We've started to order things!


Hi all,

Is there any chance to replace the .....

At this point I think any kind of hardware change or substitution -barring a real technical error or some kind of problem- is unrealistic. 'We're' nearing the final board layout for production aren't 'we'? It would be a bad time to start adding or removing features IMO.
 
The twl6040 chip handles headphone amplification.
Of course it does, almost every other DAC does the same. It provides typical smartphone power level which is good for typical 16-20 Ohm headphones.
Additional amplification is needed for higher impedance headphones and quality headphones are usually in that category.
 
Type-C requires a lot more than just exchanging the socket, because it introduced a lot of new features. The new alternate modes and the fact that it can be used upside down require a more complex wiring and you'll likely won't get around adding another IC, there's just no room to do that. There's a reason why one of the ICs near the backside uses a weird angle on the PCB (aka "wonky chip").
maybe when they do the 1st board upgrade when the pyra takes off they will look that. i also would love the usb-c also, but i can wait for it
 
Of course it does, almost every other DAC does the same. It provides typical smartphone power level which is good for typical 16-20 Ohm headphones.
Additional amplification is needed for higher impedance headphones and quality headphones are usually in that category.

Actually looking at the specs of the headphone amplifier in the Pandora vs the Pyra they're practically the same in terms of driving capabilities.

Pandora - http://www.ti.com/lit/ds/symlink/tpa6110a2.pdf
Pyra - http://www.ti.com/lit/ds/symlink/twl6040.pdf
 
Of course it does, almost every other DAC does the same. It provides typical smartphone power level which is good for typical 16-20 Ohm headphones.
Additional amplification is needed for higher impedance headphones and quality headphones are usually in that category.

Yes but those mainly use the big jacks unlike 'smartphone headphones', so you cant insert those anyway. Convertors are available but imo the soundquality that is produced with or wihout amp won't really be up to par for quality headphones.
 
Actually looking at the specs of the headphone amplifier in the Pandora vs the Pyra they're practically the same in terms of driving capabilities.

Pandora - http://www.ti.com/lit/ds/symlink/tpa6110a2.pdf
Pyra - http://www.ti.com/lit/ds/symlink/twl6040.pdf

Maybe I'm not interpreting it right, but from what I see there:
6110a2 is 150mW stereo. i.e. 75mW per channel.
6040 is 1.5W per channel

So is the 6040 a 20X more powerful amp, or what am I interpreting wrong?

Edit: To double audio volume requires 10X more amplification. So, one would not be 20X louder than the other, more like double and a bit louder. Overall volume does not denote quality, but more powerful amplification tends to maintain the cross frequency sound load with fewer artifacts or distortions.
 
Last edited:
Hmm how about some more random info from your dmesg.. Is the 4GiB part working now with LPAE? :)

Yo Tony! What brought you here?

It does run in LPAE, but with 2GB still so far. The other half is still having a bit of a fight with the init since it's dual-rank, so not much examples around, and the ram init code in u-boot is... not very transparent, especially since TI is using undocumented PHY registers there I haven't actually even managed to find yet where it's setting certain registers, but when I inspected them they turned out to contain nonsentical values for the second rank.

Note that the sort of situation the kernel gets (with a large part not accessible by other initiators) can of course already be tested with 2GB ram: just tell the kernel it has two ranges:
  • 0'800'00'000 - 0'BFF'FF'FFF (1 GB)
  • 2'C00'00'000 - 2'FFF'FF'FFF (1 GB)
 
He (m@t) is really a drolly kind of fellow, isn't he ? Always asking and moaning the same, while everybody here in the forum tried to explain him, that
  • There was never given a de facto release date for distributing the Pyra.
  • That ED always stated, he tries to deliver a maximum possible functional Pyra.
  • That it's in our all interest to clear out problems before delivering Pyras as far as possible even if this results in a further delay.
  • That this project is a community driven project and not comparable with projects driven by multi million/billion dollar companies.
  • And last but not least: We all trust in ED because of his sublime and outstanding handling of the Pandora project.
Go on ED and make our dreams of a wonderful and unrivaled Linux handheld device possible.

If he has preordered maybe his preorder should be cancelled to prevent all this stress he has.
 
Actually looking at the specs of the headphone amplifier in the Pandora vs the Pyra they're practically the same in terms of driving capabilities.

Pandora - http://www.ti.com/lit/ds/symlink/tpa6110a2.pdf
Pyra - http://www.ti.com/lit/ds/symlink/twl6040.pdf

Well, twl6040 spec isn't as detailed but if 'capability' really means the ability to provide such power for long term usage then yes, they're somewhat close.
The problem is that twl6040 was used in Galaxy Nexus and that phone didn't look like a great headphone driver.

Yes but those mainly use the big jacks unlike 'smartphone headphones', so you cant insert those anyway. Convertors are available but imo the soundquality that is produced with or wihout amp won't really be up to par for quality headphones.

There are lots of cans around 100 Ohm with the usual 3.5 mm jack. Just look at the 'better than average' category.
 
I'm still concerned at how close the bottom of the dpad and the X button are to the keyboard keys. I have large thumbs, I can see myself accidentally hitting a key when going for d-pad down or the X button.
I know ED said this was a real issue when doing early tests and was looking to address the issue. I think ED said he was adjusting the height of the keys/ buttons to address it. I'm wondering if there still is an issue there?


  1. dsc00672-jpg.30042
 
Yo Tony! What brought you here?

It does run in LPAE, but with 2GB still so far. The other half is still having a bit of a fight with the init since it's dual-rank, so not much examples around, and the ram init code in u-boot is... not very transparent, especially since TI is using undocumented PHY registers there I haven't actually even managed to find yet where it's setting certain registers, but when I inspected them they turned out to contain nonsentical values for the second rank.

Note that the sort of situation the kernel gets (with a large part not accessible by other initiators) can of course already be tested with 2GB ram: just tell the kernel it has two ranges:
  • 0'800'00'000 - 0'BFF'FF'FFF (1 GB)
  • 2'C00'00'000 - 2'FFF'FF'FFF (1 GB)

Hi, I've been lurking for quite some time, your post got me curious enough to register! Just in case this might help.. I assume you are aware that igepv5 works with LPAE and 4GB with just kernel command line:

mem=2032M@0x80000000 mem=2048M@0x300000000

And I think there's the second cpu startup loop and whatever other boot rom stuff is at the 2032-2048 so maybe it needs to be preserved for pyra too? No idea if that contains any DDR init related code, but possibly. Anyways maybe also check the extra patches igepv5 u-boot git tree at git://git.isee.biz/pub/scm/u-boot-arm.git, you can grep for IGEPv5_MEMORY_4G if you did not do that already. Also n-tuple-check the pin muxing for cke1 and ncs1..
 
I'm still concerned at how close the bottom of the dpad and the X button are to the keyboard keys. I have large thumbs, I can see myself accidentally hitting a key when going for d-pad down or the X button.
I know ED said this was a real issue when doing early tests and was looking to address the issue. I think ED said he was adjusting the height of the keys/ buttons to address it. I'm wondering if there still is an issue there?
Somewhat outdated Gamescom 2016 experience here, but it was essentially a non-issue for me. I think the action buttons stopped at a height before you'd even push down keyboard keys.
But as I said old-ish experience, so don't quote me on it. The keyboard keys themselves were a bit harder to push. Not sure if the case part was lowered or the keys are higher as of now, though.


Something what I'd rather want to know: What happened to the heat spreading solution? I know it's supposed to be a simple thing, but has any work towards that been done yet?
 
Something what I'd rather want to know: What happened to the heat spreading solution? I know it's supposed to be a simple thing, but has any work towards that been done yet?

That's something I'd like to know also, ie what is the cooling solution going into the Pyra?. Is there any?
 
Should be a kind of "magic" paper which helps to spread the heat across the device.
Some components may also help, like eh SD metal cage.
 
That's something I'd like to know also, ie what is the cooling solution going into the Pyra?. Is there any?
It's essentially just dumping the heat into the ground plane of the main PCB.
 
The second one was upside-down (which was weird).
http://www.independent.co.uk/news/science/why-do-sloths-hang-upside-down-9287482.html

See? It's not that weird...
[doublepost=1487407390,1487405862][/doublepost]
Yeah! Do that!

Thank you people who know more than me. 2 Months coming up.....

Well... knowledge is just a search away: the concept of MUX and one of the pins of the CPU is called cke1 is here for Pandora.

Of course... I know less than you in these matters. I just nod and smile. It makes me look intelligent.

So if I read "u-boot-arm.git / blob". I know what an U-boat is. And git is certainly a typo so you have arm.pit which is a nasty concept. blob is of course a movie... or another typo for byob (bring your own beer)

anyways: nod and smile...
 
Last edited:
Back
Top