This week (or more precisely: yesterday!) Global Components made a surprise attack!
They did the production run of the 30 prototype CPU boards - without telling us upfront.
1. Prototype CPU boards finished
The first thing I heard about them was from Nikolaus in the evening... when he told me he just grabbed 6 produced CPU boards, fully X-Rayed and looking fine.
What a surprise!
He made a quick boot-up test of a CPU board and it worked fine.
The remaining ones are currently being checked thoroughly by GC (X-Rayed, etc.) and Nikolaus will check the one he received next week.
The most important thing is to find out whether the system still freezes with 1,5GHz or not - as that was the only change on the CPU boards we couldn't test upfront.
The second important thing is to test the 4GB CPU boards (I'm going to send one to zmatt as he wanted to work on that).
If both works... well, then the CPU boards are ready for mass production!
2. The remaining boards
Nikolaus finished the work on the Mainboard early last week as well and we have now submitted the final design files as well as the BOM (Part List) to GC.
These should hopefully be the final revision as well as all known fixes have been included.
The production of the bare PCB will take a few weeks, but not as long as with the CPU boards, as they are A LOT simpler boards (the CPU board is 12-layer... the display board 2 layer and the mainboard 6-layer).
Once I receive more details, I'll let you know.
Well, this is kind of a milestone for the project.
With all known hardware errors squashed, it should be ready for the mass production!
3. The LCD Cable
The LCD Cable tester we created were really helpful, as it helped me a lot analysing why I had issues with the LCD Cables for the prototypes.
One important thing first: As it looks like, the latest revision of the cables works fine, so no more prototype needed!
Now the explanation what I did and why I had issues:
As mentioned, the company accidentally produced cables that were a bit too thick but those were the only ones we had ready for the GamesCom, so we used them.
They were hard to fit into the prototype units, but they worked.
Then, a few weeks ago, I wanted to build more units for developers and planned to use the newly produced LCD cables (which had proper thickness).
However, I couldn't get any unit to work properly with those... so we first had the suspicion that the ears were too close to the edge which could cause short circuits or that the too thick cables caused damage to the LCD cable connector, which would mean that we'd have to replace them on our existing prototype PCBs.
With the small tester PCB we created, I could do more tests and check how well the cables work with the connectors.
Good news: The latest revision cables fit perfectly. No shortcircuits, no broken contacts, everything working fine.
But what happens to the connectors when we use one of the thicker cables?
Tried that.
The thicker cables work fine as well, they just are hard to fit into the connectors (well, for the first 2 - 3 times... after that, they're an easy fit).
Then I used one of the new cables which has been proven to be working fine: And yes, some connections are broken!
Seems like the thicker cables bend or damage the pins inside the connector so they don't have proper contact with the normal cable anymore!
That's a VERY good reason I had issues building the prototypes!
Last test: Putting one of the thick cables into the broken connectors.
Result: Working fine!
So yes, the pins are just BENT, not broken. Which means: Once we cramped the too thick cable in there, it will only work with the thick cable in the future.
Tried that. Build a prototype with one of the older, too thick cables: Worked like a charm on first try!
So the conclusion is:
The cables are working fine, no change needed. And for the old prototype boards we will build using the thicker cables.
4. Case and keymat
I decided to let the keymat company produce the DPad as well and apply it directly to the keymat (similar to the one on the Pandora, just more reliable and we don't have to do it ourselves).
This is simply because it doesn't really add more costs and I've made a lot of tests - but without glueing the DPad, it just doesn't feel as good.
I'm waiting to receive samples of both of them.
5. The wandering prototype
I am thinking of maybe use one prototype to send it to interested preorders for a couple of days and then they'll send it to the next one on the list.
While the software surely won't be fully working, everyone can at least try the keyboard and overall design.
I'm not sure I'll do it, but it seems like an interesting idea.
The only thing the interested user has to pay are the shipping costs to the next one on the list.
It would ship without a battery though, as shipping items with batteries included isn't cheap.
What do you think?
6. Server move
You have probably found out that the current server has some networking issues right now.
Sometimes I've got up to 80% of package loss... so it feels unresponsive.
Don't worry about that though, I am currently moving to a new server anyways!
This'll take some time as moving all the files will take quite a while, but I'm already doing that.
There will be a few downtimes here and there until the move is done, but after that, the server should be faster than before.
The details of the new server:
Intel® Xeon® E3-1275 v5 Quad-Core Skylake incl. Hyper-Threading Technology
RAM64 GB DDR4 ECC (twice as much as on the old one)
Hard Drive: 2 x 4 TB SATA 6 Gb/s 7200 rpm Class Enterprise
Guaranteed Bandwidth: 1 Gbit/s (200MBit/s on the old one)
Those were the news of the week. If you got questions, feel free to ask
They did the production run of the 30 prototype CPU boards - without telling us upfront.
1. Prototype CPU boards finished
The first thing I heard about them was from Nikolaus in the evening... when he told me he just grabbed 6 produced CPU boards, fully X-Rayed and looking fine.
What a surprise!
He made a quick boot-up test of a CPU board and it worked fine.
The remaining ones are currently being checked thoroughly by GC (X-Rayed, etc.) and Nikolaus will check the one he received next week.
The most important thing is to find out whether the system still freezes with 1,5GHz or not - as that was the only change on the CPU boards we couldn't test upfront.
The second important thing is to test the 4GB CPU boards (I'm going to send one to zmatt as he wanted to work on that).
If both works... well, then the CPU boards are ready for mass production!
2. The remaining boards
Nikolaus finished the work on the Mainboard early last week as well and we have now submitted the final design files as well as the BOM (Part List) to GC.
These should hopefully be the final revision as well as all known fixes have been included.
The production of the bare PCB will take a few weeks, but not as long as with the CPU boards, as they are A LOT simpler boards (the CPU board is 12-layer... the display board 2 layer and the mainboard 6-layer).
Once I receive more details, I'll let you know.
Well, this is kind of a milestone for the project.
With all known hardware errors squashed, it should be ready for the mass production!
3. The LCD Cable
The LCD Cable tester we created were really helpful, as it helped me a lot analysing why I had issues with the LCD Cables for the prototypes.
One important thing first: As it looks like, the latest revision of the cables works fine, so no more prototype needed!
Now the explanation what I did and why I had issues:
As mentioned, the company accidentally produced cables that were a bit too thick but those were the only ones we had ready for the GamesCom, so we used them.
They were hard to fit into the prototype units, but they worked.
Then, a few weeks ago, I wanted to build more units for developers and planned to use the newly produced LCD cables (which had proper thickness).
However, I couldn't get any unit to work properly with those... so we first had the suspicion that the ears were too close to the edge which could cause short circuits or that the too thick cables caused damage to the LCD cable connector, which would mean that we'd have to replace them on our existing prototype PCBs.
With the small tester PCB we created, I could do more tests and check how well the cables work with the connectors.
Good news: The latest revision cables fit perfectly. No shortcircuits, no broken contacts, everything working fine.
But what happens to the connectors when we use one of the thicker cables?
Tried that.
The thicker cables work fine as well, they just are hard to fit into the connectors (well, for the first 2 - 3 times... after that, they're an easy fit).
Then I used one of the new cables which has been proven to be working fine: And yes, some connections are broken!
Seems like the thicker cables bend or damage the pins inside the connector so they don't have proper contact with the normal cable anymore!
That's a VERY good reason I had issues building the prototypes!
Last test: Putting one of the thick cables into the broken connectors.
Result: Working fine!
So yes, the pins are just BENT, not broken. Which means: Once we cramped the too thick cable in there, it will only work with the thick cable in the future.
Tried that. Build a prototype with one of the older, too thick cables: Worked like a charm on first try!
So the conclusion is:
The cables are working fine, no change needed. And for the old prototype boards we will build using the thicker cables.
4. Case and keymat
I decided to let the keymat company produce the DPad as well and apply it directly to the keymat (similar to the one on the Pandora, just more reliable and we don't have to do it ourselves).
This is simply because it doesn't really add more costs and I've made a lot of tests - but without glueing the DPad, it just doesn't feel as good.
I'm waiting to receive samples of both of them.
5. The wandering prototype
I am thinking of maybe use one prototype to send it to interested preorders for a couple of days and then they'll send it to the next one on the list.
While the software surely won't be fully working, everyone can at least try the keyboard and overall design.
I'm not sure I'll do it, but it seems like an interesting idea.
The only thing the interested user has to pay are the shipping costs to the next one on the list.
It would ship without a battery though, as shipping items with batteries included isn't cheap.
What do you think?
6. Server move
You have probably found out that the current server has some networking issues right now.
Sometimes I've got up to 80% of package loss... so it feels unresponsive.
Don't worry about that though, I am currently moving to a new server anyways!
This'll take some time as moving all the files will take quite a while, but I'm already doing that.
There will be a few downtimes here and there until the move is done, but after that, the server should be faster than before.
The details of the new server:
Intel® Xeon® E3-1275 v5 Quad-Core Skylake incl. Hyper-Threading Technology
RAM64 GB DDR4 ECC (twice as much as on the old one)
Hard Drive: 2 x 4 TB SATA 6 Gb/s 7200 rpm Class Enterprise
Guaranteed Bandwidth: 1 Gbit/s (200MBit/s on the old one)
Those were the news of the week. If you got questions, feel free to ask