From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mail-bw0-f210.google.com ([209.85.218.210]:48455 "EHLO mail-bw0-f210.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754863AbZJKBBx (ORCPT ); Sat, 10 Oct 2009 21:01:53 -0400 Received: by bwz6 with SMTP id 6so2408308bwz.37 for ; Sat, 10 Oct 2009 18:01:16 -0700 (PDT) Date: Sun, 11 Oct 2009 04:01:12 +0300 From: "Aleksandr V. Piskunov" To: ivtv-devel@ivtvdriver.org, linux-media@vger.kernel.org Cc: Andy Walls Subject: Re: [PATCH] AVerTV MCE 116 Plus radio Message-ID: <20091011010039.GA4726@moon> References: <20091006080406.GA22207@moon> <20091006081159.GB22207@moon> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20091006081159.GB22207@moon> Sender: linux-media-owner@vger.kernel.org List-ID: On Tue, Oct 06, 2009 at 11:11:59AM +0300, Aleksandr V. Piskunov wrote: > On Tue, Oct 06, 2009 at 11:04:06AM +0300, Aleksandr V. Piskunov wrote: > > Added FM radio support to Avermedia AVerTV MCE 116 Plus card > > > > What leaves me puzzled, radio only works ok with ivtv newi2c=1 > > With default newi2c audio is tinny, metallic, with some strange static. > Similar problem with pvr-150 was reported years ago, guess issue is still > unresolved, perhaps something with cx25840.. This particular "tinny" audio problem is definitely I2C speed related, to be more precise, audio only goes bad if i2c-algo-bit is being run with udelay less than 15, i.e. i2c bus frequency is higher than 30 KHz. So with default udelay=10 or udelay=5 (optimal for IR reciever on that board) radio goes bad. Running with newi2c=1 is ok, but again it isn't optimal for IR reciever on AVerTV M116. I2C reads/writes to cx25840 themself are ok, verified using register readback after each write/write4. Problem seems to be that with cx25840 register writes coming too fast on higher i2c bus speed, switching register 0x808 _from_ TV standard autodetection mode (0xff) _to_ FM radio mode (0xf9) leaves chip audio detection routine in inconsistent state. The only solution I found is to do standard routine (assert_reset + write + deassert_reset) followed by 50ms delay and another reset. Following patch works_for_me, can be improved to only delay/doublereset when really needed, etc. Andy, could you comment/review? diff --git a/linux/drivers/media/video/cx25840/cx25840-core.c b/linux/drivers/media/video/cx25840/cx25840-core.c --- a/linux/drivers/media/video/cx25840/cx25840-core.c +++ b/linux/drivers/media/video/cx25840/cx25840-core.c @@ -626,7 +642,13 @@ if (state->radio) { cx25840_write(client, 0x808, 0xf9); cx25840_write(client, 0x80b, 0x00); - } + /* Double reset cx2384x after setting FM radio mode, helps to + avoid "tinny" audio when ivtv I2C bus is being run on + frequency higher than 30 KHz */ + cx25840_and_or(client, 0x810, ~0x01, 0); + msleep(50); + cx25840_and_or(client, 0x810, ~0x01, 1); + } else if (std & V4L2_STD_525_60) { /* Certain Hauppauge PVR150 models have a hardware bug that causes audio to drop out. For these models the