Argh. Video out issue on prototype board, suggestions appreciated

Bill Henning

While I love my demo board (and Hydra) I wanted to do some work with more (native) I/O pins available; so I soldered up a simple Propeller setup on a prototype board - 5MHz crystal, 3 transistor RS-232 interface, 28 I/O brought to female pin headers, and three I2C sockets.

After buzzing out the circuit, correcting two minor mistakes, it worked the first time I powered it up! (powered by four NimH 2300mAh AA cells regulated by a 3.3V regulator, with a 10uf cap between Vcc and Vss at the regulator; brownout detector tied to ground)

It ran a "blink" program just fine; and I could program the eeprom at address 0 just fine (other sockets are still unused)

Then I added the resistor network for generating NTSC video, and the problems began...

- I buzzed out the circuit, it is wired correctly, following the demo boards schematic section
- NONE of the TV code that works on the demo board, and hydra board work on it!
- I moved the pin group, no difference (yes, I changed the Spin code)
- the Prop's pin's are OK, I can drive an LED (thru 270 ohm current limiting resistor)
- the power is good, even under load it stays at 3.33V (I thought it might be browning out)
- I tested the RCA connector, its good
- I checked, tip is the signal, the outside is ground

I'm getting quite frustrated! I want to use P24-P27 for a "debug" TV port, leaving P0-P23 free for other uses.

Anyone have any ideas?

I'm going to try going over the circuit again. I want to get this baby running!

p.s.

One weird thing... I accidentally touched the tip of the RCA cable from the TV to the shield on the connector on my prototype... and the board reset itself.


Post Edited (Bill Henning) : 12/6/2006 8:08:04 AM GMT

AndreL

Sounds like a short somewhere, I suggest building a prop on a solderless breadboard again, on another breadboard or area, and make it real simple. Also, obviously try another prop chip to make sure somethnig isn't fragged.

Andre'

Bill Henning

Thanks AndreL,

I checked for shorts, did not find any.

I changed the RCA jack.

I wired up four LED's with 470ohm current limiting resistors to the same pins instead of the d/a resistor network.. and it works just fine driving the LED's

Still does not drive the TV!

cgracey

Bill,

Do you have a scope that you can check the video pins on? You should see 24..26 toggling.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


Chip Gracey
Parallax, Inc.

Phil Pilgrim (PhiPi)

Your circuit description didn't mention bypass caps. The problem is similar to those of others who forgot this important detail.

-Phil

Bill Henning

Hi Chip,

Yes, I have a scope, and the pins are toggling.

Hi Phil,

Nope, I do have bypass caps, 10uF between Vout and GND at the regulator, and 10uF between +3.3 & GND at Propeller.

What's weird is that both with and without load, I am getting 1.875V / 0.948V / 0.487V at the video output when I set the MSB to LSB one at a time. I'm beginning to wonder if its the TV that's a problem/

Bill Henning

Ok, solved.

I put a 100ohm load on it instead of the monitor, and I got .22v-.89v

That clued me in.

Crappy connector!!!!!

GOT IT WORKING!!!!!

Thanks guys!

Phil Pilgrim (PhiPi)

Bill,

'Glad you got it working!

Now I do hope I'm not running this into the ground, and I sometimes wonder if people get tired of my saying it, but... A 10uF cap, unless it's a ceramic with extremely low ESR, probably doesn't qualify as a bypass cap. More capacitance, by itself, isn't necessarily better here. The objective is to deal with the fast transients; and a 0.1uF+ ceramic with short leads, located as close to the chip as possible, is the best way to do that. Your filter cap (the other 10uF device located near the regulator) can take care of the slower stuff. Also, in a breadboard situation with long power leads, two ceramic bypass caps -- one on either side of the chip -- are even better.

As Bob Pease once said, "Everything is analog." And digital can be some of the roughest analog around! But I know that you know all this and just wanted to get something working as fast as possible and planned to add ceramic bypass caps all along after you had a chance to scrounge a couple from your parts drawer. So I'll quit sermonizing now!

-Phil

Bill Henning

Thanks Phil!

And no, you are not running it it into the ground - I've been out of electronics for years, its actually the propeller getting me back into it (ok, and CPLD's, FPGA's, etc don't hurt either)

I used a 10uF electrolytic because the demo board schematic showed 10uF and I did not have any 10uF ceramics... however based on your advice, I'll remove the one at the processor and replace it with a .1uF ceramic bypass cap on either side of the Propeller.

You are right, I used to know this... but I'm rusty, and the schematic showed 10uF non-polarized; I 'hacked' by substituting 10uF electorlytic.

Any time I post, PLEASE feel free to 'sermonize' - I'm serious about getting back into electronics; I've taken delivery of an HP/Agilent 100MHz logic analyzer, Tek dual channel 100MHz scope, and I've got more gear and parts coming [noparse]:)[/noparse] [noparse][[/noparse]I'd love faster/better scope + logic analyzer, but I have a budget... and I figure these will last me until the next version of the Propeller shows up]

Best,

Bill

parts-man73

Phil Pilgrim said...
and a 0.1uF+ ceramic with short leads, located as close to the chip as possible,

On my SpinStudio, I put one 0.1uF ceramic under the Propeller Socket midway between the 2 sets of power pins. I believe that should be sufficient. There is enough clearance under the socket, I thought that rather than put one on each side, I'd use 1 and put it in a more central location.

Brian