[Milkymist-devel] Draft 1 schematics

[Sébastien Bourdeauducq]

Hi,

Thanks for the review!

On Monday 18 January 2010 02:49:31 Adam Wang wrote:
> A), Audio
> 	1, LM4550BVH pin 42 is AUDIO_AGND, pin 38 is AUDIO_AVDD. But from
> datasheet p.14, if you connected like this, means it's for second
> source application otherwise they should be left open(NC).

Yes. And it's advisable to keep the second-source possibility open :)

> Also I
> found that pin 45&46 are NC, that means it acts as primary codec
> function but you still can use register set to let 4550 to be
> secondary codec, see "Multiple Codecs" in page 26. The pin 6 of
> BIT_CLK will be act as either output(24.576MHz/2=AC97_CLK) in primary
> codec or input(=AC97_CLK) in secondary mode.

The LM4550 will always be used as primary codec, and, as you pointed out, feed 
the AC97 clock into the FPGA.

> B), Ethernet
> 	1, Can add pull high 10k at U2 pin26/27 <LED1&LED0> to 3V3, see p43
> of KSZ8001L.

Those pins have internal pull-ups, so external resistors are not necessary. 
They should not be needed even in the presence of a LED, because that LED is 
connected between the pin and 3V3 so it won't pull down the weak internal 
pull-up anyway.
I already built a KSZ8001L board without those external pull-ups and it 
worked.

By the way, Ethernet should be changed to use MII instead of the current RMII. 
I'll send an updated version.

> C, DMX512
> 	1, Vs(pin3) of U8 TSOP4838 should be connected to 5V not 3V3, see p2
> of TSOP48[2]. Supposed that it's a old version. But please be noticed
> that pin1's(IR_RX) voltage. But if use "new TSOP48xx" series, its
> operation voltage is 2.7~5.5V. From Vishay official site, should
> currently only have this new part.

Actually, my main reason for using the TSOP48xx is that their new series 
supports 3.3V operation, because the FPGA (where the IR_RX signal goes) is not 
5V-tolerant.

Sébastien