VGA scan conversion alternatives

[pentium]

After quite a bit of work I am now on the final step of finishing my VFX1 video linkbox.

The attached .doc file shows what is needed to convert a regular VGA signal to the required video output that the two LCD screens on the VFX1 need. The problem is the two scan converters at the end. They suddenly triple the size of the linkbox and make things expensive. Is there a scan converter I can build myself that does what I need or is it far more economical to get two prebuilt (like the Trust Televiewer for example) units?

[pentium]

Re: VGA scan conversion alternatives

I'm also having a bit of trouble figuring out how to wire up the DG541's to their respective left and right channels as well as where the Q and Q lines from the 74LS73 go to on each DG541 AND which pin the Csync lines from the scan converters connect to.

[pentium]

Re: VGA scan conversion alternatives

Here, I roughly pasted the chips into place. Show me how they should connect up according to the block diagram in the first post and the datasheet in the second post using paint. I am totally stumped.

[pentium]

Re: VGA scan conversion alternatives

I think I finally figured it out.

[pentium]

Re: VGA scan conversion alternatives

Okay, I have ironed out most of the wrinkles and the last hurdle I have is the scan converters.

As seen in the attachment in the first post of this thread, The scan converters
turn the VGA RGB signals to regular RGB signals as well as provide a Csync output. The thing is I am wondering how you do that but without any horizontal OR vertical sync signals.
I can find plenty of VGA to RGB + Csync converter schematics out there but they all require H and V sync and I really can't modify the current schematic without it affecting the output on the headset. I'll continue looking for the proper schematic but does anyone else have any ideas?

[pentium]

Re: VGA scan conversion alternatives

I'm pulling my hair out now.
I'm going to give up for tonight and hope someone can answer my question.

[davmax]

Re: VGA scan conversion alternatives

I have taken a look at circuits. The C synch signal can be generated from the first synch combiner circuit here: http://www.tkk.fi/Misc/Electronics/c.../vga2rgbs.html

It may be possible to eliminate th bulky and costly converters by applying this circuit. I understand that your original circuit applies alternate RGB line scans to each output complete with C synch signal. I suspect that it may not matter if the C synch signal is applied when the RGB signal is missing on one side or the other. If this is the case take the circuitry and the RGB signals as per your original diagram without the converters, add the synch combiner circuit and use the output to drive both left and right sych lines. If this is too much loading use two synch combiners, one for each side. Hoping this will not load the VGA H & V synch lines too much. If loading is too high a buffer circuit is required.

This is all new to me but I am using logic and some design experience. If the C synch line requires TTL level signals use the second circuit.

It is helpful that the same TTL chip is used in all cases.

Hopefully this is helpful.

[pentium]

Re: VGA scan conversion alternatives

Okay, I read what you said several times and correct me if wrong:

1) The RGB signals do not need to be changed.
2)adding another curcuit to the already used H and V lines will be no problem and if there is, a buffer will be needed.
3)Depending on what I need, I will require either one or two H and V sync to Csync converters.

What I have to say:
a)The Csync line should only be active when an RGB signal on the same channel (left or right) is active.
b)It is not known what will happen if both channels receive a constant Csync signal.

[davmax]

Re: VGA scan conversion alternatives

Okay, I read what you said several times and correct me if wrong:

1) The RGB signals do not need to be changed.
2)adding another curcuit to the already used H and V lines will be no problem and if there is, a buffer will be needed.
3)Depending on what I need, I will require either one or two H and V sync to Csync converters.

What I have to say:
a)The Csync line should only be active when an RGB signal on the same channel (left or right) is active.
b)It is not known what will happen if both channels receive a constant Csync signal.

The description on the link I sent indicated no need to change the RGB signal. So going with that.

Right the C synch sounds interesting. Surely it is a synch pulse ie only momentary. If it is constant as you say, surely the synch must be on all the time on a single VGA monitor and it works. I am speculating at this stage. I only suspect a synch pulse per line (H) and it should not matter if a line is missing when one side of RGB is OFF. What do you know/think?

[pentium]

Re: VGA scan conversion alternatives

I really don't know what would happen is the headset was given a constant Csync signal. I could build the circuit anyways and since we are still sketchy in this part I would leave it on a proto board.

[pentium]

Re: VGA scan conversion alternatives

Going along with what you said I will first attempt with one csync converter, as seen in figure 1 (schematicv3fig1.jpg)
How does it look?

[davmax]

Re: VGA scan conversion alternatives

Hi Pentium. I am printing out diagrams to follow what you have proposed. I work better with schematics. The wiring diagram will take some tracing to find out what component features you are connecting. Will get back.

[pentium]

Re: VGA scan conversion alternatives

Okay then.
I hope you can read my writing.

[davmax]

Re: VGA scan conversion alternatives

One problem I have is that I have no idea what the connections to the VIP card etc do.

Checking your connectivity the two FET switch blocks looks OK except I see no connection to the V+ and V- pins. These must be connected to allow the signal to pass (switches to operate). See page 9 of the data sheet found here: http://www.tranzistoare.ro/datasheets/166/494816_DS.pdf

See also typical connections and supply voltages on page 10.

The supply voltages required will depend on the signal being passed note V+ = 12V and V- = -5V to pass a +-5 volt AC signal. This need does complicate your circuit. It is a matter of knowing the the signal to be passed, the detail with the synch converter circuit states RGB signals as 0.7Vpp.

Your application of C synch to the switches is fine. I notice one error, hope I have not missed another. The RGB lines going into the Fig.1 block at the right are mislabelled ie C & E are interposed, check this.

If this config with common synch does not work then it can be arranged to switch the synch signal to alternate sides. First try as is after fixing the supply voltages.

[pentium]

Re: VGA scan conversion alternatives

Thanks for pointing that out. I was going to check everything anyways before I did a good copy.
I'll go and try to dig up the site where most of my information came from.

http://tech.groups.yahoo.com/group/forte-vr/
If you don't have a yahoo account, get one and join this group.
they have all the references you need right up to the pinouts.

[pentium]

Re: VGA scan conversion alternatives

Okay, I just fixed the mixup of the right green and blue lines so those are now fixed.
For voltages, I did not think about weather or not the FET blocks needed power as according to the document at the top of the thread they did not need it.
The RGB output is still a guess to me as I do not know how the Televiewer 11672 converts a VGA RGB signal into the RGB signal the headset needs. We are currently assuming there is no difference.
I also want to add that the other lines that are not used in the schematic (run to and from the controller and the headset) are just data, and audio lines as well as another 5V line HOWEVER that other 5V line (or the one that I used in the schematic) might have been mislabeled in the pinout and could possibly be a -5V line. I can confirm this once I get my hands on the controller card later this week.

[davmax]

Re: VGA scan conversion alternatives

Yes the switch blocks definitely need supply voltages as shown in data sheet. A real trap.

[pentium]

Re: VGA scan conversion alternatives

Well then I guess we either have to:

a)Make a circuit that gives us -5V from our existing +5V source.
b)Hope that the pinout is wrong and there is a -5V line.
c)Substitute the power source (EG, it no longer comes from the card and now comes from an external power supply).

Are you looking through that link? That's all the the known information for the headset now. There is nothing else out there.

[davmax]

Re: VGA scan conversion alternatives

If -5v is not available from video source it might be best to use the +12V and the -5V (white wire) on the computer (if it exists) or -12V.

[pentium]

Re: VGA scan conversion alternatives

Well we could always just do that since the average ATX power supply has a -5V line as well as a regular 5V line.
The problem then would be I would have to find a place on the back of my PC to allow the linkbox to hook up and receive power.
If that is the case, I can redirect the +5V and -5V lines to a second connector on the linkbox (say a DB9 socket) and using a regular serial cable, I just plug into the back of my computer (to another custom installed DB9 socket that is wired up) and there we go.

Oh yeah, and mark the port so we don't plug a serial device into it and release the magic smoke.