video overlay text brightness

Go To Last Post
11 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Guys,

I am working with a DS1820 temperature sensor in assembly (everything working ok there).

I trying to display the temperature over video on one of the security cameras at my home. I am using an example I got from the net here:http://www.circuitcellar.com/avr2004/wentries/A3754.zip.

Just testing the video overlay code without any modifications except that I adapted the code to an attiny2313 (changed delays & usart settings) it is working.

The problem is that the brightness of the text is good for about two seconds, then starts to slowly fade until you can't see it anymore. I have played with the 120 ohm resistor, however it only makes a difference at startup, then the text slowly begins to fade away.

I have used the default crystal (10MHz) and tried the 14.318 MHz crystal and also a 20Mhz crystal, ALL ARE WORKING with the same problem, fading text. The higher frequencies give less flickering in the video.

Just wondering if anyone has an idea what the problem is? I have included the circuit here, the code is in the link above.

Thanks

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Add a 200..470 Ohm resistor in parallel with J3/J4. Your video monitor/TV appears to be purely AC coupled inside (w/o this DC shunt resistor).

Warning: Grumpy Old Chuff. Reading this post may severely damage your mental health.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry I don't follow you. You mean install a 200 - 470 ohm resistor from J3, J4 to GND and remove the 10K pulldown resistor (R1) between C3 and D1?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Just a single resistor from J3/J4 center pin to GND. I haven't said a single word about R1.

Warning: Grumpy Old Chuff. Reading this post may severely damage your mental health.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks for the help. Briefly touching the resistor to J3/J4 center pin to GND restores the text to its full brightness (depending on the setting of R5).

If left connected, it scrambles the text and even removes it completely. I am using a variable resistor. I tried everything from 50 ohms to 2K.

With the pot removed from J3/J4 to GND, the text slowly disappears within 20-30 seconds. Touching the resistor to J3/J4 to GND again restores the text for another 30 seconds.

I wanted to do it all on one avr, without having to spend $50 on a max7456 board.

Any help is appreciated.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Try connecting the pot in a 3-wire manner - edges to GND/VCC, and the wiper via 10k..100k resistor to J3/J4 center. Adjust the pot until your text is visible and stable.

Warning: Grumpy Old Chuff. Reading this post may severely damage your mental health.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Decoupling J3/J4 from R5/C3 by 0.1mF capacitor should improve situation, imho. DC shunt shall be in place, too.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks MBedder for the help. It is working with the pot & 10K resistor. Working without any flickering using an attiny2313 & a 20MHz clock. (Text is very small at this clock frequency, text is good at the original 10MHz, except too much flickering in the video signal)

The circuit only works on a very small Black & White TV. Tried it on the colour TV's here, no text displayed no matter what the pot setting is. Video has some abnormalities with the device powered up when connected to the colour TV. The circuit is also very particular in the input video signal.

With a Black & White security camera as the input video signal, no text is ever displayed, no matter what the pot setting is.

The only combination that worked was a VCR composite output (movie in colour) and a very small Black and White (5 inch screen) TV.

I have a max7456 breakout board on order from Sparkfun. I hope it is more reliable.

Thank you again for the help.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A couple of final notes then.
the real issues are two:
DC level of video signal and
DC decoupling.
I would strongly suggest hooking a scope to that working and not so working configurations. By connecting OVL You should no way change existing DC offset; this may be not the same for different in-out combinations.
For this quite a huge decoupling capacitor would be needed at OVL output.
The second thing imho is output impedance of OVL and amplitude of output signal when connected. Assuming this, a kind of emitter follower or other driver circuit should be necessary. OVL output should pump enough to exceed video amplitude.
Again, checking video by a scope with DC input is at the first place.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I would guess that, if you get flickering or no text at all, the DC offset problem is also affecting the sync detection circuitry.

Quebracho seems to be the hardest wood.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I am the original designer of that circuit. I have a few notes that may be of interest to you or anyone else trying to get it to work.
First, it is VERY sensitive to the incoming video signal level. The diode and two resistors used as a voltage reference create about 0.25 Volt reference input to the AVR comparator. If the sync level of the video doesn't consistently drop below that it won't be detected. There will be no video, except maybe some occasional random garbage/interference. It works well if the levels are just right, that is "standard" NTSC video level of 1V p-p. By replacing the two 10K resistors in the divider with a 10K pot you can adjust the sync detect level. I haven't tried it but I think that will help a lot. Anything more complicated and you may as well use an LM1881 sync separator, which is what I was trying to avoid.

The AC coupling of the TV/monitor is something I never encountered and didn't take into account. In addition to what has been mentioned, changing the value up or down of the video output resistor from the AVR might help.

And finally, the flicker is most likely caused by the mismatch of frequency between the AVR and the video signal. There are some hooks and constants in the original code for using a 14.31818 MHz crystal which should give a much better match. The ATTiny2313 was just becoming available when I built this, and I never got around to trying it. I didn't want to overclock the 90s2313 and 10 MHz worked, so I left it.

If anyone is or wants to use this, I would like to hear about it. I may even be able to offer some helpful advice :-).