Ambilight mit STM32F4discovery und TVP5150AM1

pitschu schrieb:
> sync lock and H

Waldemar schrieb:
> zoe schrieb:
>> Waldemar schrieb:
>> Maybe you have to Power the hdmi2av converter with USB?
>>
>> yeah I have powered it already. i think it's some kind of malfunction
>> because it's very hot, is there anyway for me to check with
>> oscilloscope?
>
> Check your converter with an old TV which has an AV Input

thanks both of you
I think that was problem
and not because my converter.
I have connected the converter with an HDMI signal from my second HDMI 
port on the VGA (my computer) and set the desktop mode to duplicate and 
extend
may be that was the problem. I will try it again and hope you will be 
here to help me
thanks again

hello, do you think the resistor and câpcitors between in put and in put 
chanel on the chip is the problem of video signal?

No, the resistors are needed to simulate an input impedance of nearly 75 
ohm, the capacitor is needed do decouple the input regarding DC voltage 
from the chip input. You should not remove them.

pitschu schrieb:
> No, the resistors are needed to simulate an input impedance of
> nearly 75
> ohm, the capacitor is needed do decouple the input regarding DC voltage
> from the chip input. You should not remove them.

thank you. how can i test the board tvp5150 if it has problems. i think 
the converter is fine because the computer regconized it

To use the board 'out of the box' with the software published here you 
need a video signal which is compatible to one of the supported video 
standards (look at the datasheet). The mode should be detected 
automatically but it's only tested with the European PAL 50Hz standard. 
Look at the logged status registers of the TVP, especially Status Reg 5 
shows the actual video standard which was detected.If your signal cannot 
be decoded you could try to change the initial settings for various 
registers. May be the documentation of your hdmi2av converter gives some 
hints of the output signal (Vpp + coding).

my HDMI2AV converter is the 1080p only. was that the problem?

the hdmi2av converts FROM hdmi 1080p to another AV format. This format 
should be supported by the ambilight processor. Do you have an 
osziloscope to check the output of the hdmi2av?

oh god. i think my converter not support 1080p input

pitschu schrieb:
> the hdmi2av converts FROM hdmi 1080p to another AV format. This
> format should be supported by the ambilight processor. Do you have an
> osziloscope to check the output of the hdmi2av?

I have and when i connnect the red wire of oscilloscope to the output. i 
saw like sin wave. after that. i connect the black wire to gnd. it 
straight out

I've check and buy a new converter. so the converter is not the problems
can you tell me is there any factor cause no signal to TVP5150AM1??
(capacitor, resistor?) the comunicate between 5150AM1 and stm32 board is 
ok

I got exactly this problem like Pat Wat has before, no video signal
can anyone check is there anything fail in this schematic and PCB? 
(atached)

I'm trying to replace 39 ohms with exactly 37.4 ohm and see what happens

zoe schrieb:
> I'm trying to replace 39 ohms with exactly 37.4 ohm and see what
> happens

I'm sure that this will not solve your problem. Can you measure the 
video signal directly at he input pin of the TVP and post an screen 
image here? Your log shows, that the video status changes from 61, 71 
and sometime 73. This means that no sync could be detected.

The file formats in your rar file are not usable for me. Can you make 
PDF files of the schematic? Especially the TVP input and connection to 
the STM32 are important.

pitschu schrieb:
> zoe schrieb:
> I'm trying to replace 39 ohms with exactly 37.4 ohm and see what happens
>
> I'm sure that this will not solve your problem. Can you measure the
> video signal directly at he input pin of the TVP and post an screen
> image here? Your log shows, that the video status changes from 61, 71
> and sometime 73. This means that no sync could be detected.

you want me to put the video signal directly to tvp5150 video input pin? 
bypass the resistor and capacitor?

or you want me to measure with Oscilloscope?

im getting an 8ch logic analyzer and will post the screen shot
thanks

Zoe schrieb:
> im getting an 8ch logic analyzer and will post the screen shot
> thanks

No, simply feed the video signal into the board (including the resistors 
and capacitor) and measure the resulting signal at the input pin of the 
TVP. You cannot use a logic analyzer for that because it's a analog 
signal. A logic analyzer can only measure digital signals.

this is when i only connect the red wire of ócilloscope to the input pin 
chanel 1 of tvp5150( the strange is it showed the same when connect to 
chanel 2) so both no signal

and this when i connect the black wire of oscilloscope and the red to 
AGND and input chanel on chip

Zoe schrieb:
> and this when i connect the black wire of oscilloscope and the red
> to
> AGND and input chanel on chip

You have to measure always with connected GND. This means you have no 
signal on your AL-Board. Measure the output (yellow chinch) of your 
hdmi2av converter

this is when I power up the converter with usb port
black wire to AGND
red wire to input chanel
(actualy the converter no need power support)

and this is from yeallow chinch (disconnected with the board)
( i changed the voltage of the oscilloscope)

We can not open your schematic files, please post a PDF of it, because 
the issue is on your board....

The FBAS-Signal of your converter looks fine

Waldemar schrieb:
> We can not open your schematic files, please post a PDF of it,
> because the issue is on your board....
>
> The FBAS-Signal of your converter looks fine

ok bro

here is the pdf. if you need anything i will convert

what if my TVp5150 chip malfunction??
I got it from china,

I used altium to wiring

my quartz

I'm not sure whether you soldered the TVP right. It seems that you 
connected the input signals to pins 17 + 18 instead of 1 + 2. But I 
can't really check it with your picture (could you please mark the pin 1 
on the picture).
Does the crystal 143F3I really have 14.31818 MHz?

I will check the crystall again
I think the tvp5150 is soldered right because it respond to the stm32 
when i press N in putty. it reset and report things about device id 
..etc

Yes, you are right, the I2C bus seems to be working. But the signal on 
pins 1 or 2 looks wrong on your scope screenshot. It should look exactly 
like the signal out of the chinch socket (second screen shot) with only 
half of the amplitude. Maybe you damaged the TVP chip.

strange... if you can measure a video signal on your converter, then you 
should measure a signal at PIN1 / PIN2 from TVP if the converter is 
connected to the board. maybe check your wiring on board with multimeter 
to capacitor C13 and from C13 to Pin1 of TVP and also the same for 
INPUT2

Waldemar schrieb:
> strange... if you can measure a video signal on your converter,
> then you should measure a signal at PIN1 / PIN2 from TVP if the
> converter is connected to the board. maybe check your wiring on board
> with multimeter to capacitor C13 and from C13 to Pin1 of TVP and also
> the same for INPUT2

yeah i have measured it and it show like this. just a little bit lower 
than the original
output from yellow chinch on converter

it's only half because 2 39 ohms resistor reduce the voltage under 0.78 
volts, right?

That's right, but the peaks to GND are missing. These peaks (you should 
see them at least every 20ms with your 'scope') are the X/Y sync signals 
are are important for correct video detection. The status log shows, 
that the sync signals are missing and the scope shows, that it's true 
:-)
Please check the 1.8V and 3.3V supply of the TVP.

this picture is when I soldered all FBs with solder bridge, the signal 
from AGND and input chanel 1,
I think my TVP chip was gone, i will try to replace it and check again

i used 1uh FBs for FB1 2 and 3,

the 1.8 voltage supply is 1.69 volt measured
the 3.3 is 3.14

zoe schrieb:
> the 1.8 voltage supply is 1.69 volt measured
> the 3.3 is 3.14

not good... but i don't know that this coases the problems

Datasheet says min 1,65V. Maybe you have some error in your measurement 
and then you get in trouble

Is the voltage regulator getting hot? Maybe your TVP is bad and draining 
to much current, so this will cause the lower voltage on 1,8V rail

have you looked with your oscilloscope: are the voltages stable?

Waldemar schrieb:
> zoe schrieb:
> the 1.8 voltage supply is 1.69 volt measured
> the 3.3 is 3.14
>
> not good... but i don't know that this coases the problems
>
> Datasheet says min 1,65V. Maybe you have some error in your measurement
> and then you get in trouble
>
> Is the voltage regulator getting hot? Maybe your TVP is bad and draining
> to much current, so this will cause the lower voltage on 1,8V rail
>
> have you looked with your oscilloscope: are the voltages stable?

the oscillo show all the voltage not stable. i dont know why, it show 
min max and average. 1.69 is average, and 3.14 too

the tvp 5150 is normal. no hot, im ordering 10 more tvp to find out what 
is the problem :D

Zoe schrieb:
> the tvp 5150 is normal. no hot, im ordering 10 more tvp to find
> out what
> is the problem :D

please post photos of your measurements of 1,8v and 3,3v supply where we 
can see the instability of the voltage. i asked if the 1,8v regulator 
(LM1117) is getting hot or not

this is the video about 3.3 supply
measured gnd and 3.3v pin
https://youtu.be/Ep6mfYqr_XA

and this is 1.8
i change the adapter and the 3.3 is 3.22
1.8 is 1.69 and nothing happen
https://youtu.be/lGX7lG6fv4U

Hallo Frank

Hättest Du noch eine Platine über?

Wer sonst noch eine abgeben möchte, gern auch bestückt - ich hätte 
Interesse.

Viele Grüße
Jens

Hallo Jens,

vor etwas über einem Jahr habe ich mir von pitschu eine Bestückte 
Platine incl. SW gekauft. Bin dann aber nicht dazu gekommen das Projekt 
fertig zu stellen. Wenn Du willst kannst Du sie haben.

LG EDE

EDE schrieb:
> Wenn Du willst kannst Du sie haben.
>
> LG EDE

Hallo EDE

Zur weiteren Absprache melde Dich bitte mal unter ambilight -at- 
ist-einmalig.de.

Viele Grüße
Jens

Hallo zusammen,
Wegen Umstieg auf original Philips und fehlender Zeit verkaufe ich meine
2 Sets Ambilight.

1. Set beinhaltet das STM32F4, das bestückte und getestete 555 Board,
HDMI Splitter, HMDI-> Analog Converter, 5V 5A Schaltnetzteil (Meanwell)
und 5m WS2812B (neu) 60 LEDs/m und Gehäuse.

2.Set beinhaltet das STM32F4, das bestückte und getestete 555 Board und
einen ALU-Profil U Rahmen mit darin verlegten WS2812B 60LEDs/m passend
für TV ab 42 Zoll (wurde bei 55 Zoll verwendet).

Bitte nur seriöse Angebote unterbreiten. Fotos kann ich bei Interesse
per Mail versenden.

hello pitschu
after 1 busy month i got back to this board
change the 5150 AM1 chip vendor and got this like you said
status 0E

the log lock 0E and nothing show out to the led.led is off
what is the problem here? the tvp board is working right? because the 
video status is 0E

finally the led turn on and reacto to the screen but with the weird way
the green on the screen is the white, the color is not true and only 
afte the log show
IR data: 0040, rep count: 20
Mood/Ambi mode switched to 02
IR data: 0040, rep count: 20
Mood/Ambi mode switched to 02
IR data: 0040, rep count: 20
Mood/Ambi mode switched to 02
IR data: 0040, rep count: 20
Mood/Ambi mode switched to 02
IR data: 0040, rep count: 1
Mood/Ambi mode switched to 01

the led turn on,
otherwise. it's off
help

ah yes it's happen when i press the button on stm32 board. then log show 
ir data received and led turn on. the black is pure green (wrong color)
and it's not stable
this is the log
IR data: 0040, rep count: 1
Mood/Ambi mode switched to 01
q
Current dynRows:
 AVG:  89 107 111 115 112 114 105  96  96  96  97  95 102 100  96 103 
83  94 12 160 137 153 152 152 153 157 138 129 29 130 131  76  91  96   0 
0   0   0   0   0
 CON: 197 341 350 332 313 302 302 209 203 208 214 214 212 199 191 196 
160 200 245 271 260 291 288 287 282 283 269 267 269 260 270 146  69 110 
0   0   0   0   0   0
 ACH:   0  18   4   4  -3   2  -9  -9   0   0   1  -2   7  -2  -4   7 
-20  11  30  36 -23  16  -1   0   1   4 -19  -9   0   1   1 -55  15   5 
-96   0   0   0   0   0
 CCH:   0 144   9 -18 -19 -11   0 -93  -6   5   6   0  -2 -13  -8   5 
-36  40  45  26 -11  31  -3  -1  -5   1 -14  -2   2  -9  10-124 -77 
41-110   0   0   0   0   0
 dynTop = 0000, dynBot =   33
 Black level =    0


Current dynCols:
 AVG:  45  48  46  45  45  45  45  45  46  47  45  55  56  57  57  57 
57  55  50  56  58  57  63  69  66  65  65  62  51  55  55  55  58  52 
49  50  50  48  50  64  65  68  71  71  75  91  99 102 101 100  95  89 
84  82  75  65  56  56  56  55  55  54  56  44
 CON: 140 153 154 150 145 147 146 153 160 179 185 229 281 290 296 302 
300 288 292 298 295 295 283 323 316 325 317 307 249 286 303 315 321 256 
248 254 252 225 246 323 330 335 339 341 358 396 395 404 404 400 391 381 
347 336 303 258 199 204 207 202 204 190 174 118
 ACH:   0   3  -2  -1   0   0   0   0   1   1  -2  10   1   1   0   0 
0  -2  -5   6   2  -1   6   6  -3  -1   0  -3 -11   4   0   0   3  -6 
-3   1   0  -2   2  14   1   3   3   0   4  16   8   3  -1  -1  -5  -6 
-5  -2  -7 -10  -9   0   0  -1   0  -1   2 -12
 CCH:   0  13   1  -4  -5   2  -1   7   7  19   6  44  52   9   6   6 
-2 -12   4   6  -3   0 -12  40  -7   9  -8 -10 -58  37  17  12   6 -65 
-8   6  -2 -27  21  77   7   5   4   2  17  38  -1   9   0  -4  -9 -10 
-34 -11 -33 -45 -59   5   3  -5   2 -14 -16 -56
 dynLeft = 0000, dynRight =   63
 Black level =    0

aaaaaaaaaaaaaaaaaaaaaaaaaaaah yeasssssssssssss everything is working 
now. just solder wrong pin E6 so all green color.
thank you all

Great project! Did you try "Embedded Synchro" mode with DCMI on 
STM32F04? I think you used the "8-Bit 4:2:2 With Discrete Syncs" mode of 
TVP5150, fed that to STM32 together with VSYNC / HSYNC lines, in 
"External Synchro" mode.

This is exceptional property of TVP5150AM1 that it can strip out ITU-R 
BT 656 embedded sync codes (SAV / EAV) out of data stream. DCMI does 
support embedded sync – no VSYNC, HSYNC connections are required. 
However – not in interlace mode! Only in progressive mode (8-bit). 
That's my problem. I want to use ADV7280 because it allows component 
input YPbPr, not only CVBS and S-Video. But it cannot strip out the 
embedded codes SAV / EAV. It however allows interlace -> progressive 
conversion – at cost of 54 MHz pixel clock (2x 27MHz). So I'll probably 
be fine, will use "Embedded Synchro" with 54 MHz input (supported by 
STM32F767). I'm just curious of your experience, because it's the first 
post on the internet that successfully integrated video-decoder IC with 
STM32 DCMI. Did you try embedded synchro mode with DCMI?

Hallo. Beide Sets sind noch zu haben.

Ausserdem verkaufe ich noch einen
ANALOG DEVICE ADV7612 Dual Port, Xpressview, 225 MHz HDMI Receiver

http://www.analog.com/media/en/technical-documentation/data-sheets/ADV7612.pdf

NEU und unbenutzt.

I am trying to get data from tvp5150 and send it to LCD. I have 
horizontal and vertical  synchronization problems. Seems like synhro 
isn't generated properly (in right time) by TVP. Do you have any idea of 
this?

Hallo,

ein großes Lob an alle, die zu diesem Projekt beigetragen haben, vor 
allem an pitschu!
Habe das Projekt mit China-Bestellungen im Wert von unter 50 € pro 
pitschulight umgesetzt. STM32-Board, PCB und LED-Stripes mit 60 LEDs/m 
sind bereits eingerechnet. Die ersten beiden fertigen Boards 
funktionieren tadellos.
Bevor die Fragen aufkommen: Ich möchte keine überschüssigen Teile und 
auch keine fertigen Leiterplatten verkaufen. Musste bestimmt über 300 € 
ausgegeben, um alle Teile für die ersten 5 pitschulights zu kaufen. PCBs 
und Bauteile sind für insgesamt 10 Stück vorhanden, für die Bastelkiste 
wird am Schluss auch noch was übrig bleiben. Ein einzelnes Exemplar kann 
man sich also nicht für 50 € zusammenbauen.

Wenn Interesse besteht, kann ich die Artikel vom Auktionshaus mal 
zusammenschreiben, das PCB Layout kann ich auch weitergeben. Allerdings 
ist das Layout grottenschlecht, da zum Großteil vom Eagle Autorouter 
übernommen.

Nun zu meinem eigentlichen Anliegen: Habe ein STM32-Board mit 
STM32F407VET6 (512 kB Flash). Die hex-Datei aus dem Beitrag unten läuft, 
ich kann aber keine Änderungen am Code vornehmen. Z.B. möchte ich die 
IR-Codes ändern, da ich noch ein paar Wochen auf die Fernbedienung aus 
der Anleitung warten muss. Kann bitte jemand den Quellcode für 512 kB 
zur Verfügung stellen bzw. die nötigen Änderungen mitteilen? pitschu 
vielleicht?

Bereits jetzt vielen Dank!

Michael

Peter S. schrieb:
> I just tried the HEX attached in this post on a brand new
> STM32F4dicso
> board. I made some changes for your CPU (only 512K flash). I'm sure it
> works and USART out is on PD8. Please use this HEX and tell what's
> happening.
>
> pitschu

hello pitschu, I've sent you an email. please check it
thanks :D

Hello everyone, It's me bothering you guys again
It's been a long time I haven't touch the Peter's project,
I decided to modify the code just to make the board works with Ws2811 
12v LED (which has RBG order instead of RGB as ws2812B) I used custom 
stm32f407vgt6 board and everything worked when flash Peter's original 
code (terminal, LED mode, IR...)
after edit WS2812.h (to swap G and B value) everything worked fine 
except USART3, no responding and transmitting  at all.
at first I thought it's because I edited the library but after revert to 
original, USART3 still dead, I re-flashed peter's original Hex and 
everything worked again
-I've never get USART3 (pd8 pd9) work on my own compiler,just Peter's 
hex file worked
-anything to work with these warnings below?

FB1 nd FB3 are open (3rd image) so no 1.8vA and AGND for tvp5150. just 
bridge them with solder or use ferrite beads

Good day.
Pitschu, your project is great!
I've read the forum but still don't understand is it open source project 
or not =)
You has uploaded scheme, .hex code.
But there is a site with commercial boards:
http://www.keiang.de/Content-pid-61.html

I have F4-disco board, some stm32f407vgt6 mcu and I can produce pcb by 
scheme from your post.
Need I extra source files or .hex only?
By the link below there is a last firmware that is assume presence of 
bootloader http://www.keiang.de/Downloads-op-view-lid-83.html

Answer please =)
Best regards, Yuriy

Hi, pitschu! =)
My TV use only)
I was searching for the easiest solution and everywhere was foundes 
Raspbery Pi.
Your project most acceptable for me)
You use connectivity like on the image and 
"Ambilight-STM32F4-GNUARM.hex" firmware, right?.
Does your project need some configurations?
How many WS2812 leds I can use?

Best regards, Yuriy

Good day!
Pitschu, I just want to thank you for this design :)
Everything works, but there are minor glitches.
Please tell me, because of what can there be a long response and a 
periodic reset of the display?
Sometimes between the change of image and the change of backlight, up to 
two seconds pass.
It is rare to turn off the backlight (feeling like a signal malfunction) 
and turn it on after a few seconds.
At the same time, the area for display on the screen is not black, but 
the STM32F407 works (I look at 4 LEDs).
At the moment, it is not possible to connect through the terminal while 
the entire structure is working, therefore
I don’t know the state of the controller during the failure.
I am attaching an archive with a video and a couple of pictures (I 
apologize for shaking hands))).


I connect it like this: PlayStation4-> HDMI Splitter-> HDMI to RCA 
converter-> PCB + DiscoF4.
I tried to switch PAL / NTSC on the converter and did not notice the 
difference.
Directly to connect to the TV via SCART-> RCA is not yet possible, I 
will soon buy an adapter and try.
I use only 3 zones, without lower backlight. 16 LEDs high, 28 wide.
Power supply 220  12V and step-down stabilizer 12  5V 5A.
I also tried to power it from the laboratory power supply KA3005D, 
failures still remain.

I configured the board through the terminal, the settings are as 
follows:
H = 270, T = 30, W = 720, L = 88,
P = 28, R = 16, A = 1, M = 0,
F = 0, S = 128, B = 85, C = 90
I = 90, E = 5, Y = 40, X = 64.

Tell me what I can check.
Thank you in advance)
Best regards, Yuri

Hi Yuri,

I think the delay time of your device chain (Playstation -> splitter - 
converter -> Ambilight) could be long, depending on the type of the 
devices. One thing you could do in the Ambilight is to increase the 
I-parameter to 128 (= instant display without integration).
Try to lower the G-parameter (# of frames for black-border-detection). 
The default is 100 frames. Try smaller values.

pitschu

Good day.
Pitschu, thanks a lot for your project!!!
Everything works perfectly! =)
Finaly I use the following parameters:
I = 90, G = 80.
It was useful for me to reduce the length of the wires to the LED 
backlight.
Also added a 100pF capacitor from PE9 to GND.

Thanks again!
Regards, Yuri

Hi Dear,
Firstly thanks for good project.
I want to use project could anybody share last version schematic and 
source code for project.

Best Regard
Otto