- This topic has 12 replies, 4 voices, and was last updated 11 years, 6 months ago by
.
-
Posts
-
Hello,
I am having a problem that I haven’t seen here yet – hopefully I didn’t miss the question being asked already.
The video overlay is very light – almost not visible. If I switch it from overlay to sync only, the text and simulated graph is very easy to see – but as soon as I switch it back to video overlay – then it ghosts almost to the point of not being able to see it at all. I did verify with two different displays with the same result.
The video source is a 5.8ghz remote video feed – the video is crisp either through the shield or direct, but can just barely make out the overlay.
I have also tried switching the mode from NTSC to PAL in the code and re-ran with same results.If I try a sketch with it sourced from the D9 and shield set as such, video overlay is out of sync/unreadable until I switch it from overlay to sync – at which point I can see the text and graphics from the included demo tv sketch.
Any support would be appreciated!
Thanks. 🙂Are you overlaying onto a very bright image? Is overlay light even on a dark image? I would expect the overlay to be washed out on a bright image…
@Michael wrote:
Are you overlaying onto a very bright image? Is overlay light even on a dark image? I would expect the overlay to be washed out on a bright image…
Hello Michael
I tested it as follows:
Living room
Outside
In dark closetAll three have the same result – almost unreadable display. Im not exactly a EE, but I have been a hobbyist for quite some time, so I am not afraid to test/try various options and ideas you may have.
Thanks for your quick response – hopefully we can figure it out. 🙂
With the switch set to sync only, the white pixels are bright and opaque? But when in overly mode and the camera image is black (e.g. hand over camera) the pixels are faint gray? These two cases should not differ.
Did you buy as a kit or assembled? If as a kit, have you checked to see if you put the resistors in the right places?
Bought it as a kit – and yes the resistors are all in the correct place. Verified with a fluke that resistors etc have the correct values etc… So not a bad component that I can test (I also tested components before assembly – resistors and pots).
Yes, it is as you describe above as well – as long as there is a video signal it is almost unreadable.
Odd. Have you tried different video sources? It’s almost as if the voltage levels are too high from your source and the TV is scaling it down.
Have you tried different TVs?
You should find 75ohms resistance between the TV input and ground. This is standard and internal to the TV.
I will ohm it out – curious about that.
But yes, two different TV’s and a portable DVD player as well. I haven’t tried a different video source however, will need to try and find something that has composite video out, only thing I have currently is for my RC FPV stuff.
ThanksI have fabricated a circuit card with an ATMEGA328 micro-controller and an LM1881 Sync Separator with a hardware design similar to a combination of an Arduino Uno and Video Experimenter. This design has worked great (thanks to Nootropic for an excellent design) for a variety of video sources. Recently I purchased a low cost camera sensor ($15), and I experienced an identical problem to that described in this post. The video overlay was barely discernible when overlaid on either a bright image or dark image. Looking at the video output from the camera sensor on the scope, I observed a DC bias. The vertical/horizontal sync pulses were at +.25 volts and bright images were running between +1.2 and +1.5 volts DC. These video amplitudes swamped out the overlay pulses that have an amplitude of +.8 volts DC. The solution is to reduce the value of the series 330 ohm resistor connected to pin 13 of the ATMEGA328 (Pin D7 of the Arduino Uno) through a diode. For a supply voltage of +5 volts, a series resistance of 150 ohms will increase the amplitude of the overlay pulses to approximately 1.5 volts (based on the monitor having an input resistance of 75 ohms and a diode in series with the resistor). This value of resistance results in the processor sourcing 20 mA which is well within its rating of 40 mA max.
ricki, thanks very much for sharing your fix. It seems that some video sources provide a voltage that is too high relative to the signal voltage generated by the Video Experimenter. The solution is to reduce the resistance of the 330 ohm resistor on the board. For those that have a Video Experimenter, you could add another 330 ohm resistor in parallel to the existing one (just solder the leads together). This effectively halves the resistance to 165 ohms which will allow a higher overlay voltage to be given to the TV. Thanks again, ricki.
Hi Michael,
For those that have a Video Experimenter, you could add another 330 ohm resistor in parallel to the existing one (just solder the leads together)
Which specific resistor on the VE board are you referring to? I’m colour blind and resistors are something of a challenge!
Many thanks,
alec
- You must be logged in to reply to this topic.