Monday, 2016-10-03

*** tpb has joined #timvideos00:00
thaytanxfxf, that's more a question for one of the python binding guys - bilboed or thibault01:38
thaytanwe used to have support in the old manual bindings though, I think01:39
thaytanso probably just requires some overrides in gst-python01:39
thaytansome reference to the GStreamer 0.10 python bindings might help01:40
*** sb0 has quit IRC03:38
mithro_florent_: ping me if you are around this evening03:45
*** rohitksingh_work has joined #timvideos04:20
*** sb0 has joined #timvideos04:24
*** rohitksingh_wor1 has joined #timvideos04:51
*** rohitksingh_work has quit IRC04:52
*** rohitksingh_wor1 has quit IRC05:21
*** rohitksingh_work has joined #timvideos05:43
*** Bertl_zZ is now known as Bertl06:11
*** rohitksingh_wor1 has joined #timvideos06:39
*** rohitksingh_work has quit IRC06:39
*** Bertl is now known as Bertl_oO06:45
*** NTU has quit IRC07:57
*** RattusRattus has joined #timvideos08:02
RattusRattushay mithro sorry, work has gone carzey, I've not been reading IRC for weeks.08:03
* RattusRattus needs to recap.... give me half hour or so08:07
mithroRattusRattus: okay08:10
RattusRattusSo DCC is just an I2c bus.  As such the SDA signal is bi-directional, but the clock signal needs only be in one direction:  however the datasheet states "Bidirectional Level-Shifting N-Channel FETs Provided for DDC_CLK and DDC_DATA Channels08:13
RattusRattusso yes this is suitable for both VGA IN and VGA output ports08:13
* RattusRattus looks further and sees that the data sheet refers to the pins as In and Out. this is just a nominal name, if you look at section 7.2 you can see that they are FET level shifters and therfore bi-directional.08:17
RattusRattusthe important bit is the "Out" side has the ESD clamp diodes...08:18
RattusRattusmithro: please also note the layout guidelines on the datasheet (figure 9)08:21
RattusRattus(surface routin, no vias)08:21
RattusRattusrouting even!08:21
mithroRattusRattus: The I2C is bidirectional08:26
mithroRattusRattus: the HSYNC / VSYNC signals are not08:26
*** NTU has joined #timvideos08:28
RattusRattusyes - I need to add a couple of resistors to the schematic for the input ~ A simple devider will bring the signals down (1k to SYNC in, 2k to GND, mid point is 3V3 sync signal)08:45
RattusRattusDo not connect pins 13 or 15, but still connect 14 and 16 to give ESD protection08:47
RattusRattusno need for active level shift here08:47
mithroRattusRattus: But it looks like the drivers would cause the output always be driven to a value?09:19
mithroRattusRattus: the SYNC pins on the AD9948 are 5V tolerant09:19
RattusRattusyes I had spotted that and was looking at alternatives:  options include:   seporate clamp diodes (with / without resistor devider - with is probably safest), seporate clamp diodes and level shifter (but looks more expensive)09:26
RattusRattusehhh  hang on....  just looking again.  we can connect vsysnc seporatly.....09:29
RattusRattusso vsync can be 3v3 for driving in reverse09:29
RattusRattusbut we would need to add the resistors still.09:32
mithroRattusRattus: hrm? These VSYNC / HSYNC signals are coming from somewhere else09:33
mithroRattusRattus: Looking at AD9984, it seems like we can just connect them directly to it, with some ESD protection09:33
mithroRattusRattus: an alternative options is to replace the TPD with a PACVGA200 which does seem to separate the drivers from the ESD protection09:34
RattusRattusso the supply for tyhe vsync pin of the shifetr can be 3V3 (it doesn't have to be the same as VCC_video - thats why it is a seporate pin)09:35
* RattusRattus also looks at PACVGA20009:35
mithroRattusRattus: The problem isn't the supply for the shifter - the ESD protection is only on the output side of the shifter, not the input - hence we can't just connect it backwards09:39
RattusRattusno I haven't been clear.  Vcc_sync = 3V3, Sync_In* to Vga connector (via resistor devider), sync_out* to sync input of ADC09:41
RattusRattuswhich is pretty much the same as PACVGA200.09:42
RattusRattusI can change to PACVGA200 if you prefer thoughj09:42
RattusRattusnope no good.  the TPD7S019 sync input pins don't have ESD protection....  so it should be the PACVGA20009:48
* RattusRattus updates schematics09:48
mithroRattusRattus: so, as I mentioned above....09:56
mithroThe PACVGA200 is quite expensive :(09:57
mithroBut I wasn't able to find anything else09:57
mithroRattusRattus: oh, I had unpushed changes to the schematic....09:58
RattusRattusyes as you said above :-(  sorry09:58
mithroRattusRattus: btw - if you say my nick it flashes a notification :)09:58
RattusRattuswill I can price up just adding ESD diodes to sync lines and still using TPD7S01909:58
mithroWell, I guess the PACVGA200 saves the term resistors - which is kind of nice10:10
mithroBut it does seem to be double the TPD7S019 price on Digikey10:11
*** rohitksingh_wor1 has quit IRC10:11
tpbTitle: pacvga200 Stock and Price by Distributor (at
*** rohitksingh_work has joined #timvideos10:26
*** rohitksingh_work has quit IRC10:39
*** rohitksingh_work has joined #timvideos10:44
RattusRattusmithro: I guess we could use a TPD2S017 on the SYNC lines and keep the TPD7S019....  still costs less than the PACVGA20011:13
mithroRattusRattus: Sure I guess?11:25
RattusRattusok.  /me will need to add symbol, will do this this evening when I get home11:26
mithroRattusRattus: okay, make sure you pull the latest changes before doing anything11:27
*** rohitksingh_work has quit IRC12:06
*** rohitksingh_work has joined #timvideos12:09
*** cr1901_modern1 is now known as cr1901_modern12:29
*** sb0 has quit IRC12:31
*** sb0 has joined #timvideos13:09
mithroRattusRattus: just pushed the latest update from the PCB designer to the VGA repo13:10
*** rohitksingh_work has quit IRC13:16
*** rohitksingh has joined #timvideos14:53
*** Bertl_oO is now known as Bertl_zZ16:31
*** rohitksingh has quit IRC20:24
*** sb0 has quit IRC21:27
mithroRattusRattus: any luck?21:48

Generated by 2.13.1 by Marius Gedminas - find it at!