... but as he has an older prototype that's especially picky about whether it wants to boot or not, it's not that easy (he'll receive a new one soon though).
I'll try to build more prototypes as soon as possible (though I've still got plenty of work to do), so I can send aTc, JohnnyH, notaz and Andrej a unit as well (and maybe some more devs).
Let the fun begin!
Once we have the kernel and OS running well enough for "almost" normal usage, I'll start sending the prototypes to the prototype orders as well as auction some more here at the boards.
2. The CPU Boards
Nikolaus finished the CPU board layout with the help of Julianos simulations.
We've checked and compared them with the actual stackup from the manufacturer and have now achieved the best possible results.
While it's not 100% according to TIs suggestions, it should easily be good enough to run 1,5GHz.
But now here is the important message for you:
600 CPU boards have now been ordered.
Yes, there's not another prototype run, we've placed the first mass production order!
Why only 600?
Well, we're building the units in batches of 500 (and while we're assembling and shipping, the next 500 units are being produced, so there shouldn't be a huge gap).
And in case there's REALLY something wrong with the CPU Board layout (the chance for that is REALLY really low, otherwise I would've decided to do another prototype run), then 600 non-usable boards would mean a loss of 10000 EUR.
And even less, in case we sell them as collectors item for 17 EUR each
)
The more likely case is that we save at least 7000 EUR because we're not doing another prototype run - so that's a good option.
3. The Mainboard
Nikolaus found another small fix for the mainboard which will be manually fixed on the prototype boards but of course will be fixed in the layout for the mass production.
Right now, Nikolaus and Juliano are working on Pyras Wifi. While the normal Wifi works about as good as Pandoras Wifi right now, 5GHz Wifi has a really bad reception.
Luckily, Juliano can simulate that as well, so we've already applied quite a few fixes to the antenna circuit already.
This will also improve normal Wifi a lot. These are the same tools used to simulate antennas in normal commercial devices, so we're having an advantage thanks to that (compared to Pandoras Wifi).
It can't be fixed for the prototype boards sadly, but they're working well enough for normal usage unless you try to use 5GHz Wifi
We're doing a small testing mainboard run for that (only partly populated, can be done for 200 EUR) and once that has been confirmed to be working fine, it's time to order the mainboards as well.
Remember: The CPU Boards are more complex and take more time to produce than the mainboards, do it was important to get the CPU boards produced first.
4. The keymat
After testing and playing around with the keymat a bit more, me (and the austrian team working on it) found out the following things:
So everything is correct physically - but the material and color needs to be improved a bit.
We're almost done here as well!
5. The case
Still waiting for the next versions, sadly.
But as the last one was almost perfect (except for some minor issues), I'm not too worried here.
Unlike the PCB production, the cases can be produced within a very short time once they're 100% finished.
I can't wait to receive the next batch though. They won't be transparent anymore, but really dark with a VERY slight tint (which is almost invisible).
Well... that's it for todays news.
I hope you liked them. And I'll let you know once I've got more news
For those who didn't read it:
I've catched a flu and wasn't really able to do a lot of work or write lenghty posts last week, but now I've fully recovered!
So MORE than time to do a news post, as some important things have been going on.
1. Prototype Units
Thanks to the flu, I was only able to build one prototype so far, and that one has been sent to zmatt (the one who fixed the TILER for rotation).
He has already received it and is already poking around on it for a few days.
With good success! Thanks to some help from him, Nikolaus could boot from the internal MicroSD card as well, which was one of the last tests we were still missing.
Right now, he's working on improving U-Boot even further (with patches specifically for our OMAP5) as well as getting a good stable LPAE kernel running.
aTc is helping him as good as possible... but as he has an older prototype that's especially picky about whether it wants to boot or not, it's not that easy (he'll receive a new one soon though).
I'll try to build more prototypes as soon as possible (though I've still got plenty of work to do), so I can send aTc, JohnnyH, notaz and Andrej a unit as well (and maybe some more devs).
Let the fun begin!
Once we have the kernel and OS running well enough for "almost" normal usage, I'll start sending the prototypes to the prototype orders as well as auction some more here at the boards.
2. The CPU Boards
Nikolaus finished the CPU board layout with the help of Julianos simulations.
We've checked and compared them with the actual stackup from the manufacturer and have now achieved the best possible results.
While it's not 100% according to TIs suggestions, it should easily be good enough to run 1,5GHz.
But now here is the important message for you:
600 CPU boards have now been ordered.
Yes, there's not another prototype run, we've placed the first mass production order!
Why only 600?
Well, we're building the units in batches of 500 (and while we're assembling and shipping, the next 500 units are being produced, so there shouldn't be a huge gap).
And in case there's REALLY something wrong with the CPU Board layout (the chance for that is REALLY really low, otherwise I would've decided to do another prototype run), then 600 non-usable boards would mean a loss of 10000 EUR.
And even less, in case we sell them as collectors item for 17 EUR each
)The more likely case is that we save at least 7000 EUR because we're not doing another prototype run - so that's a good option.
3. The Mainboard
Nikolaus found another small fix for the mainboard which will be manually fixed on the prototype boards but of course will be fixed in the layout for the mass production.
Right now, Nikolaus and Juliano are working on Pyras Wifi. While the normal Wifi works about as good as Pandoras Wifi right now, 5GHz Wifi has a really bad reception.
Luckily, Juliano can simulate that as well, so we've already applied quite a few fixes to the antenna circuit already.
This will also improve normal Wifi a lot. These are the same tools used to simulate antennas in normal commercial devices, so we're having an advantage thanks to that (compared to Pandoras Wifi).
It can't be fixed for the prototype boards sadly, but they're working well enough for normal usage unless you try to use 5GHz Wifi
We're doing a small testing mainboard run for that (only partly populated, can be done for 200 EUR) and once that has been confirmed to be working fine, it's time to order the mainboards as well.
Remember: The CPU Boards are more complex and take more time to produce than the mainboards, do it was important to get the CPU boards produced first.
4. The keymat
After testing and playing around with the keymat a bit more, me (and the austrian team working on it) found out the following things:
- The height of the keys is fine. Nothing needs to be changed here.
- The ABXY-buttons feel almost identical to the one on the Pandora. They're not too hard to push and have a good tactile feedback.
- We need to use a different color for the keymat though, as it is a bit too thick, so it has a negative effect on the tactile feedback. The color can be completely left out on the DPad though.
- Another way to improve this is by using a different material (the austrian team is currently testing multiple ones).
So everything is correct physically - but the material and color needs to be improved a bit.
We're almost done here as well!
5. The case
Still waiting for the next versions, sadly.
But as the last one was almost perfect (except for some minor issues), I'm not too worried here.
Unlike the PCB production, the cases can be produced within a very short time once they're 100% finished.
I can't wait to receive the next batch though. They won't be transparent anymore, but really dark with a VERY slight tint (which is almost invisible).
Well... that's it for todays news.
I hope you liked them. And I'll let you know once I've got more news
begin..
