From LinuxTVWiki
Revision as of 23:27, 28 February 2013 by Gomboc (Talk | contribs)

Jump to: navigation, search

The page is to describe and provide information for a device called EZgrabber2 which appears to be produced Mygica and Geniatech.


The EZGrabber2 device shares the windows drivers for the similar devices Capit and EZGrabber, so likely much of the information will be in common with those. Please clarify if you have any of these products:

MyGica/Geniatech USB EZGrabber2 and similar analog video capture cards
Device Product url(s) USB ID Chipset
EZGrabber2 Mygica and Geniatech 1f4d:0102


CX78921-11z, compatible with cx231xx linux framework
EZGrabber My Gica EZGrabber ? ?
Capit MyGica Capit MyGica Capit Old site Geniatech ? ?

Supposedly supports output of compressed MPEG1/2/4 stream(on windows). On linux, using a patched cx231xx driver, it is at least possible to get uncompressed YUV4:2:2 video and 48000Hz stereo through the corresponding alsa device.

It has the standard USB 2.0, stereo, composite and S-video inputs.

Components Used

The single chip on the board is labeled as a CX78921-11z. This is no reference to this chip on Conexant's website nor practically any the entire internet, so likely this is just a renumbered chip for these specific vendors. However the indication that is can output compressed MPEG4 video does suggest that it is not your standard cx23100/1/2.

However examining the Windows drivers provided by the vendors, they refer to the device Polaris.CVBS or POLARIS.OTG102, and contain many of the same driver files as other devices containing the cx23100/1/2 chips .


  • until I have permission to upload the images to this site, they are posted on google sites, click the links beside the description.

Making it work

We need to add the device definition and some related details to the cx231xx driver and have installed the appropriate firmware (see below). Also we need to have built and load the kernel modules for cx25840 and cx231xx before plugging in the device (#modprobe cx25840;modprobe cx231cc). If I don't do this, the cx231xx module will autoload on detecting the device, but the cx25840 module will not autoload, and the device doesn't function. I don't know why, go ask your dad.

The following is a patch against a kernel source 3.6. It should be fairly applicable to sources from 2.6.32 on through 3.7, with the possible note of the device number in cx2311.h. Note that in 3.8 and presumably onwards, the structure of the drivers/media/ directory has been changed and no longer is there a drivers/media/video directory, but rather the video drivers are regrouped by interface (usb, pci), so the cx231xx drivers are under drivers/media/usb/cx231xx while the related and also needed driver for the cx25840 is under drivers/media/pci/.

The has also been a lot of recent (early 2013) patches submitted to patchwork.kernel.org/project/linux-media that may alter future the viability of the patch below once they make it to mainline, without subsequent modifications.

Kernel Patch

I'll also post a downloadable patch once I have permission to upload to this site. Until then, and for posterity, the text is below.

diff -uprN linux-3.6/drivers/media/video/cx231xx/cx231xx-avcore.c linux-3.6.new/drivers/media/video/cx231xx/cx231xx-avcore.c
--- linux-3.6/drivers/media/video/cx231xx/cx231xx-avcore.c	2012-09-30 17:47:46.000000000 -0600
+++ linux-3.6.new/drivers/media/video/cx231xx/cx231xx-avcore.c	2013-02-26 19:58:51.096793077 -0700
@@ -352,6 +352,7 @@ int cx231xx_afe_update_power_control(str
 	case CX231XX_BOARD_CNXT_RDE_253S:
 	case CX231XX_BOARD_CNXT_RDU_253S:
+	case CX231XX_BOARD_OTG102:
@@ -1719,6 +1720,7 @@ int cx231xx_dif_set_standard(struct cx23
 	case CX231XX_BOARD_CNXT_RDU_250:
+	case CX231XX_BOARD_OTG102:
 		func_mode = 0x03;
diff -uprN linux-3.6/drivers/media/video/cx231xx/cx231xx-cards.c linux-3.6.new/drivers/media/video/cx231xx/cx231xx-cards.c
--- linux-3.6/drivers/media/video/cx231xx/cx231xx-cards.c	2012-09-30 17:47:46.000000000 -0600
+++ linux-3.6.new/drivers/media/video/cx231xx/cx231xx-cards.c	2013-02-28 12:23:58.925869674 -0700
@@ -280,6 +280,37 @@ struct cx231xx_board cx231xx_boards[] =
+	[CX231XX_BOARD_OTG102] = {
+                .name = "Geniatech OTG102",
+                .tuner_type = TUNER_ABSENT,
+                .decoder = CX231XX_AVDECODER,
+                .output_mode = OUT_MODE_VIP11,
+                .ctl_pin_status_mask = 0xFFFFFFC4,
+                .agc_analog_digital_select_gpio = 0x0c, /* According with PV CxPlrCAP.inf file */
+                .gpio_pin_status_mask = 0x4001000,
+                .norm = V4L2_STD_NTSC,
+                .no_alt_vanc = 1,
+                .external_av = 1,
+                .dont_use_port_3 = 1,
+		//.has_417 = 1,
+		/* this board has hardware encoding chip supporting mpeg1/2/4, but as the 417 is apparently not working for the
+		   reference board it is not on this one either. building the driver with this option and then loading the module
+		   creates a second video device node, but nothing comes out of it.  */
+                .input = {{
+                                .type = CX231XX_VMUX_COMPOSITE1,
+                                .vmux = CX231XX_VIN_2_1,
+                                .amux = CX231XX_AMUX_LINE_IN,
+                                .gpio = NULL,
+                        }, {
+                                .type = CX231XX_VMUX_SVIDEO,
+                                .vmux = CX231XX_VIN_1_1 |
+                                        (CX231XX_VIN_1_2 << 8) |
+                                        CX25840_SVIDEO_ON,
+                                .amux = CX231XX_AMUX_LINE_IN,
+                                .gpio = NULL,
+                        }
+                },
+        },
 	[CX231XX_BOARD_CNXT_RDE_250] = {
 		.name = "Conexant Hybrid TV - rde 250",
 		.tuner_type = TUNER_XC5000,
@@ -620,6 +651,8 @@ struct usb_device_id cx231xx_id_table[]
 	 .driver_info = CX231XX_BOARD_CNXT_RDU_253S},
 	{USB_DEVICE(0x0572, 0x58A6),
 	 .driver_info = CX231XX_BOARD_CNXT_VIDEO_GRABBER},
+	{USB_DEVICE(0x1F4D, 0x0102),
+	 .driver_info = CX231XX_BOARD_OTG102},
 	{USB_DEVICE(0x0572, 0x589E),
 	 .driver_info = CX231XX_BOARD_CNXT_RDE_250},
 	{USB_DEVICE(0x0572, 0x58A0),
@@ -904,6 +937,12 @@ static int cx231xx_init_dev(struct cx231
 		cx231xx_set_alt_setting(dev, INDEX_VIDEO, 3);
 		cx231xx_set_alt_setting(dev, INDEX_VANC, 1);
+	if (dev->model == CX231XX_OTG102) {
+		cx231xx_set_alt_setting(dev, INDEX_VIDEO, 3);
+		cx231xx_set_alt_setting(dev, INDEX_VANC, 1);
+	}
 	/* Cx231xx pre card setup */
@@ -1295,6 +1334,12 @@ static int cx231xx_usb_probe(struct usb_
 		cx231xx_set_alt_setting(dev, INDEX_VIDEO, 3);
+	}
+	if (dev->model == CX231XX_BOARD_OTG102) {
+		cx231xx_enable_OSC(dev);
+		cx231xx_reset_out(dev);
+		cx231xx_set_alt_setting(dev, INDEX_VIDEO, 3);
 	if (dev->model == CX231XX_BOARD_CNXT_RDE_253S)
diff -uprN linux-3.6/drivers/media/video/cx231xx/cx231xx.h linux-3.6.new/drivers/media/video/cx231xx/cx231xx.h
--- linux-3.6/drivers/media/video/cx231xx/cx231xx.h	2012-09-30 17:47:46.000000000 -0600
+++ linux-3.6.new/drivers/media/video/cx231xx/cx231xx.h	2013-02-26 16:01:58.924653199 -0700
@@ -68,6 +68,7 @@
 #define CX231XX_BOARD_ICONBIT_U100 13
+#define CX231XX_BOARD_OTG102 16
 /* Limits minimum and default number of buffers */
 #define CX231XX_MIN_BUF                 4


Like for many similar devices of various chipsets, a firmware image or images are required that the driver uploads to the device. These are often closed source binary blobs under unclear license as far as their redistributeability. To further complicate matters, there are various versions of these firmwares with the same name that are not (always) cross-compatible. To complicate matters even further, there is apparently a mix-up one of the firmware files distrubted though linuxtv.org and linux-firmware.git. See [1] and especially [2] for more information.

I have these in my /lib/firmware folder with the device working, although I still get messages in the system log that "unable to open firmware v4l-cx23885-avcore-01.fw". These came from [3]. The Geniatech driver also contains a firmware image called cx416enc.rom which is very similar to the v4l-cx23885-enc.fw file (which is a renamed hcw85enc.rom from the HVR1800 driver).

   16382 Jun 22  2010 /lib/firmware/v4l-cx231xx-avcore-01.fw   (doesn't try to load)
   16382 Feb 27 17:49 /lib/firmware/v4l-cx23885-avcore-01.fw  (doesn't load)
   16382 Feb 25 22:23 /lib/firmware/v4l-cx23885-avcore-01.fw2 (alternate version, also doesn't load)
  376836 Feb 25 22:23 /lib/firmware/v4l-cx23885-enc.fw
   16382 Jun 22  2010 /lib/firmware/v4l-cx25840.fw

#md5sum /lib/firmware/v4l-cx2*
a9f8f5d901a7fb42f552e1ee6384f3bb  /lib/firmware/v4l-cx231xx-avcore-01.fw
a9f8f5d901a7fb42f552e1ee6384f3bb  /lib/firmware/v4l-cx23885-avcore-01.fw 
b3704908fd058485f3ef136941b2e513  /lib/firmware/v4l-cx23885-avcore-01.fw2
1cb3c48a6684126f5e503a434f2d636b  /lib/firmware/v4l-cx23885-enc.fw
dadb79e9904fc8af96e8111d9cb59320  /lib/firmware/v4l-cx25840.fw

Notes about the Windows drivers

I purchased an EZGrabber2 from Meritline early February 2013. It was the MyGica variant although the product page links to the Geniatech driver and software.

MyGica provides drivers at ezgrabber2driver.zip and otg103driver.zip (for the Capit) , both are the exact same file.

Geniatech lists a single driver for the EzGrabber/EzGrabber2/Capit products, ezgrabber2driver.zip.


The install .inf file from the MyGica driver is called CxPlrCap.inf and contains the following device definitions:

%CXPOLARIS_NAME.U237%        =POLARIS.U237,    USB\VID_1F4D&PID_0237&MI_01

The driver directory contains the following files:

  23678 Oct 28  2009 CxPlrCap.inf
 187776 Jan  6  2010 CxPlrCap.sys
  29184 Sep  8  2009 CxPolaris.ax
  18432 Sep  8  2009 cpnotify.ax
  10154 Jan  7  2010 cxplrcap.cat
  24576 Sep  8  2009 cxtvrate.dll
    336 Sep 24  2009 maxcir.bin
   8404 Jan  7  2010 maxcir.cat
   1440 Sep 17  2009 maxcir.inf
  32256 Oct  9  2009 maxcir.sys
  16382 Jan 20  2009 merlinD.bin
  28672 Sep  8  2009 y8cnvt.ax


The install .inf file from the Geniatech driver is called CxCVBS.inf and contains the following device definitions:

%CXPOLARIS.DeviceDesc-VGrabber%=POLARIS.VGrabber,  USB\VID_0572&PID_58A6&MI_01
%CXPOLARIS.DeviceDesc-EZCap%=POLARIS.EZCap,        USB\VID_0572&PID_58A7&MI_01

The driver directory contains the following files:

   18432 Oct 23 16:12 CPNotify.ax
   31704 Oct 23 16:11 CxCVBS.inf
  195712 Oct 23 16:13 CxCVBS.sys
    3449 Oct 23 17:36 CxUnCir.inf
   47104 Oct 23 16:12 EncPrxyPage.ax
 2944000 Sep  7 15:31 Setup.exe
   33792 Oct 23 16:12 Y8Cnvt.ax
  376836 Jul 18  2012 cx416enc.rom
   15360 Oct 23 16:12 cxEZCAP.ax
   47104 Oct 23 16:12 cxTVRate.dll
   11407 Oct 23 16:13 cxcvbs.cat
   30720 Oct 23 16:12 cxpolaris.ax
    8578 Oct 23 17:36 cxuncir.cat
   16382 Jul 18  2012 merlinD.rom

It is also worth noting that the Geniatech ezgrabber2driver.zip file contains both a Windows and Linux directories. It was due to this that I originally purchased this device hoping they had included some support for Linux and they do provide some linux drivers for some of there other products. However, it appears to be a mistake. The only modification of drivers in the tree I can find are some alternate definitions for au0828 cards, specifically defining 05e1:0400 and 1f4d:6011 as AU0828_BOARD_HAUPPAUGE_WOODBURY.

Other notes

In some places it is noted that some versions of the Diamond VC500 EZgrabber are also identified as OTG102, a cursory web search suggests this has a TM5600/6000 chipset and the windows drivers provided by diamond appear totally different.