A different approach to video modifications

Started by viletim, December 16, 2008, 05:20:53 PM

Previous topic - Next topic

viletim

When modifying consoles to output any kind of video (which is not immediately available from the video connector) most of the time require an active video driver circuit. In the past I've always tried to design the video driver with the minimum of number of components which can easily be bought from any electronics shop. I'd use discrete transistors and rely heavily on the support of the circuitry already on the board to ensure the video output was within spec.

I think this approach has largely been a failure. Previous work like my N64 RGB driver and Colecovision composite video mod which work well on my test console but don't work so well when somebody on the other side of the planet tries the same thing on their console. There are so many different revisions of any given console. Not only that, the component tolerance between two identical consoles can vary enough to make a driver circuit which works perfectly on one particular unit work less than optimumly on another unless a lot of care is taken.

Another problem with modifying the circuit is that there are often unwanted side-effects. For example, a lot of composite video modifications leave you with no or poor(er) looking RF video output.

I now think a better approach is to leave the original circuit alone as much as possible and draw the required signals with a high impedance buffer. High speed video opamps which operate from only a signle 5 volta rail were once hard to get and expensive. Now they are readily available in single peice quantities from a lot of places. Some one "size fits all" video buffer circuits could be designed which would be pretty much guarenteed to work on any varient of a given console or perhaps any console at all. Such a circuit could have a gain adjustable with pots and have the output Z fixed at 75 ohms.

Any thoughts?

Link83

#1
I think it would be a fantastic idea  ;D

So what your proposing is a standard Video Amplifier circuit that could be used externally or internally for any given console?

Would the idea be to have a number of standard circuits made (making the cost lower) so that you can easily use it on any console? (If so please count me in for quite a few!) Or were you thinking self build?

Just one question - would you need to use an oscilloscope in order to set the pots gain correctly for each specific console? I only ask as I hate adjusting things by eye, as im forever thinking 'just one more tweak'. Maybe would it be possible to have a space for a fixed resistor instead of a pot, for those people who just want to use it internally for one specific console? Then we could have a small list of the correct resistor value to use for any given console - if that makes sense!

btw was it me you were thinking of in regards to the N64 RGB Driver not working as expected? I only ask as it does work fabulously if I dont connect the amp up to the ENC-NUS chip (So lose RF/Composite/S-Video) but its worth it for the great RGB picture you get. I am just hoping I havent put you off designing more amps in the future - as the N64 RGB driver does work great for its intended purpose of amplifiing the RGB signals to the correct level.

I do agree though that it would be nice to leave the consoles original video circuit intact if possible (I didnt enjoy removing those surface mount resistors!) and I think the idea of just tapping the RGB from the video chip and connecting it to the output withing messing with any other parts is quite desirible (Although I remember you saying that this was possible with normal components but would make the amps component count alot higher)

Your hard work and effort is really appreciated Viletim ;D

undamned

@ viletim:  I agree w/ that approach.  Not that I always adhear to that, but, the theory is spot on.  Two major benefits I see are:
1.) Uniformity across hardware platforms by using the same amp
2.) [By modifying each console as little as possible] you eliminate the need to grossly rework another console if one of your consoles dies.

@ Link83:
Maybe I'm wrong, but I think having pots means adjusting each color's brightness by eye.  Color bar generators can be used to calibrate pretty easily.  One could make a batch of these and calibrate them on something w/ a color bar generator (for example, Capcom's CPS2 games have a nice one), and then leave it to other people's discretion to fine tune.
-ud
"Don't need to ask my name to figure out how cool I am."

RGB32E

#3
Quote from: viletim on December 16, 2008, 05:20:53 PM
I now think a better approach is to leave the original circuit alone as much as possible and draw the required signals with a high impedance buffer. High speed video opamps which operate from only a signle 5 volta rail were once hard to get and expensive. Now they are readily available in single peice quantities from a lot of places. Some one "size fits all" video buffer circuits could be designed which would be pretty much guarenteed to work on any varient of a given console or perhaps any console at all. Such a circuit could have a gain adjustable with pots and have the output Z fixed at 75 ohms.

Any thoughts?
Didn't you post about a month or two ago how you were going to build a circuit around a video amp (TLS1233), but didn't have time?  What ever happened to that project?  ???

albino_vulpix

Quote from: viletim on December 16, 2008, 05:20:53 PM
Another problem with modifying the circuit is that there are often unwanted side-effects. For example, a lot of composite video modifications leave you with no or poor(er) looking RF video output.

So that's why my Atari's RF went bye-byes :p I haven't noticed a difference with the Master System's RF though, hmm.

How much of a difference in component count/cost are we talking? I'd imagine most of the members here wouldn't mind a more complex circuit at all.



duo_r

But what is the issue with harming the RF signal? Who cares about those these days??? I am all for the external RGB amp idea, and have thought this is a great approach -standardize the output and amp externally = less work in the long run.

Quote from: viletim on December 16, 2008, 05:20:53 PM
When modifying consoles to output any kind of video (which is not immediately available from the video connector) most of the time require an active video driver circuit. In the past I've always tried to design the video driver with the minimum of number of components which can easily be bought from any electronics shop. I'd use discrete transistors and rely heavily on the support of the circuitry already on the board to ensure the video output was within spec.

I think this approach has largely been a failure. Previous work like my N64 RGB driver and Colecovision composite video mod which work well on my test console but don't work so well when somebody on the other side of the planet tries the same thing on their console. There are so many different revisions of any given console. Not only that, the component tolerance between two identical consoles can vary enough to make a driver circuit which works perfectly on one particular unit work less than optimumly on another unless a lot of care is taken.

Another problem with modifying the circuit is that there are often unwanted side-effects. For example, a lot of composite video modifications leave you with no or poor(er) looking RF video output.

I now think a better approach is to leave the original circuit alone as much as possible and draw the required signals with a high impedance buffer. High speed video opamps which operate from only a signle 5 volta rail were once hard to get and expensive. Now they are readily available in single peice quantities from a lot of places. Some one "size fits all" video buffer circuits could be designed which would be pretty much guarenteed to work on any varient of a given console or perhaps any console at all. Such a circuit could have a gain adjustable with pots and have the output Z fixed at 75 ohms.

Any thoughts?

kendrick

The ideal mod of any sort will add a capability but not remove an existing capability. As unlikely as it might be, you might want to have the RF output at a later date. If you were able to add a great feature at the cost of the second player controller port, would you do it? The same thinking applies here.

Link83

Quote from: kendrick on December 17, 2008, 11:28:34 AM
The ideal mod of any sort will add a capability but not remove an existing capability. As unlikely as it might be, you might want to have the RF output at a later date. If you were able to add a great feature at the cost of the second player controller port, would you do it? The same thinking applies here.

I have to say I do prefer adding abilities instead or removing/replacing one. Also its great to be able to change between each video output method and see the difference on each system  :)

I have just realised I was talking rubbish before - if you use a pot you could easily set it to any given value using a multimeter, no need for a fixed resistor!

This amp circuit sounds better and better to me  ;D I hope it comes to fruition.

viletim

This would be an internal amp with 3 or 4 channels which runs from the consoles's +5v rail. Wired into the console close to the point of the signal source (which would reduce noise, stray capacitance, etc).

Immediately useful for NTSC N64 RGB, PC Engine RGB, SMS S-video, intellivision composite -video, colecovision component video. Probably a few more with some extra parts added. Also good as a general purpose module which could be used when modifying some unfamilier/obscure piece of hardware for RGB, etc.

I'll keep this one pretty simple. Not much more than the op-amps (three or four on one chip), some pots for gain adjust, and some resistors.

QuoteDidn't you post about a month or two ago how you were going to build a circuit around a video amp (TLS1233), but didn't have time?  What ever happened to that project?
I'll be getting back to that. This idea is a bit like a little baby version of that.

duo_r

I agree with the general idea of what you are saying, but to compare removing RF modulator to a controller port is well....just..... ::). Anyone that removes RF from a system is doing a favor for the future generations....  ;D

Quote from: kendrick on December 17, 2008, 11:28:34 AM
The ideal mod of any sort will add a capability but not remove an existing capability. As unlikely as it might be, you might want to have the RF output at a later date. If you were able to add a great feature at the cost of the second player controller port, would you do it? The same thinking applies here.

Leynos

Hi,

I'd been thinking about modding my Laseractive for RGB, and was planning on using this as an external video amp:

http://www.lenexpo-electronics.com/pdf/SB-2820.pdf

Do you think it would do the job of the device you're describing, or would it be inappropriate?

Cheers.

ken_cinder


viletim

Quote from: Leynos on January 08, 2009, 10:52:08 AM
Hi,

I'd been thinking about modding my Laseractive for RGB, and was planning on using this as an external video amp:

http://www.lenexpo-electronics.com/pdf/SB-2820.pdf

Do you think it would do the job of the device you're describing, or would it be inappropriate?

Cheers.

No, not without modification at least. This device needs a 75 ohm signal to begin with. It'll load the internal circuitry of you Laseractive just like a TV will.


budo-ka

Back in 2006 I discovered the Bakuchikujuu's way to make a PCE amp. Trying to understand why he did it that way I manage to reverse engineering the video output of the PCE.

That gave me that : http://users.teledisnet.be/mbe48577/electronics//pictures/pce_cvbs2.png

Trying to apply this schematic for doing the same amp for RGB will be too much trouble. So I manage to use a simple quad op-amp, the lt6551.

A long time has passed, and finally found some hours during this last Christmas holiday to complete this project.
Here is the result : http://users.teledisnet.be/mbe48577/electronics/PCEngine_RGB_Amplifier.html

As I noted below we can replace one component by a potentiometer to (almost?) achieve your goal.

I like to say  "Electronic is like the arts, there is 1 way to do things wrong but 10,000 others ways to do them right". Sure that we can find one more solution to the eternal problem of RGB amp :)

have fun



viletim

budo-ka,

The LT6551 isn't an opamp at all (there's no feedback) but a fixed gain amplifier. It would work very well in applications where a fixed gain of 2 is desired but what if you need more?

It's a nice little device though, very low bias current (and hence high input resistance). That was what was bothering me about all the high slew rate opamps I've looked at so far, massive bias current.

budo-ka

Quote from: viletim on January 21, 2009, 10:46:34 PM
budo-ka,

The LT6551 isn't an opamp at all (there's no feedback) but a fixed gain amplifier. It would work very well in applications where a fixed gain of 2 is desired but what if you need more?

Correct! This is an 'technical language misuse'.  meaculpa ;D
The LT6551 contains 4 opamp with internal resistors but the ic as a whole is a fixed gain amplifier.

This amplifier board must be considered as a video buffer as the overall voltage gain will always be 1 (x2 then /2). It's only amplifying current coming from video DAC that cannot drive heavy load like a 75Ohm input.
By using an input trimpot we could use it to adapt system like a jamma board (2~4V pp) to the TV input (0.7V pp). Who can do more can do less.
Of course if we have a video signal less than 0.7V we will have some trouble and a dark picture. I don't know if there is some system (video DAC), especially video game hardware, that have that kind of output.

bye