ice1724/Audiotrak Prodigy 7.1 HiFi

ice1724/Audiotrak Prodigy 7.1 HiFi

[Julian Scheel]

I just got a Audiotrak Prodigy 7.1 HiFi card, which is baed upon a Envy24HT 
chipset in combination with one WM8776 codec for front out and a WM8766 for 
channel 3-8.
I made the front out working, taking the code from pontis.c - Now I'd like to 
add support for the other channels, therefore I need to figure out the 
I2C-base-address for the WM8766 registers.
Can you provide me with some help on how to figure that out? I'm a newbie to 
alsa/soundcard driver development.
Reading the envy24ht-specs didn't help me, yet.

One more question: Should I merge my code into pontis.c, which would blow it a 
bit up, but would avoid duplicate code, or should I open a new 
prodigy_hifi.c/h?

Cheers,
Julian

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: ice1724/Audiotrak Prodigy 7.1 HiFi

[Takashi Iwai]

At Fri, 15 Sep 2006 21:11:03 +0200,
Julian Scheel wrote:
> 
> I just got a Audiotrak Prodigy 7.1 HiFi card, which is baed upon a Envy24HT 
> chipset in combination with one WM8776 codec for front out and a WM8766 for 
> channel 3-8.
> I made the front out working, taking the code from pontis.c - Now I'd like to 
> add support for the other channels, therefore I need to figure out the 
> I2C-base-address for the WM8766 registers.
> Can you provide me with some help on how to figure that out? I'm a newbie to 
> alsa/soundcard driver development.

First you should make clear whether the codec chip is connected via
the vt1724's standard connection or 3-wired SPI connection via GPIO.
The most boards use the latter method while pontis board uses the
former.  In the latter case, the address is usually 2 bit, so you can
do a wild guess more easily.

> Reading the envy24ht-specs didn't help me, yet.
> 
> One more question: Should I merge my code into pontis.c, which would blow it a 
> bit up, but would avoid duplicate code, or should I open a new 
> prodigy_hifi.c/h?

Depends.  Send us a patch, and let's decide this later.


Takashi

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

Re: ice1724/Audiotrak Prodigy 7.1 HiFi

[Julian Scheel]

Am Dienstag, 19. September 2006 11:27 schrieb Takashi Iwai:
> At Fri, 15 Sep 2006 21:11:03 +0200,
> First you should make clear whether the codec chip is connected via
> the vt1724's standard connection or 3-wired SPI connection via GPIO.
> The most boards use the latter method while pontis board uses the
> former.  In the latter case, the address is usually 2 bit, so you can
> do a wild guess more easily.

I figured out that the WM8766 is connected through a 3-wired SPI-connection. I 
also figured out that the GPIOs 16,17,18 are used.
pontis.c already includes a SPI-implementation, but according to the WM8766 
specs (look at wolfsonmico.com) this device does not await an address, but 
only 2 bytes of data: register address/values.
Actually trying to initializing the chip leads to silence, in comparison to 
the noise coming out before, when the chip was in default mode.
I assume that there must be some timing problem with the data-transmission or 
similiar. I will do some more testing today.
Maybe I'll send in a patch if I can't find a problem, anyway.

-Julian

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

Re: ice1724/Audiotrak Prodigy 7.1 HiFi

[Julian Scheel]

[-- Attachment #1: Type: text/plain, Size: 1460 bytes --]

Am Dienstag, 19. September 2006 11:40 schrieb Julian Scheel:
> Am Dienstag, 19. September 2006 11:27 schrieb Takashi Iwai:
> > At Fri, 15 Sep 2006 21:11:03 +0200,
> > First you should make clear whether the codec chip is connected via
> > the vt1724's standard connection or 3-wired SPI connection via GPIO.
> > The most boards use the latter method while pontis board uses the
> > former.  In the latter case, the address is usually 2 bit, so you can
> > do a wild guess more easily.
>
> I figured out that the WM8766 is connected through a 3-wired
> SPI-connection. I also figured out that the GPIOs 16,17,18 are used.
> pontis.c already includes a SPI-implementation, but according to the WM8766
> specs (look at wolfsonmico.com) this device does not await an address, but
> only 2 bytes of data: register address/values.
> Actually trying to initializing the chip leads to silence, in comparison to
> the noise coming out before, when the chip was in default mode.
> I assume that there must be some timing problem with the data-transmission
> or similiar. I will do some more testing today.
> Maybe I'll send in a patch if I can't find a problem, anyway.

Ok, attached is now what I've done so far. (patch against 1.0.13rc2) This 
causes the WM8766 to be silent. Only way to get the noise back is restarting 
the system without the init-code.
Couldn't find a reason for that behaviour. Maybe you have an idea what could 
be wrong with that code?

-Julian

[-- Attachment #2: prodigy71_hifi.patch --]
[-- Type: text/x-diff, Size: 5622 bytes --]

diff -ru alsa-driver-1.0.13rc2.orig/alsa-kernel/pci/ice1712/pontis.c alsa-driver-1.0.13rc2/alsa-kernel/pci/ice1712/pontis.c
--- alsa-driver-1.0.13rc2.orig/alsa-kernel/pci/ice1712/pontis.c	2006-09-12 15:40:57.000000000 +0200
+++ alsa-driver-1.0.13rc2/alsa-kernel/pci/ice1712/pontis.c	2006-09-19 13:52:55.000000000 +0200
@@ -67,6 +67,22 @@
 #define WM_OUT_MUX		0x16
 #define WM_RESET		0x17
 
+/* WM8766 registers */
+#define WM8766_DAC_CTRL		0x02   /* DAC Control */
+#define WM8766_INT_CTRL		0x03   /* Interface Control */
+#define WM8766_DAC_CTRL2	0x09
+#define WM8766_DAC_CTRL3	0x0a
+#define WM8766_RESET		0x1f
+#define WM8766_LDA1		0x00
+#define WM8766_LDA2		0x04
+#define WM8766_LDA3		0x06
+#define WM8766_RDA1		0x01
+#define WM8766_RDA2		0x05
+#define WM8766_RDA3		0x07
+#define WM8766_MUTE1		0x0C
+#define WM8766_MUTE2		0x0F
+
+
 /*
  * GPIO
  */
@@ -75,6 +91,10 @@
 #define PONTIS_CS_RDATA		(1<<6)	/* CS8416 -> VT1720 */
 #define PONTIS_CS_WDATA		(1<<7)	/* VT1720 -> CS8416 */
 
+#define WM8766_SPI_CLK		(1<<17) /* CLK, Pin97 on ICE1724 */
+#define WM8766_SPI_MD		(1<<16) /* DATA VT1724 -> WM8766, Pin96 */
+#define WM8766_SPI_ML		(1<<18) /* Latch, Pin98 */
+
 
 /*
  * get the current register value of WM codec
@@ -353,7 +373,12 @@
 		tmp |= bit;
 	else
 		tmp &= ~bit;
-	snd_ice1712_gpio_write(ice, tmp);
+/*printk("BIT: %08x\n", bit);
+printk("GPIO W: %08x\n", tmp);*/
+        snd_ice1712_gpio_write(ice, tmp);
+/*        udelay(1);
+tmp = snd_ice1712_gpio_read(ice);
+printk("GPIO R: %08x\n", tmp);*/
 }
 
 static void spi_send_byte(struct snd_ice1712 *ice, unsigned char data)
@@ -428,6 +453,39 @@
 	return val;
 }
 
+/*
+ * SPI implementation for WM8766 codec - only writing supported, no readback
+ */
+
+static void wm8766_spi_send_byte(struct snd_ice1712 *ice, unsigned char data)
+{
+	int i;
+	for (i = 0; i < 8; i++) {
+		set_gpio_bit(ice, WM8766_SPI_CLK, 0);
+		udelay(1);
+		set_gpio_bit(ice, WM8766_SPI_MD, data & 0x80);
+		udelay(1);
+		set_gpio_bit(ice, WM8766_SPI_CLK, 1);
+		udelay(1);
+		data <<= 1;
+	}
+}
+
+static void wm8766_spi_write(struct snd_ice1712 *ice, unsigned int reg, unsigned int data)
+{
+	snd_ice1712_gpio_set_dir(ice, WM8766_SPI_MD|WM8766_SPI_CLK|WM8766_SPI_ML);
+	snd_ice1712_gpio_set_mask(ice, ~(WM8766_SPI_MD|WM8766_SPI_CLK|WM8766_SPI_ML));
+	set_gpio_bit(ice, WM8766_SPI_ML, 0); /* latch must be low when writing */
+	wm8766_spi_send_byte(ice, reg); /* REGISTER ADDRESS */
+	wm8766_spi_send_byte(ice, data); /* DATA */
+	/* release latch */
+	set_gpio_bit(ice, WM8766_SPI_ML, 1);
+	udelay(1);
+	/* restore */
+	snd_ice1712_gpio_set_mask(ice, ice->gpio.write_mask);
+	snd_ice1712_gpio_set_dir(ice, ice->gpio.direction);
+}
+
 
 /*
  * SPDIF input source
@@ -776,6 +834,21 @@
 		WM_DAC_MUTE,	0x0000,	/* DAC unmute */
 		WM_ADC_MUX,	0x0003,	/* ADC unmute, both CD/Line On */
 	};
+        static unsigned short wm8766_inits[] = {
+/*                WM8766_RESET,		0x01,*/
+		WM8766_DAC_CTRL,	0x90,
+                WM8766_INT_CTRL,	0x22, /* I2S Normal Mode, 24 bit */
+                WM8766_DAC_CTRL2,	0x00,
+		WM8766_DAC_CTRL3,	0x10,
+		WM8766_LDA1,		0xff,
+		WM8766_LDA2,		0xff,
+		WM8766_LDA3,		0xff,
+		WM8766_RDA1,		0xff,
+		WM8766_RDA2,		0xff,
+		WM8766_RDA3,		0xff,
+		WM8766_MUTE1,		0x20,
+		WM8766_MUTE2,		0x10,
+        };
 	static unsigned char cs_inits[] = {
 		0x04,	0x80,	/* RUN, RXP0 */
 		0x05,	0x05,	/* slave, 24bit */
@@ -788,12 +861,19 @@
 	ice->vt1720 = 1;
 	ice->num_total_dacs = 2;
 	ice->num_total_adcs = 2;
+	ice->akm_codecs = 1;
+
+        switch (ice->eeprom.subvendor) {
+        case VT1720_SUBDEVICE_PRODIGY71_HIFI:
+		ice->vt1720 = 0;
+                ice->num_total_dacs = 8;
+                ice->akm_codecs = 2;
+        }
 
 	/* to remeber the register values */
-	ice->akm = kzalloc(sizeof(struct snd_akm4xxx), GFP_KERNEL);
+        ice->akm = kcalloc(ice->akm_codecs, sizeof(struct snd_akm4xxx), GFP_KERNEL);
 	if (! ice->akm)
 		return -ENOMEM;
-	ice->akm_codecs = 1;
 
 	/* HACK - use this as the SPDIF source.
 	 * don't call snd_ice1712_gpio_get/put(), otherwise it's overwritten
@@ -807,6 +887,13 @@
 	for (i = 0; i < ARRAY_SIZE(wm_inits2); i += 2)
 		wm_put(ice, wm_inits2[i], wm_inits2[i+1]);
 
+        /* initialize WM8766 codec */
+        for (i = 0; i < ARRAY_SIZE(wm8766_inits); i += 2)
+{
+printk("SET WM8766 REGISTER\n");
+                wm8766_spi_write(ice, wm8766_inits[i], wm8766_inits[i+1]);
+}
+
 	/* initialize CS8416 codec */
 	/* assert PRST#; MT05 bit 7 */
 	outb(inb(ICEMT1724(ice, AC97_CMD)) | 0x80, ICEMT1724(ice, AC97_CMD));
@@ -853,5 +940,14 @@
 		.eeprom_size = sizeof(pontis_eeprom),
 		.eeprom_data = pontis_eeprom,
 	},
+	{
+		.subvendor = VT1720_SUBDEVICE_PRODIGY71_HIFI,
+		.name = "Audiotrak Prodigy 7.1 HiFi",
+		.model = "prodigy71hifi",
+		.chip_init = pontis_init,
+		.build_controls = pontis_add_controls,
+		.eeprom_size = sizeof(pontis_eeprom),
+		.eeprom_data = pontis_eeprom,
+	},
 	{ } /* terminator */
 };
Only in alsa-driver-1.0.13rc2/alsa-kernel/pci/ice1712: pontis.c.orig
diff -ru alsa-driver-1.0.13rc2.orig/alsa-kernel/pci/ice1712/pontis.h alsa-driver-1.0.13rc2/alsa-kernel/pci/ice1712/pontis.h
--- alsa-driver-1.0.13rc2.orig/alsa-kernel/pci/ice1712/pontis.h	2006-09-12 15:40:57.000000000 +0200
+++ alsa-driver-1.0.13rc2/alsa-kernel/pci/ice1712/pontis.h	2006-09-19 13:52:55.000000000 +0200
@@ -27,6 +27,7 @@
 #define PONTIS_DEVICE_DESC 	       "{Pontis,MS300},"
 
 #define VT1720_SUBDEVICE_PONTIS_MS300	0x00020002	/* a dummy id for MS300 */
+#define VT1720_SUBDEVICE_PRODIGY71_HIFI	0x38315441	/* audiotrak prodigy */
 
 extern struct snd_ice1712_card_info  snd_vt1720_pontis_cards[];
 

[-- Attachment #3: Type: text/plain, Size: 348 bytes --]

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

[-- Attachment #4: Type: text/plain, Size: 161 bytes --]

_______________________________________________
Alsa-devel mailing list
Alsa-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/alsa-devel

Re: ice1724/Audiotrak Prodigy 7.1 HiFi

[Julian Scheel]

Am Dienstag, 19. September 2006 13:56 schrieb Julian Scheel:
> Ok, attached is now what I've done so far. (patch against 1.0.13rc2) This
> causes the WM8766 to be silent. Only way to get the noise back is
> restarting the system without the init-code.
> Couldn't find a reason for that behaviour. Maybe you have an idea what
> could be wrong with that code?
>

One more note about that patch:
I know coding-style is quite horrible and it'd probably break pontis. It's 
just what I hacked together for testing, and as the DAC doesn't seem to 
understand what I am sending I'd like to get some feedback on the 
SPI-implementation, maybe that one is faulty? (The WM8766 sheet isn't very 
detailed about that)

-Julian

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

Re: ice1724/Audiotrak Prodigy 7.1 HiFi

[James Courtier-Dutton]

Julian Scheel wrote:
> Am Dienstag, 19. September 2006 13:56 schrieb Julian Scheel:
>   
>> Ok, attached is now what I've done so far. (patch against 1.0.13rc2) This
>> causes the WM8766 to be silent. Only way to get the noise back is
>> restarting the system without the init-code.
>> Couldn't find a reason for that behaviour. Maybe you have an idea what
>> could be wrong with that code?
>>
>>     
>
> One more note about that patch:
> I know coding-style is quite horrible and it'd probably break pontis. It's 
> just what I hacked together for testing, and as the DAC doesn't seem to 
> understand what I am sending I'd like to get some feedback on the 
> SPI-implementation, maybe that one is faulty? (The WM8766 sheet isn't very 
> detailed about that)
>
> -Julian
>   

After our conversation on IRC and the resulting help I was able to give 
you, I understand that you managed to fix this.
When are you going to post the completed patch?

James


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

Re: ice1724/Audiotrak Prodigy 7.1 HiFi

[Julian Scheel]

James Courtier-Dutton schrieb:
> Julian Scheel wrote:
>> Am Dienstag, 19. September 2006 13:56 schrieb Julian Scheel:
>>  
>>> Ok, attached is now what I've done so far. (patch against 1.0.13rc2) 
>>> This
>>> causes the WM8766 to be silent. Only way to get the noise back is
>>> restarting the system without the init-code.
>>> Couldn't find a reason for that behaviour. Maybe you have an idea what
>>> could be wrong with that code?
>>>
>>>     
>>
>> One more note about that patch:
>> I know coding-style is quite horrible and it'd probably break pontis. 
>> It's just what I hacked together for testing, and as the DAC doesn't 
>> seem to understand what I am sending I'd like to get some feedback on 
>> the SPI-implementation, maybe that one is faulty? (The WM8766 sheet 
>> isn't very detailed about that)
>>
>> -Julian
>>   
> 
> After our conversation on IRC and the resulting help I was able to give 
> you, I understand that you managed to fix this.
> When are you going to post the completed patch?
> 
> James
> 

Sadly if been out of work for several days now, because of a tonsilitis.
That's the reason, why I haven't finished up the work yet. Hopefully 
I'll do so next week and post the result then.
Though I haven't yet worked with digital out, which might need some more 
work, too. But I think I'd post analog-only patches first.

-Julian

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

Re: ice1724/Audiotrak Prodigy 7.1 HiFi

[Takashi Iwai]

At Fri, 29 Sep 2006 17:37:31 +0200,
Julian Scheel wrote:
> 
> James Courtier-Dutton schrieb:
> > Julian Scheel wrote:
> >> Am Dienstag, 19. September 2006 13:56 schrieb Julian Scheel:
> >>  
> >>> Ok, attached is now what I've done so far. (patch against 1.0.13rc2) 
> >>> This
> >>> causes the WM8766 to be silent. Only way to get the noise back is
> >>> restarting the system without the init-code.
> >>> Couldn't find a reason for that behaviour. Maybe you have an idea what
> >>> could be wrong with that code?
> >>>
> >>>     
> >>
> >> One more note about that patch:
> >> I know coding-style is quite horrible and it'd probably break pontis. 
> >> It's just what I hacked together for testing, and as the DAC doesn't 
> >> seem to understand what I am sending I'd like to get some feedback on 
> >> the SPI-implementation, maybe that one is faulty? (The WM8766 sheet 
> >> isn't very detailed about that)
> >>
> >> -Julian
> >>   
> > 
> > After our conversation on IRC and the resulting help I was able to give 
> > you, I understand that you managed to fix this.
> > When are you going to post the completed patch?
> > 
> > James
> > 
> 
> Sadly if been out of work for several days now, because of a tonsilitis.
> That's the reason, why I haven't finished up the work yet. Hopefully 
> I'll do so next week and post the result then.
> Though I haven't yet worked with digital out, which might need some more 
> work, too. But I think I'd post analog-only patches first.

Good to hear.  The patch will be a 2.6.20-kernel material, so we have
a couple of months until a complete version :)

Anyway, ALSA-1.0.13 is out, and the new patches are open for merge
now.  Please post a patch fore review at any time.


thanks,

Takashi

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

ice1724/Audiotrak Prodigy 7.1 HiFi

[Konstantin Kletschke]

Hi Folks!

Currently I am awaiting my Audiotrak Prodigy 7.1 HiFi soundcard to be
delivered as I realize, that some work needs to be put into the driver
for it. I revealed a thread from September started by julian
investigating the stuff. As I found nothing of the results in 1.0.14_rc1
(did not investigate mercurial so far) I ask here, if some help is need
to finish and if I can help sorting things out.

I could investigate the hardware and code kernel driver with datasheets
and oscilloscope (doing this mainly @work with very boring stuff :-P),
or, julian, if you are too busy to tidy up your patches I could check
them also and commit them to ml.

Regards, Konstantin Kletschke

PS.: Is there any other soundcard with an analog output stage as good as
of this card (terratec phase22, audiophile192, and maya44 are not
looking bad, but not as good as the prodigy 7.1 hifi...)?

-- 
GPG KeyID EF62FCEF
Fingerprint: 13C9 B16B 9844 EC15 CC2E  A080 1E69 3FDA EF62 FCEF

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV