## 2-bit video DAC questions

5 posts / 0 new
Author
Message

Hi all,

I read the document http://instruct1.cit.cornell.edu... on how generate TV images (on NTSC TVs) using a 2-bit DAC controlled by the ATMEGA644. Please see the attached image of the schematic of the 2-bit DAC.

Basically, it's a resistive voltage divder to calculate out the voltage for white and black levels, depending on the output voltages from pinD.5 and pinD.6

We operate at 5V. I know not exactly 5V output from pins, more like 4.75V something, but, let's just use 5V.

Here is my calculations. Are they correct? And what would be Vout for item (4), when pinD.5 and pinD.6 both output 5V? Thanks!

1. pinD.5 = 0V and pinD.6 = 0V

Vout = 0

2. pinD.5 = 5V and pinD.6 = 0V

Vout = (75ohm/(300ohm + 75ohm)) x 5V = 1.0V WHITE LEVEL

3. pinD.5 = 0V and pinD.6 = 5V

Vout = (75ohm/(1000ohm + 75ohm)) x 5V = 0.35V BLACK LEVEL

4. pinD.5 = 5V and pinD.6 = 5V

Vout = ???

## Attachment(s): b_w_video_dac.jpg
Last Edited: Sun. Feb 20, 2011 - 06:33 AM

Item 4.

PinD5 nd D6 at 5V

work out paralel equivalent of 1K and 300 ohm resistor and then work out voltage divider outcome as before.

EDIT:

your calculations when one of the inputs is differentfrom theother are not correct. Rethink the circuits in these instances.

ignoramus wrote:
Item 4.

PinD5 nd D6 at 5V

work out paralel equivalent of 1K and 300 ohm resistor and then work out voltage divider outcome as before.

EDIT:

your calculations when one of the inputs is differentfrom theother are not correct. Rethink the circuits in these instances.

http://www.rickard.gunee.com/pro...

Last Edited: Mon. Feb 21, 2011 - 12:14 AM

Also assume that there's a further 75 ohm impedance in the TV input termination.

I calculate four states, assuming 5v outputs:

```sync video R from 5v    R to ground      Ratio   Vout  comment
0    0     open cct     75//75//330//1k  n/a       0v  line and field sync
0    1     330          75//75//1k       36/330  0.55v never used
1    0     1000         75//75//330      34/1000 0.17v black level
1    1     1000//330    75//75           38/248  0.75v white level
```

A little bit low for standard levels, or in the absence of the terminating resistor a little high, but they should work on most TVs.

In the absence of the terminating resistor, the voltages for black level and white level evaluate to 0.31 and 1.51 volts - black level is fine but the white level is well over, though it's probably within the limits of the input driver.

If you have no terminating resistor (i.e. the total load to ground is simply 75 ohms) then values of 1100 and 560 ohms for sync and video resistors respectively work to a percent or so of the expected values. If the signal is terminated, so the load to ground is 37.5 ohms, then values of 560 and 270 ohms will be very close (I've used standard values).

Neil

Edit - note also that I have ignored the source impedance of the driver port and assumed that a solid 5v is available. As you point out, it probably isn't quite that, so the output will be commensurately lower.

barnacle wrote:
Also assume that there's a further 75 ohm impedance in the TV input termination.

I calculate four states, assuming 5v outputs:

```sync video R from 5v    R to ground      Ratio   Vout  comment
0    0     open cct     75//75//330//1k  n/a       0v  line and field sync
0    1     330          75//75//1k       36/330  0.55v never used
1    0     1000         75//75//330      34/1000 0.17v black level
1    1     1000//330    75//75           38/248  0.75v white level
```

A little bit low for standard levels, or in the absence of the terminating resistor a little high, but they should work on most TVs.

In the absence of the terminating resistor, the voltages for black level and white level evaluate to 0.31 and 1.51 volts - black level is fine but the white level is well over, though it's probably within the limits of the input driver.

If you have no terminating resistor (i.e. the total load to ground is simply 75 ohms) then values of 1100 and 560 ohms for sync and video resistors respectively work to a percent or so of the expected values. If the signal is terminated, so the load to ground is 37.5 ohms, then values of 560 and 270 ohms will be very close (I've used standard values).

Neil

Edit - note also that I have ignored the source impedance of the driver port and assumed that a solid 5v is available. As you point out, it probably isn't quite that, so the output will be commensurately lower.

Thanks!