Re: Pixie video from ELF2000
- Thanks David. I've done some experiments with various TVs around the house and web searches to confirm your numbers which as we all knew were correct. So, I did some algebra and came up with a video out circuit and values that works (at least on my PIC 1861 emulator http://www.tedrossin.0sites.net/Electronics/RCA/RCA.html#ElfClone). On the PIC the output high voltage is 5V as I would expect the 1861 to be. The PIC has more drive current(about 20mA instead of about 4ma for the 1861).
If the monitor you are connecting up to has a 75 Ohm termination (which they all do). Then skip the pull down resistor (the 75 ohm terminator in the monitor will do) and just use a resistor from the SYNC pin to the video connector and the VID pin to the video connector:
The value for the NSYNC resistor is 1K and the value for the VID resistor is 429 but just use two 220 Ohm resistors in series to get 440 Ohms which is close enough. So you still only have 3 resistors which are common values.
To solve for the values I created two equations which have two unknows:
R1=Rsync = 75R2(1-Sb)/(Sb(R2+75))
Where Sb is the black scale factor and Sw is the white scale factor.
Sb = 0.3/5 = 0.06 and Sw = 1.0/5 = 0.2
R1=Rsync = 1175R2/(R2+75)
R2=Rvid = 300R1/(R1-300)
Solving and using the quadradic equation gives:
R1=Rsync = 1000
R2=Rvid = 429
The max current draw is 13mA which is more than the 1861 can drive with one pin but there are two pins trying to drive this so it may work ok.
--- In firstname.lastname@example.org, "David G Williams" <davidg.williams@...> wrote:
> Hi Ted,
> May I elaborate on your comment about the Video spec.
> A monochrome TV signal should have a maximum signal level of 1 volt peak to peak when properly terminated by a 75 ohm load. Of the 1 volt peak to peak signal the lowest point is the synchronization tip. Then you have 0.3 volts to get to the Black level. The actual video signal is from the 0.3 volt level to the 1 volt peak - which is pure white. What can confuse some people is that if you use an oscilloscope to look at the video output from a computer & you don't terminate it into 75 ohms then the voltage will be 2 volts peak to peak.
> ----- Original Message -----
> From: ted_rossin
> To: email@example.com
> Sent: Friday, February 19, 2010 3:32 AM
> Subject: [cosmacelf] Re: Pixie video from ELF2000
> When I first built my 1861 circuit it did not sync. I lowered the 10K to 680 Ohms and it worked great. So I use 680, 2K, 1K instead of 10K,2K, 1K.
> All of these values suck and violate the composite video spec which should produce a 75 Ohm impedence and create a black level of 0.5 V and white level of 2V.
> I've been meaning to do the math one day to pick correct values.
> --- In firstname.lastname@example.org, "waltermi@" <waltermi@> wrote:
> > the ELF 2000 uses three resisters to produce the video from the pixie chip outputs. there is a 3.3k from composite sync to the rca out,
> > along with a 1 k from video to the same rca connection, and a 200 ohm from there to gnd...
> > Comp sync -----XXXXXX---------|
> > 3.3 k |
> > |----- video out
> > video out -----xxxxxxx--------|
> > 1K x
> > x RCA JACK
> > 200 x
> > x
> > --------- gnd
> > ( I hope the bove looks fine when posted. )
> > in any case, the video out is very dull and I have a couple of LCD screens that are unstable when connected to my ELF2000.
> > The original ELF pixie cicruit had 10K and 2 K and 1 K
> > in the same order.
> > does any one have their elf connected to an inexpensive lcd?
> > and had similar issues?
> [Non-text portions of this message have been removed]