[gta02-core] ECN0010: LDO2/LDO4 swap implemented

Rene Harder rehar at saweb.de
Mon Nov 23 19:55:48 CET 2009


Werner Almesberger wrote:
> Joerg Reisenweber wrote:
>   
>> * C4119..22, +..28 move to other side of choke
>>     
>
> This change is in ECN0038, which has been in "execute" for a while.
> Rene, do you want to implement it ?
>
>   

This is on my to-do list but unfortunately I couldn't find time to do
much work on gta02-core in the last couple of weeks. Hopefully I'll have
some spare time to catch up during the upcoming holidays.


>
>>   thought we go make that path symmetric like formerly done
>>   anyway, so r4303 mustn't 0R
>>     
>
> Indeed. What happened with that ?
>   

This one is on my to-do list as well. (new ECN)
IIRC, we agreed to change R4303 & R4305 to 1k each. However, I just saw
that R4303 and R4305 are 2.2k in the big-C buzz fix, so I  would suggest
to stick to those values. We know for sure that this works and the
higher the resistance the better the noise performance.

>   
>>   c4307 is a nonsense, even to have pads for it. Remove C4307 footprints
>>     

Indeed, this cap is not useful at all.

>>   we should replace (A)-R4305-(B) by (A)-R4305-(X), (X)-C4306_NC-(GND),
>> (X)-R4399_0R-(B)
>>   move C4303, C4304, R4305 next to wm8753, behind B3005
>>     

I think that's a really good idea. In this configuration you completely
prevent RF from getting into MICBIAS through JK4401.

>> *make B3005 (and B4105, B3004) look like decent C-L-C filters, like e.g B4419
>>     
>
>   
Having caps on each side of the chokes will definitely harden the audio
subsystem against EMI and is probably a save move. However this will add
another 9 components to the schematics in an already crowded area.
So I'm not totally sure if that is really necessary.

>> *kill R3004, R3005 finally!
>>     
>
>   

Okay, let's kill them.

>   
>> *c3017 -> 1uF ?
>>     
>
>   

Agreed!

>   
>> *c3003 AND c3008??? Seems we don't need both
>>  c3002 same
>>     
>
> Looks like the usual HF/LF pair, with the 100 nF cap right next to
> the chip and the much larger 10 uF cap at some distance.
>
>   

Wolson recommends to have a 100nF decoupling cap for each power supply
pin, which leads to multiple 100nF caps on the power rails of the codec.
Actually, there is one 100nF cap missing. DVCC_CODEC is connected to 3
pins however only two 100nF caps are in the schematic.

>>  do we need B3001??
>>     
>
> I'm not sure. Neither side of the bead should be infested with RF noise.
> If in doubt, I'd keep it.
>
>   

I'd also keep the bead,  just in case.





More information about the gta02-core mailing list