I just bought one of these from Sparkfun, put it together, and stuck it on top of an Arduino Uno.
I tried asking a question in the Wayne & Layne forums, but things seem pretty dead over there.
Anyway, it seems to work ok on a crappy old CRT TV I have. The NTSC demo runs and draws a rotating cube.
However, when I connect it to my big LCD TV, the screen is just blank. For a fraction of a second, you can see the shield output, then the screen goes blank. I’m assuming the LCD TV thinks there is no video signal and is blanking the screen? How can I get it not to do this?
I only thing I can think of checking is the voltage levels of the NTSC video signal. Either the LCD TV whats a larger signal levels or is loading the Shield output.
Take a look at the voltage levels using an O’scope when the Shield is connected to the OLD CRT verse Lnew LCD TVs.
I’ve only got a DSO203 and it looks like it can’t keep up with the PWM? I can’t get it to trigger properly, but what I see flashing by are a lot of 3 to 4 volt spikes and the occasional 5-6V spike. With the timescale of about .5ms. (At the shortest timescale, 100ns, there is just some kind of wierd, goofed-up-looking square-ish wave with a lot of ringing.) The scope looks the same when I have it hooked up to the big TV or the little one: lots of little spikes and the occasional big one. I have the DSO hooked up to pin 7, which is labeled “Video” on the schematic.
I’m not entirely sure my VGS is working right: there could be some problem with the nunchuk stuff. I don’t have one to test. With no controllers, the Conway’s Game of Life demo won’t run unless I comment the nunchuk stuff. I did all the 4066 diagnostic stuff they have in the W&L forums (with the DSO, not a LED) and it passed every test.
Anyway, the LCD TV definitely blanks the screen, say .3 secs after I plug in the cable. I can see that the (Conway) sketch is still running, because removing and reinserting the plug gives me another brief view of the screen.
I am plugging the VGS output into the yellow RCA jack marked Y/Video. I’ve used that jack before with a DVD player, so I know it’s working.
Even free-running and under sampling you should get something like the proper voltage. Those don’t seem correct to me and yet it works with the CRT TV. I wonder if the output has some corrupting noise on it that’s getting filtered out by the analog TV but is confusing the crap out of the LCD TV ?
Just a wild idea. Place a 75 Ohm resistor across the Shield’s video output and try both TVs again (also look at the signal with your scope to see if there is a difference).
Another thing to try a a small value cap (~100pF) across the video output.
waltr:
Just a wild idea. Place a 75 Ohm resistor across the Shield’s video output and try both TVs again (also look at the signal with your scope to see if there is a difference). Another thing to try a a small value cap (~100pF) across the video output.
You’ve sparked an idea in my head ! I wondered how the shield makes it’s video signal and that’s all explained here.
Note the part where the 2 PWM outputs are combined. I really do expect that the TVs will provide the req’d 75 ohm termination so what else could be wrong ? Well if R1 and R2 (above) are the wrong values then the output will be wrong.
Looking at the [schematic, “R1 and R2” (above) are really R4 and R5 for the sync and video signals (respectively). They are supposed to be :
The more I think about it, the more puzzling this is. I can’t see getting the full 5V peak signal when plugged into both TVs and have this work at all on the CRT TV. Where was this signal measured ? At the P2 test point or ??
You say you can see the expected output briefly on the LCD but then it blanks. That says to me there’s enough signal there but something else goes wrong. The only thing I can now imagine is that the black level output is somehow dipping below the LCD TV’s blanking threshold (but is above the CRT’s). Perhaps you can measure the output signal levels.
You should be able to see the sync period with your o-scope. Set the trigger for neg edge and the threshold to be just a bit > 0 V. You should see a signal that grossly resembles …
… with a period of about 64 usec. The video signal btw the blanking intervals should be bouncing btw 0.3 and 1V. See what you see but I’m curious about what that lower level is. Be sure to connect the o-scope gnd to the Arduino gnd. Perhaps use an ohmmeter and verify continuity btw the RCA jacks gnd contact (outer barrel) and the shield ground.
OK, I was off swimming in a spring the last couple days.
Getting back to it, here is a shot of what my DSO203 screen looks like with the probe on the pin 9 SYNC and with conway’s game of life displaying on a little TV:
The probe is on the green wire and the ground is on the white one.
It doesn’t matter what I set the triggering level or mode to on the DSO203, the messed up signal just dances around. I know the DSO works ok b/c I tested it with a signal gen. and have used it for other really slow signals–mostly feeding it slow square waves for figuring out which pin is which on confusing level shifter ICs.
This is as stable I can get it at 10 us/div:
I’ve triple checked my soldering job and it looks ok and in agreement with the W&L instructions. Hell, I’ve soldered much harder kits than this.
Just for fun I tried another old CRT and it works fine on that one, too. There is a difference between the 2 CRTs and the LCD TV: they have only one yellow video jack and the LCD has three, none of which are yellow. (Green “Y/Video”, Red “Pr”, Blue “Pb”).
If this is all that is wrong with my shield, it’s not a show stopper b/c I actually prefer the crappy little CRTs. I have a whole bunch of them for my retro game systems.
Does the reisitor mentioned a few posts up have to be exactly 75 Ohm? The only thing I have in the ballpark is a 12 Ohm and I’d hate to have to string together six of them. My next biggest is 460.
If you probe pin 7 (video) or pin 9 (sync) what you’re going to see is the PWM output(s) from the Arduino. They should both be a 0 - 5v signal. I think your problem may be that your LCD TV doesn’t have a composite video input, or at least not a default one. The RCA jacks you describe are component video inputs. The Y input is the luminance input, the same as the B/W composite signal, I think. Perhaps there’s a voltage scaling difference ??
In any case you might want to peruse your TV’s setup menu and see if there’s a way to designate those jacks as component or composite. You want the latter if it’s there.
ps - if you want to see the video signal being output to the TV, use the test point (P2), which I think is the one below.
Your o-scope shot of the sync pin looks good (@ 10 usec/div). It’s a pulse every 64 usecs.
I think your problem may be that your LCD TV doesn’t have a composite video input
In any case you might want to peruse your TV’s setup menu and see if there’s a way to designate those jacks as component or composite. You want the latter if it’s there.
I think I have it !?! The component luma signal is the same as the composite BW luma signal but it’s shifted down by ~285 mV. That is, the sync pulse for component luma is expected to go to -285 mV where as the composite only goes to 0V. So if the LCD TV is expecting component levels (but getting a composite signal), it’s probably not seeing a sync pulse and blanking as a result. IF the TV can recognize composite levels AND can be told to do so, it might well work.
