Page 1 of 3

UZEBOX AVCore up and running!

Posted: Sat Sep 13, 2008 11:53 pm
by havok1919
I like it when things work first try. :D

Other than being outsmarted by the AD723, I didn't have any problems getting things running. (I was trying to probe the outputs from the AD723 with a scope before hooking anything up-- nada, nichts, nothing... Finally remembered that the AD723 disables the outputs if it doesn't sense a load from a TV on the other end. :oops: )

The board is seriously small. Hard to appreciate until you see it next to a standard RCA connector.

Anyway, I'm still not sure if my zener clamp will be OK for the SD card or not (I haven't fully populated everything yet). Video output looks good-- stable, no funny stuff. I couldn't find my 75 ohm resistors, so I used 68's instead so my video levels are probably a little low at the moment. The shot below is composite video only, no trap yet. (could only find 15pF caps too... I need to reorganize my reels!)

So, I'm pretty stoked at the moment. :lol:

Image
Image
Image

-Clay

Re: UZEBOX AVCore up and running!

Posted: Sun Sep 14, 2008 12:18 am
by ladyada
awesome!!!
i cant wait till my boards come in *yeeek*
there will be much uzzing

for the sd card zener clamp, if you're running at 5v, here is a schem that is known good
http://ladyada.net/media/wavshield/wave ... 0schem.png
i dont know if you posted your schem so i cant tell what your zener is doing

Re: UZEBOX AVCore up and running!

Posted: Sun Sep 14, 2008 4:53 am
by CompMan
Looks great, I can't wait to get my parts.

Compman

Re: UZEBOX AVCore up and running!

Posted: Sun Sep 14, 2008 5:09 am
by havok1919
ladyada wrote:for the sd card zener clamp, if you're running at 5v, here is a schem that is known good
http://ladyada.net/media/wavshield/wave ... 0schem.png
i dont know if you posted your schem so i cant tell what your zener is doing
Yeah, I just have a 3.3V zener with a ~300ohm resistor in series on MOSI to clamp it to 3.3V. Signal looks OK at a ~1.8MHz ISP clock. In my enthusiasm to get a fab ordered I forgot to do something similar for SD_CS and SD_CLK. :oops: (I was mainly thinking about signal integrity-- I've had issues with the AT91SAM9261 running the SD card interface above ~15MHz. I confess that I haven't checked what the max SPI clock rate would be from a ~28.6MHz AVR though... I was a little worried about just using resistor dividers on either the clock or data lines (dunno what the capacitance on the uSD might be like and how the slew rate will be affected). How fast have you run your SD interface?)

For SD_CS I was thinking of just putting a pullup to 3.3V and then use the AVR's GPIO as a tristate. (AKA, write a '0' to the output latch and leave it there. Then to select the SD card write to the data direction register and select it as an output (sinking the 3.3V to ground through the AVR) and to deselect the SD card set the data direction register as in input and let the pullup yank it back up to 3.3V. Saves a diode or resistor.)

-Clay

Re: UZEBOX AVCore up and running!

Posted: Sun Sep 14, 2008 5:29 am
by ladyada
i was running the SD interface at 'max speed' on a 16mhz avr, so 4mhz. an earlier rev had zeners instead but i that found that the zener's capacitance slowed down the signal so i used resistors and it is nice and clean
the only reason i had one on CS is to allow use of a pullup so programming the avr wouldnt wonk the SD card. your idea would probably work just fine...i wanted mine a little bit more foolproof since it was designed for use by fools :D

Re: UZEBOX AVCore up and running!

Posted: Sun Sep 14, 2008 8:18 am
by nubie
Looks awesome :).

Re: UZEBOX AVCore up and running!

Posted: Mon Sep 15, 2008 12:53 am
by uze6666
I'm a very happy man...the first 'third party' Uzebox up and running! :mrgreen: I couldn't help myself, I had to promote this topic to 'Anouncement' ;)

Congratulations Clay, it's just perfect! And even using your balanced DAC, the colors appears just like mine. I'd like to see some pics from the main game, since there a lot of blue in there, but I think it will be just fine too.

From another topic:
Interestingly enough, at room temperature my Mega644 starts up and runs from a 3.3V (and even lower) supply just fine. I suspect that 5V will give plenty of margin for the clock.
We're you saying that the 644 was running fine overclocked and at that voltage...for the AVCore? :shock: That would be awesome and very good news if thats the case. It could reduce the part count and lower the cost a bit further.

I also like the AVCore name you just coined. I pronounce it "hav-core" and it sound just great, even evokes "AVR" a bit. :P Very nice indeed.

Best regards,

Uze

Re: UZEBOX AVCore up and running!

Posted: Mon Sep 15, 2008 4:25 am
by havok1919
uze6666 wrote:I'm a very happy man...the first 'third party' Uzebox up and running! :mrgreen: I couldn't help myself, I had to promote this topic to 'Anouncement' ;)

Congratulations Clay, it's just perfect! And even using your balanced DAC, the colors appears just like mine. I'd like to see some pics from the main game, since there a lot of blue in there, but I think it will be just fine too.
Thanks! I need/want to make a few tweaks still. I had CXA2075 NTSC encoder on the brain and forgot to tie the VSYNC input on the AD723 high (the '2075 only has CSYNC in). That might have been what I was thinking was a reset issue-- since fixing that it seems to come up fine everytime (and I've tried a few different power supplies). I'm running it off USB at the moment with it PIP'd on my monitor. Kinda fun.

I also spazzed and didn't include level shifters on the SD clock and chip select. Ladyada says that resistor dividers worked fine for him/her, so I'm just going to do that instead.

I integrated a small R/C filter on the audio output. On my old Magnavox/Commodore monitor it sounds fine without it, but that's a little paper cone speaker that's firing sideways-- probably making for a very good high pass filter. ;-)

Lastly, I decided to muck with the pinout a little bit on the stamp. I convinced myself that it'd be nice to be able to get at the RGB signals (for driving LCDs, component video conversion, etc.), but there was just *no* way to route those to the existing free pins without a 'do-over' in that entire corner of the board. Turns out there was a much more 'routing friendly' pinout to use, so I made the change. It puts RGB and CSYNC right next to each other which is handy. OTOH it splits the joypad inputs around a bit more than I'd like, but it did have the benefit of getting the joypad clock trace away from all the analog stuff, so all in all I think it's a win.

I'm going to hand mod one and make sure that I can talk to the SD card OK, then I need to test out Y/C outputs, but otherwise I'm not expecting any surprises. I'll probably order another set of prototypes, although I'm thinking of gambling and just going straight to a production spin.
We're you saying that the 644 was running fine overclocked and at that voltage...for the AVCore? :shock: That would be awesome and very good news if thats the case. It could reduce the part count and lower the cost a bit further.
On the one that I've built, if I give it a 3.3V power rail instead of 5.0V it does indeed seem to work. Granted that was just a room temperature test on a sample of one, but it would seem encouraging anyway. There's an easy spot on the board where a zero ohm 0603 (or a bit of wire) can go that would effectively short the +5V rail to the 3.3V rail-- so to get 3.3V operation (although possibly at a reduced operating range) it would just take replacing the pulldown resistors with higher values (say ~121ohms instead of 75).

Actually, that's not a bad way to go-- since I brought the RGB out directly to the pins, an "other voltage" version could just not have the pulldown resistors and those could be loaded by the user on their baseboard, etc. by connecting the resistors to the RGB pins. It would also allow for a few pots there if people want to have variable outputs (to push a hotter signal to an RGB monitor that can take 1V P-P for example).

3.3V would make a big difference for battery operation. (ie, I see about ~90mA @ 5V running tetris and driving composite video out. At 3.3v it's less than half that.)
I also like the AVCore name you just coined. I pronounce it "hav-core" and it sound just great, even evokes "AVR" a bit. :P Very nice indeed.
Thanks-- I thought it had a decent ring to it.

-Clay

Re: UZEBOX AVCore up and running!

Posted: Fri Sep 19, 2008 5:03 am
by uze6666
At the risk of repeating myself, I just looked (again) at that pic with you holding the module with your fingers...and now I realize, how small and optimized it is! I'm *really* impressed (trust me, it doesn't happen that often ;) ) and I'm glad it's been done for this project of mine. Can't hide it, I'm very happy to have so many talented peoples interested in this! :mrgreen:

Best regards,

Uze

Re: UZEBOX AVCore up and running!

Posted: Fri Sep 19, 2008 7:15 am
by havok1919
uze6666 wrote:At the risk of repeating myself, I just looked (again) at that pic with you holding the module with your fingers...and now I realize, how small and optimized it is! I'm *really* impressed (trust me, it doesn't happen that often ;) ) and I'm glad it's been done for this project of mine. Can't hide it, I'm very happy to have so many talented peoples interested in this! :mrgreen:
Glad you like it. And kudos to you for putting all that time and effort in developing the software base to make it an interesting platform. I've had "AVR video widget" on my todo list for *years* and just could never bring myself to put the kind of time required in on the software to make it general purpose enough to be useful, so suffice it to say I was happy when I stumbled across your project. :-)

I've been struggling for a good size comparison on the AVCore myself... It's a 40-pin DIP, so that's easy enough, but not everyone has one of those on their desk. I finally found something "close enough" that most everyone knows. A good old AA cell!

Image

-Clay