And here it is, as promised: The second part of my Pyra News post 
1. The Hardware
Well, we had quite a few issues to solve which slowed things down a bit, but it seems the solution has been found.
I'll try to explain what happened and what took us four weeks to solve.
As you know, we were trying to get the Solomon (Rotator) chip to work.
This is not an easy task - there are quite a lot of timing settings between the OMAP5, the Solomon Chip and the Panel Driver which need to be properly setup and which even affect each others.
Calculating proper, usable values alone is quite a time consuming task, but it's even harder when you have a hardware issue you didn't really know about.. as you first need to find that out...
Well, without the rotator chip, the display works without any issues.
With the rotator chip included in the setup, we didn't really get a picture anymore - sometimes a bit of flickering, but nothing really.
Of course, we thought the timings were not properly done yet and tried to fix them... but without much success.
Nikolaus checked the hardware and found out that the voltage for the solomon chip was too low. So he started looking for the issue,
After some more tests, he found a quick workaround by adding some wires to the setup. Apparently the LCD cable and some traces were too thin for the power needed, which was a good guess.
With that workaround, the timings could be tried. Nikolaus managed to get a picture... but slightly shifted (as you can see at the picture).
However, Nikolaus needs to continue to work on the hardware, so I made a setup at home involving the OMAP5 devboard connected to the Pyra Mainboard and the LCD, a Pandora and a webcam.
notaz can use that setup to fiddle around with the LCD and see the result - and he got it to work stable and properly, no shift anymore.
That's for the unrotated image, but fully going through the Solomon chip, which is the most important step! Next step is to enable rotation and find the proper setup, but with the timings we already found out now, it should (hopefully) be not a too hard step.
In the meantime, Nikolaus made a lot more tests found the weaknesses and improved the setup to be A LOT more stable.
Since today he knows what needs to be changed to solve these power issues, and once that's done, he can finally finish the CPU Board.
So it took quite a while, but the important thing is that there's a solution
As we got all parts to populate the prototypes already, once the CPU Board layout has been finished, populating it should happen within short time frame.
2. Financing
The closer we get to the production, the more detailled our cost calculation is.
As promised, there's enough money to fund the full development - but we will not have the money to fund the full production.
Which is nothing to be surprised with - as we need to buy a lot of parts upfront in high quantities, we need about 500,000 EUR (half a million EUR!) for the first batch.
This is quite a bit of money (at least for us )
About half of that doesn't need to be paid before the production of the units is finished, which is great.
And I'll also try raising as much money upfront so that preordering will cost you as little money as possible!
There are multiple possibilities (looking for some more investors, getting a short-frame credit for the production run, the upcoming RetroPi2-project at Kickstarter, etc.)
I don't want to scare you, as we know we can do it (I managed to survive the Pandora!), but as financing the production is also part of the whole process, I thought I'd let you know that.
Well - and donations help as well, which I want to mention here.
As you might've noticed, we got a whopping 9000 EUR donation yesterday - THANKS A LOT to the sponsor
We'll have the next goal up within the next few days, as I'll get the accurate pricing for the case mold then.
1. The Hardware
Well, we had quite a few issues to solve which slowed things down a bit, but it seems the solution has been found.
I'll try to explain what happened and what took us four weeks to solve.
As you know, we were trying to get the Solomon (Rotator) chip to work.
This is not an easy task - there are quite a lot of timing settings between the OMAP5, the Solomon Chip and the Panel Driver which need to be properly setup and which even affect each others.
Calculating proper, usable values alone is quite a time consuming task, but it's even harder when you have a hardware issue you didn't really know about.. as you first need to find that out...
Well, without the rotator chip, the display works without any issues.
With the rotator chip included in the setup, we didn't really get a picture anymore - sometimes a bit of flickering, but nothing really.
Of course, we thought the timings were not properly done yet and tried to fix them... but without much success.
Nikolaus checked the hardware and found out that the voltage for the solomon chip was too low. So he started looking for the issue,
After some more tests, he found a quick workaround by adding some wires to the setup. Apparently the LCD cable and some traces were too thin for the power needed, which was a good guess.
With that workaround, the timings could be tried. Nikolaus managed to get a picture... but slightly shifted (as you can see at the picture).
However, Nikolaus needs to continue to work on the hardware, so I made a setup at home involving the OMAP5 devboard connected to the Pyra Mainboard and the LCD, a Pandora and a webcam.
notaz can use that setup to fiddle around with the LCD and see the result - and he got it to work stable and properly, no shift anymore.
That's for the unrotated image, but fully going through the Solomon chip, which is the most important step! Next step is to enable rotation and find the proper setup, but with the timings we already found out now, it should (hopefully) be not a too hard step.
In the meantime, Nikolaus made a lot more tests found the weaknesses and improved the setup to be A LOT more stable.
Since today he knows what needs to be changed to solve these power issues, and once that's done, he can finally finish the CPU Board.
So it took quite a while, but the important thing is that there's a solution
As we got all parts to populate the prototypes already, once the CPU Board layout has been finished, populating it should happen within short time frame.
2. Financing
The closer we get to the production, the more detailled our cost calculation is.
As promised, there's enough money to fund the full development - but we will not have the money to fund the full production.
Which is nothing to be surprised with - as we need to buy a lot of parts upfront in high quantities, we need about 500,000 EUR (half a million EUR!) for the first batch.
This is quite a bit of money (at least for us
)About half of that doesn't need to be paid before the production of the units is finished, which is great.
And I'll also try raising as much money upfront so that preordering will cost you as little money as possible!
There are multiple possibilities (looking for some more investors, getting a short-frame credit for the production run, the upcoming RetroPi2-project at Kickstarter, etc.)
I don't want to scare you, as we know we can do it (I managed to survive the Pandora!), but as financing the production is also part of the whole process, I thought I'd let you know that.
Well - and donations help as well, which I want to mention here.
As you might've noticed, we got a whopping 9000 EUR donation yesterday - THANKS A LOT to the sponsor
We'll have the next goal up within the next few days, as I'll get the accurate pricing for the case mold then.
