XMOS_Master review

XCore Project reviews, ideas, videos and proposals.
muntablues
Active Member
Posts: 34
Joined: Tue Dec 20, 2011 2:45 pm

XMOS_Master review

Post by muntablues »

Hi all

It would be great if someone of you could have a look at my prototype board.

It is based on "USB Audio 2.0" and "USB Audio 2.0 DJ Kit" reference designes, so most of the parts should work fine, but maybe I have missed something!

The plan is, to have one "controller" board which can be used for different USB audio devices. So this board is used as "universal mainboard" and I only change the analog part for the differnt devices, e.g. 4x stereo out or 2x stereo out/in and so on.

Thanks in advance!

MB
Attachments
XMOS_Master_1.0_SCH.pdf
Schematic
(33.75 KiB) Downloaded 417 times
XMOS_Master_1.0_SCH.pdf
Schematic
(33.75 KiB) Downloaded 417 times
XMOS_Master_1.0_PCB_BOT.pdf
PCB Bottom
(37.96 KiB) Downloaded 401 times
XMOS_Master_1.0_PCB_BOT.pdf
PCB Bottom
(37.96 KiB) Downloaded 401 times
XMOS_Master_1.0_PCB_TOP.pdf
PCB Top
(74.37 KiB) Downloaded 441 times
XMOS_Master_1.0_PCB_TOP.pdf
PCB Top
(74.37 KiB) Downloaded 441 times


User avatar
EdB
Member++
Posts: 17
Joined: Tue Mar 19, 2013 11:58 am

Post by EdB »

The layout of the DCDC components is not optimal, probably due to the power copper pours.

If you look at the reference layout in the U8 datasheet you will see that the power supply ground paths return directly to the relevant decoupling cap ground, and then return to the plane.

Also the ground plane on the bottom side looks a little narrow where the ground balls of the device are connected, and there is a relatively long path from them to the main ground entry point.

I would recommend routing the power supplies as traces and have top and bottom ground pours for a 2 layer design, as per the reference layout.

Ed
muntablues
Active Member
Posts: 34
Joined: Tue Dec 20, 2011 2:45 pm

Post by muntablues »

Hi Ed

Thank you for your answer!

I looked at "USB Audio 2.0 DJ Kit" PCB files and so I thought my GND plane should be ok. If you look at them you will see (in my opinion) that XMOS missed their rules as well ;-)

But I will give it a try to route two GND planes...

MB
muntablues
Active Member
Posts: 34
Joined: Tue Dec 20, 2011 2:45 pm

Post by muntablues »

Hi Ed

OK I have changed the routing, but not exactly as you said. But it should be much better now. I have a real solid ground plane, every GND connection is as close as possible and that should be ok (hope).

Maybe someone or you could have a look at the schematic, that would be great!

Thanks MB
Attachments
XMOS_Master_1.0_SCH.pdf
(34.1 KiB) Downloaded 369 times
XMOS_Master_1.0_SCH.pdf
(34.1 KiB) Downloaded 369 times
XMOS_Master_1.0_PCB_Top.pdf
(73.1 KiB) Downloaded 452 times
XMOS_Master_1.0_PCB_Top.pdf
(73.1 KiB) Downloaded 452 times
XMOS_Master_1.0_PCB_Bot.pdf
(34.94 KiB) Downloaded 365 times
XMOS_Master_1.0_PCB_Bot.pdf
(34.94 KiB) Downloaded 365 times
User avatar
EdB
Member++
Posts: 17
Joined: Tue Mar 19, 2013 11:58 am

Post by EdB »

Hi,

That looks better, it would be worth considering swapping the positions of L4<->C13 and L3<->C14 so you can avoid running the power supply ground paths through vias. These traces need to be able to handle transient currents peaking in the amps.

Note the arrangement below, this allows for the shortest, widest path for the DCDC currents.
U8 power.PNG
(195.33 KiB) Not downloaded yet
U8 power.PNG
(195.33 KiB) Not downloaded yet
Ed
muntablues
Active Member
Posts: 34
Joined: Tue Dec 20, 2011 2:45 pm

Post by muntablues »

Hi Ed

Ok, that makes sense, thank you for that hint!!!

MB
muntablues
Active Member
Posts: 34
Joined: Tue Dec 20, 2011 2:45 pm

Post by muntablues »

Hi Ed

I think I got it and it should ok now. Maybe you could have a last look on it.

Did you see any mistakes on the schematic?

Thanks again!!!

MB
Attachments
XMOS_Master_1.0_PCB.pdf
(100.38 KiB) Downloaded 381 times
XMOS_Master_1.0_PCB.pdf
(100.38 KiB) Downloaded 381 times
User avatar
EdB
Member++
Posts: 17
Joined: Tue Mar 19, 2013 11:58 am

Post by EdB »

I would just remove the GND stub connected to pin L2 and that should do it.

I can't see anything obvious with the schematic. The only note is that if you ever intend for the design to be self powered (rather than bus powered) then you'll need USB_VBUS connected.

Ed
muntablues
Active Member
Posts: 34
Joined: Tue Dec 20, 2011 2:45 pm

Post by muntablues »

Hi Ed

OK, thanks again. I think if I make it self powerd +5V connection should be ok as well. I just leave L1 and it is fine.

So I think I will order the first proto boards and hope that everything is working rigth...

MB
User avatar
EdB
Member++
Posts: 17
Joined: Tue Mar 19, 2013 11:58 am

Post by EdB »

To make it self powered you'd just need to supply the 5V externally, that's fine. You will however need to have the USB_VBUS pin connected to the VBUS line from the USB connector so the U8 can detect the presence of the bus and swithc the phys pull-up resistors accordingly.

Let me know how the proto goes.

Ed
Post Reply