VGA connection doesn't work

qsuscs

Hello,

for some days, I own a Propeller Professional Development Board (Rev. A), which also has a VGA connector. So I tried various demos and also CoggyTerm, but none of them worked.
I wired it like this: P16=V, P17=H, P18=B0, P19=B1, P20=G0, P21=G1, P22=R0, P23=R1. I also checked the resistors, none of them was more than 5Ω off the value it should be.
When I connected the Propeller to my TFT, it just said "no signal" and went to standby, my good pl'' Eizo CRT just showed a "thing" almost shaped like a recangle (It was more like ")_̄("), its color was almost black, but when i disconnected the VGA cable, the screen turned totally black and went to standby
The wiring was not perfect (some loose cables hanging in the air and), but shouldn't I get at least a malformed picture?

qsuscs

English is not my native language, but I hope that you can understand me.

Rayman

I'd guess either the Prop chip isn't working, the wiring is wrong, or your monitors can't accept the output signal...

Some of the demos are at kindof a strange refresh rate that I know some monitors can't handle.
Try the "VGA_512x384_Bitmap_Demo.spin" demo included with the Prop Tool.
This one is 1024x768 at 67 Hz that I think most monitors can handle.
(Kye has a 640x480x60Hz demo on OBEX, I think, that may be even easier to sync to).

If that doesn't work, then I'd double check the wiring.

If it still doesn't work, then I'd try some other types of demos to make sure your board is working right...

Toby Seckshund

H and V signals the wrong way around is common.

I also have a hatred of the +5V on the VGA pin 9, if that gets taken to ground by the entire middle row getting commoned to ground then the Prop would be stopped, and the 5 Volt regulator gets hot. Are you useing a proper VGA cable? or a hand wired arrangement? The 5 Volts on pin 9 is only there for real video cards to get the details back from the monitor's EEPROM, the Prop doesn't use any of that.

Alan

qsuscs

I tried with another Propeller chip, but this didn't work as well (although both chips might be defective), I checked the wiring and even changed it (swapped X0 with X1, I read that the Dev Board Rev. A has the pins arranged the other way round than the other boards), I also treid another screen (CRT), but it just went to standby (standby, not "no signal mode", in this case, the LED of the CRT went yellow instead of orange).
EDIT: The VGA cable doesn't even have a PIN 9, at the other end, there are just 5 BNC connectors (HVRGB)

Toby Seckshund

Well at least it isn't a Pin 9 problem.

The way round a crystal gets connected doesn't matter, presumably it is the right frequency. This is usually 5MHz but many use higher frequencies for overclocking.

At the monitor end the only ones that matter are the RGB and H & V, on the 15 Pin I believe 5, 6, 7, 8, 10 should be ground.http://www.theavguide.co.uk/view_page.php?page=18 gives the general idea.

Are you using a 15 pin -> 5 BNCs cable, if so H and V swapping is very easy



The Props themselves are very robust little creatures (heaven knows that I have been cruel to them and all but once( 9 volts, not 3.3 Volts!) they survived)

qsuscs

Toby Seckshund wrote: »

The Props themselves are very robust little creatures (heaven knows that I have been cruel to them and all but once( 9 volts, not 3.3 Volts!) they survived)

Seems like, while I was trying to get that damn DIL package into the socket, I must have touched the pins with my fingers about hundred times.

Toby Seckshund

Hopefully there are no nylon carpets or Van der Graff generators around you :-)

Most of my Props are 40 pin DIPs and they get used in many different projects. I think you said that you are using a "Bread-Board" type of kit, make sure that the Prop is sat fully down so that the pins make contact.

BOTH sets of power pins must be connected and decoupling capacitors placed as close to those as possible.

The /BOE pin should be grounded (to start with) as this will enable the internal pull up resistor on the /RST pin. A 10K resistor pulling up /RST wouldn't hurt anyway.

Do you have any test kit such as 'scopes etc?

Edit

Something that has just occured to me. When you say that you are wiring to P16 - P23 you are actually on pins 21- 28 of the DIP .

qsuscs

I have a Parallax Propeller Professional Development Board, it has a socket for the Propeller, a socket for the E²PROM and loads of other stuff (all but one IC are soldered directly to the board). However, I have to make the Connections Propeller --- Other Stuff by myself, but I checked the VGA connections some times, and as I already said, I tried swapping X0 and X1 or H and V, didn't work.
I don't have an oscilloscope (although I could do something in software with the Propeller or the audio card of my PC), but I do have a multimeter.

Toby Seckshund

I have just looked at the schematics of that board and it looks as if there should be no problem in getting everything right (I have always made my own setups).

http://obex.parallax.com/objects/68/ is Chip's original VGA with a Demo attached, download, unzip it all, compile and load into RAM (F 10) and it should be working.

The loading and verifications take place whilst the Prop is using its internal RC oscillator, but the VGA will invoke the PLL multiplier, The PLL can become damaged on poorly designed circuits but you have an official type so that shouldn't be a problem.


If you do have a problem with dead ICs etc then get in contact with Parallax, they are very understanding.

Rayman

I would measure the DC voltage at the VGA pins then...
If they are all zero volts, then something is wrong...

Tubular

The "almost black" thing is interesting. Perhaps there is an almost-short somewhere

I would forget the VGA stuff for a moment and just check you get the DC voltages you expect from the resistor networks. Each color has two resistors, a 240 ohm (MSB) and a 470 ohm (LSB). Their action combines with a 75 ohm terminating resistor (inside the monitor) to produce a pin voltage of about 0.00, 0.35v, 0.70v or 1.06v for bit patterns 00b, 01b, 10b, 11b. You can check all this at DC using a multimeter.

You can just set pins P22 and P23 both high, with monitor connected, and check the "red" signal has about 1.06v on it. In fact just set all pins P16..23 to high, and check each of the 3 colors has approximately 1.06v on each of them. The H and V sync signals may have a different voltage on them, eg 0.79v if terminated 75 ohm, or even 3v3 if unterminated, it doesn't matter so much for those signals.

Let us know what you find. I'm guessing something is crossing over or shorting somewhere.

edit: Also, what frequency of crystal are you using, and does your code reflect that same crystal value?

qsuscs

I think the H line is not properly connected. I set the Pins to HI (dira[16..23]~~; outa[16..23]~~) and R, G, B and V were at ~3.3V, but H at 0V. So I checked the resistance between P17 at the Propeller and P13 at the VGA connector, it was around 30MΩ (yep, mega-ohm). Then I checked the connection between the H header on the top and the soldered pin at the bottom, 25MΩ, with some pressure around 10MΩ.
BTW, the crystal is at 5MHz, but every program I saw till now used 5MHz (or the internal oscillator)

I checked it again, the connection is broken somewhere between the resistor and the VGA plug.

Rayman

Sounds like a defective board then... Maybe a bad trace...
If you have a magnifying glass, maybe you could look at the trace and see if it is broken.

Also, I'd check the solder connection on the H pin.

Maybe also check the resistor. Maybe it's not soldered on right or maybe it's the wrong value...