Finishing up the hardware


EvilDragon

Administrator
Staff member
Joined
Mar 4, 2003
Messages
29,986
Age
46
Location
Ingolstadt
We've made some good progress - so time again to make another summary :)

 

1. The Rotator Chip

 

The only issue left right now is that we got some tearing when using the rotator chip.

For those who don't know what that is: When writing to and reading from the framebuffer is not properly synchronized, the picture you can see of the LCD is from two different frames (i.e. top half frame 1, bottom half frame 2).

This gives an effect as if the screen is torn into two halfs when you got videos or games with fast movement.

 

To prevent this, you need to synchronize the OMAP5 with the SSD (rotator chip) so that the framebuffer is not being overwritten while it's being read.

 

This is a bit more software work than we thought. There's a special signal line from the SSD to the OMAP that is being used to synchronize both devices, but that only works in MIPI Command Mode, whereas we are running the OMAP5 using MIPI Video Mode.

So as a solution, we'd either need to change the existing driver somehow so that the TE line can be used in Video Mode as well OR try to use the OMAP5 in Command Mode instead of Video Mode.

 

We don't know yet which one of those two will be easier to achieve, but both of them might be a bit time consuming.

 

Nevertheless, the hardware itself is working fine, so Nikolaus now concentrates on the remaining hardware - fixing the software can happen while we're waiting for the mass production to start (or even afterwards).

 

So the display including the rotator works fine, but we currently have some tearing issues with fast moving content until we improved the driver.

 

 

2. The case

 

Basically, the case is finished.

I've printed the last revision and am just waiting for some final prototype boards to assemble everything - and once I confirm that everything fits, the real mold is being created.

I expect it to be finished about October.

 

Of course, as soon as I assemble it, I'll create a video of it!

 

 

3. The keyboard

 

The keyboard design is basically finished.

The normal keys will be surrounded by a bit of transparent silicone mat (but will be 0,1mm higher), which will give a nice feeling and also light up the key surroundings as well as the characters.

 

keymat.jpg

 

The only exception will be the gaming buttons:

Here, the plastic of the buttons will be glued on top of the mat, so they feel like standard gaming buttons (yes, this has changed within the last few days).

 

cap.jpg

 

Mold production here should start within the next week and will take approximately one month.

 

The means: Yep, there's no way I can run away from reading the keyboard layout threads anymore ;)

 

 

 

 

4. The hardware

 

Nikolaus was working on the rotator chip as well, to find out what was hardware and software related.

As you probably can imagine, fixing hardware later will be hard, but fixing software is possible :)

 

A week ago, Nikolaus restarted the work on the hardware.

Within this week, he created 106 traces within 15 hours work (so about 7 traces per hour).

 

cpulayout.jpg

 

410 traces are left to be created - so if he continues in the same pace, the CPU board will be finished in about 15 days.

Could be a bit faster as well though (some non-critical traces could be made using the auto-router), but we'll see.

 

Our goal is to have the full hardware finished within July, so we actually have real prototypes ready to show at the GamesCom in August.

Yep, if you're visiting there, you might be able to play on a real Pyra prototype!

 

DSC02309.JPGDSC02306.JPG

 

Shortly after that preorders will start (unless we find some major issue with the prototype) - with a lot of videos following.

 

Well, that's what we're planning anyways - let's hope we can keep that and it will work out!

 

Nikolaus also started to work on the e-SATA adaptor which plugs into our USB port for those who want to connect a SATA device to the Pyra.

 

DSC02304.jpeg

 

Some good progress, I'd say :)

Hopefully back with more good news soon.
 
Glad to see the gaming buttons will now be glued on. Was wondering if the earlier design with plastic just in the centre would feel okay. Thanks for the heaps of work. May the July target be hit!
 
Last edited by a moderator:
Awesome news.  I still don't quite understand how a software change could make this TE line operational, but as long as notaz and anyone else involved does, that's fine by me.

I'm also glad you listened to people's concerns about the gaming buttons and changed them to be more like the Pandora buttons (but just with the plastic caps glued on instead of free-floating).  I wasn't too concerned about how the keyboard-style buttons would feel as gaming buttons personally, but I think these will look better at least.

Regarding the SATA adaptor, I'm a little surprised to see the mask including space for an IC on there.  I expected the adaptor to be a passive adaptor from the extra USB3 lines to SATA lines (plus ground from the original USB lines), but I guess it's not quite as simple as that.
 
Out of curiosity: will any keys in the middle of the mat, like F and J, have any markings to help key-navigation by touch?

Looking good, I hope to be able to pre-order soon. Would it be possible to order a Pyra including the cellular modem and satellite navigation device at the earliest time?
 
Awesome news.  I still don't quite understand how a software change could make this TE line operational, but as long as notaz and anyone else involved does, that's fine by me.
Well, one way is to use the MIPI Command Mode instead of video mode.

That mode officially supports the TE signal out of the box. 

Apart from that, the TE line is simply a signal coming from the SSD. It triggers whenever it awaits a new frame.

This can be read out in software and be hacked into the driver so synchronize the video stream it sends.

Regarding the SATA adaptor, I'm a little surprised to see the mask including space for an IC on there.  I expected the adaptor to be a passive adaptor from the extra USB3 lines to SATA lines (plus ground from the original USB lines), but I guess it's not quite as simple as that.
The parts needed for the eSATA port have been removed from the PCB to save costs and moved onto the adapter.

Without the adaptor, you couldn't use eSATA anyways, so no need to have all the needed parts inside the normal device.

Out of curiosity: will any keys in the middle of the mat, like F and J, have any markings to help key-navigation by touch?
Can be done, but first we need to finish the keymat design for that.
 
The parts needed for the eSATA port have been removed from the PCB to save costs and moved onto the adapter.

Without the adaptor, you couldn't use eSATA anyways, so no need to have all the needed parts inside the normal device. 
That makes me very sad, as I had planned to mod in a SATA device internally to it.
 
The parts needed for the eSATA port have been removed from the PCB to save costs and moved onto the adapter.

Without the adaptor, you couldn't use eSATA anyways, so no need to have all the needed parts inside the normal device.
 That makes me very sad, as I had planned to mod in a SATA device internally to it.
Well, what keeps you from adding that few parts into your mod as well? ;)
 
Any chance we could get a tail on or included with that eSATA adapter? I want to minimize the amount of torque being applied to the port by gravity, a solid adapter is quite a lever, and finding really short (~0.5 ft.) USB 3.0 extensions that aren't also angle adapters or converters is currently a bit tricky.
 
Last edited by a moderator:
Is there a reason to use the non-standard sata adaptor rather than just use a USB3.0 to SATA device?

Would it be possible to add a cam to the case mold so that some small percentage of the cases can be made without that double wide space bar that some of us hate with a passion? You don't need to make a different keymat since it would be easy to cut and splice a pair of keymats into a single keymat with no double wide space bar. Trying to mod the case to add the missing part over the double wide space bar would be very difficult, and near impossible to make it look professional.
 
Is there a reason to use the non-standard sata adaptor rather than just use a USB3.0 to SATA device?
USB 3.0 has a theoretical speed of 5Gbit/s, and taking encoding overhead etc into account, you can get about 3.2Gbit/s (400MB/s) in practice.

eSATA has a theoretical speed of 6Gbit/s and has somewhat less overhead, so you can get about 4.8Gbit/s (600MB/s) in practice.

If you happen to have a device that reaches higher speeds than what USB 3.0 offers, then if you just use a standard SATA-to-USB adapter and plug it into a USB port, the speed will be capped at 400MB/s. The point of this non-standard Pyra-specific adapter is to bypass the USB port completely, and only use its physical wires to get access directly to the OMAP5 SATA port. This necessarily has to be something non-standard, because the USB port is used in a non-standard way.
 
3. The keyboard

 

The keyboard design is basically finished.

The normal keys will be surrounded by a bit of transparent silicone mat (but will be 0,1mm higher), which will give a nice feeling and also light up the key surroundings as well as the characters.

 

attachicon.gif
keymat.jpg

 

The only exception will be the gaming buttons:

Here, the plastic of the buttons will be glued on top of the mat, so they feel like standard gaming buttons (yes, this has changed within the last few days).

 

attachicon.gif
cap.jpg

 

Mold production here should start within the next week and will take approximately one month.
I suppose this also means that there's slightly more room on the action buttons for a label, since there's no circle padding anymore. Will the action button keycaps be printed on the underside as well? (that could be harder to do, since they are not flat pieces of plastic like the rest of the keyboard)

Are all 6 action buttons like this, or just the 4 big ones?
 
Last edited by a moderator:
Is there a reason to use the non-standard sata adaptor rather than just use a USB3.0 to SATA device?
Don't forget that the USB port in question is still only a USB2.0 port despite the USB3.0 style connector being used, so you would still need to use an adaptor to connect a USB3.0 device anyway.
- Neelix
 
Is there a reason to use the non-standard sata adaptor rather than just use a USB3.0 to SATA device?
Don't forget that the USB port in question is still only a USB2.0 port despite the USB3.0 style connector being used, so you would still need to use an adaptor to connect a USB3.0 device anyway.
Oh, that too. Though you wouldn't need an adapter to connect a USB 3.0 device since it is backwards compatible. But if you would connect, say, an external hard disk which can reach 700MB/s and which has both an eSATA and a USB 3.0 connector, then if you would use the USB 3.0 one on a big USB port on the Pyra, it would get only USB 2.0 speeds (35 MB/s). If you would use the USB 3.0 connector with a µUSB cable and connect it to the Pyra's USB 3.0 OTG port, you would get USB 3.0 speeds (400MB/s). If you would use the eSATA connector, you would get 600MB/s.
 
Great news! About the NUBs, will be improved this time? I have heard that the second production for Pandora still had a high failure rate.
 
Good to know about the progresses,

I'd need the Pyra pretty soon -- just lost my Pandora this week in the train (RE, Euskirchen to Trier). I currently hope to get it back, but as always .... you  never know.

thumbs up on the progress and the nice work, one of the pre-orders will be mine (hopefully .-))

peb
 
Back
Top