RGB Driver for NTSC Nintendo 64

Started by viletim, September 21, 2008, 03:34:14 PM

Previous topic - Next topic

viletim

Here's a good quality RGB driver for those early model N64s which have RGB available inside. I put it in the wiki too. http://gamesx.com/wiki/doku.php?id=av:n64rgb-ntsc

Thanks to Link83 for lending me the board. I'll post it back when I'm done with Mario Tennis.

Link83

#1
Thanks so much for this Viletim, I cant wait to try it  ;D

So if I wanted to make this amp on veroboard would this be the right sort of layout for each line?:-

Or is there a better/smaller layout I should use?

Also, im just curious to know to know if it would be possible to make a similar amp which wouldnt require removing any components/lifting chip legs, or would this compromise the RGB quality?

Lastly, if you have time I would love to know what the 'linking two points' boost method actually does to the RGB signal? Its shown here under 'another solution':-
http://www.eurasia.nu/wiki/index.php?pagename=N64RgbBooster
You can do the same thing by linking pins 22 & 23 of the VDC-NUS chip.

For me it made the picture much brighter but also added interference to the picture  :(

Thanks again  :)

<EDIT - Slightly revised pic)

viletim

Quote from: Link83 on September 22, 2008, 06:02:51 AM
Thanks so much for this Viletim, I cant wait to try it  ;D

So if I wanted to make this amp on veroboard would this be the right sort of layout for each line?:-

Or is there a better/smaller layout I should use?
I hate veroboard. The one pad per hole stuff is so much more flexable. See here for a layout (amplifier circuit is very similar).
Quote
Also, im just curious to know to know if it would be possible to make a similar amp which wouldnt require removing any components/lifting chip legs, or would this compromise the RGB quality?
It's possible but would require an amp with voltage gain. It would be fairly complex with just transistors, a high speed opamp or similar would reduce the parts count but the're not always easy to obtain.
Quote
Lastly, if you have time I would love to know what the 'linking two points' boost method actually does to the RGB signal? Its shown here under 'another solution':-
http://www.eurasia.nu/wiki/index.php?pagename=N64RgbBooster
You can do the same thing by linking pins 22 & 23 of the VDC-NUS chip.

For me it made the picture much brighter but also added interference to the picture  :(
One end is ground, the other is some kind of DC voltage generated by the IC for its internal workings. It seems to be brought out to a pin just to add a small filter cap. I'd rather not fiddle with it, might damage something...

Link83

#3
Thanks  :)

Quote from: viletim on September 23, 2008, 01:29:26 AM
I hate veroboard. The one pad per hole stuff is so much more flexable. See here for a layout (amplifier circuit is very similar).

Its does look like a much better idea than veroboard, but I cant seem to find it available anywhere  :( Do you know what its called/brand name?

Also, Just looking at the amp schematic some more and I am wondering if removing the capacitors C124, C125 and C126 from the board is actually necessary? Since they are all 1uf couldnt I use them to replace the C1 1uf capacitor used in the amp? (eg. remove the 110ohm resistors from the board, cut the RGB traces or lift VDC-NUS chip legs, then connect the ENC-NUS connection from the amp (from between R1 and R2) to the bottom of the C124, C125 and C126 capacitors?

Also, I take it if I am not interested in keeping the Composite/S-Video output that I could just removed C1 entirely, and combine R1 and R2 into one resistor? Would 660ohms be the correct value to use?

Lastly, could I just ask if you recommend using Carbon film or Metal film resistors for amps/video work? (In the Game Gear RGB amp it looks like you used a mixture of both?)

Thanks again Viletim, I really appreciate the effort/work you have put into this - I kind of expected a few more replies!

Hope you enjoy Mario Tennis  ;D

(P.S. If you have a moment id appreciate your insight into the PAL N64 S-Video missing components I mentioned in this thread:-
http://nfggames.com/forum2/index.php?topic=3203.40
Expecially now that I know C14 should be 68nf - I just need to know if the component ordering is important, and if I can add the parts into the cable rather than the console)

viletim

Quote from: Link83 on September 23, 2008, 10:41:55 PM
Thanks  :)

Quote from: viletim on September 23, 2008, 01:29:26 AM
I hate veroboard. The one pad per hole stuff is so much more flexable. See here for a layout (amplifier circuit is very similar).

Its does look like a much better idea than veroboard, but I cant seem to find it available anywhere  :( Do you know what its called/brand name?

It doesn't really have a name, usually it's listed as prototype board, experementers board or something like that. Usualy unbranded or branded with mysterious chinese company names. Verroboard is much more common which is shame cause it's not nearly as good as this stuff.

Quote
Also, Just looking at the amp schematic some more and I am wondering if removing the capacitors C124, C125 and C126 from the board is actually necessary? Since they are all 1uf couldnt I use them to replace the C1 1uf capacitor used in the amp? (eg. remove the 75ohm resistors from the board, cut the RGB traces or lift VDC-NUS chip legs, then connect the ENC-NUS connection from the amp (from between R1 and R2) to the bottom of the C124, C125 and C126 capacitors?

I just transfered the caps from the n64 board to my amp board. They're not hard to remove. Add a bit of extra solder, apply a hot iron and solder wick across the side and they just slide off the PCB. I wouldn't cut a track unless it was absolutely necessary.

Quote
Also, I take it if I am not interested in keeping the Composite/S-Video output that I could just removed C1 entirely, and combine R1 and R2 into one resistor? Would 660ohms be the correct value to use?

Sure. The prefered value would be 680 ohms.

Quote
Lastly, could I just ask if you recommend using Carbon film or Metal film resistors for amps/video work? (In the Game Gear RGB amp it looks like you used a mixture of both?)

I wrote on the diagram "all resistors 1% tolerance" so that pretty much means metal film resistors. If you use the 5% type for matched amplifiers then you might end up with one amp outputing slightly a bigger signal than the others.

I don't like metal film resistors much though, the ones I buy locally have blue bodies which makes the values hard to read. I use normal carbon ones whenever I can get away with it.

Quote
(P.S. If you have a moment id appreciate your insight into the PAL N64 S-Video missing components I mentioned in this thread:-
http://nfggames.com/forum2/index.php?topic=3203.40
Expecially now that I know C14 should be 68nf - I just need to know if the component ordering is important, and if I can add the parts into the cable rather than the console)

The component order isn't important at all, and neither is the placement.

ken_cinder

This sort of stuff?

http://www.radioshack.com/product/index.jsp?productId=2104052

What's the difference between the board you're using viletim and veroboard (aka stripboard), they look the same to me?

viletim

Quote from: IJTF_Cinder on September 24, 2008, 09:17:17 AM
This sort of stuff?

http://www.radioshack.com/product/index.jsp?productId=2104052

What's the difference between the board you're using viletim and veroboard (aka stripboard), they look the same to me?

Veroboard:


Stuff I use:

RARusk

Excellent work. But I noticed the U5 chip in the diagram. If this chip can be safely removed from a dead N64 motherboard how useful could it be for making Composite Video and S-Video? This may be useful for consoles like the Sega Master System, TG-16, and the SNES Jr.
Console hacking is like sex. For best results you got to know where to poke.....

Link83

#8
Quote from: RARusk on September 24, 2008, 01:11:09 PM
Excellent work. But I noticed the U5 chip in the diagram. If this chip can be safely removed from a dead N64 motherboard how useful could it be for making Composite Video and S-Video? This may be useful for consoles like the Sega Master System, TG-16, and the SNES Jr.
I dont know if the ENC-NUS would be suitable or not, but if it is then I thought I should add that it is still available from some Hong Kong suppliers - so theres no need to butcher an N64 (unless its already dead)

Also, if the ENC-NUS would work, what would probably be even better is to use the 'S-RGB A' chip used in late model SNES consoles, as it not only produces S-Video and Composite, but also retains RGB aswell. Even better, if you google the chips code 'BA6596F' you should get quite alot of hits for Hong Kong suppliers who sell them for $2-3 each. I guess they were overproduced!

viletim

Quote from: RARusk on September 24, 2008, 01:11:09 PM
Excellent work. But I noticed the U5 chip in the diagram. If this chip can be safely removed from a dead N64 motherboard how useful could it be for making Composite Video and S-Video? This may be useful for consoles like the Sega Master System, TG-16, and the SNES Jr.

Sure, just make sure of a few thing things:


  • Input is 0.7Vpp (if replacing a CXA chip the signal has to be attenuated by 1/3)
  • Subcarrier must be generated by the console, not and extra oscillator or you'll get heaps of dot crawl (shouldn't be a problem)
  • You need to find a Back Porch pulse from somewhere. Maybe H-Sync would work... if all else fails you can generate one with H-sync and a LM1881

QuoteAlso, if the ENC-NUS would work, what would probably be even better is to use the 'S-RGB A' chip used in late model SNES consoles, as it not only produces S-Video and Composite, but also retains RGB aswell. Even better, if you google the chips code 'BA6596F' you should get quite alot of hits for Hong Kong suppliers who sell them for $2-3 each. I guess they were overproduced!

These are custom chips made for one particular device that have been out of production for 13 years or so. I think $2-3 a piece is being a bit optomistic... That's less than the cost of an AD724 !

Link83

#10
Quote from: viletim on September 24, 2008, 02:42:58 PM
These are custom chips made for one particular device that have been out of production for 13 years or so. I think $2-3 a piece is being a bit optomistic... That's less than the cost of an AD724 !
Actually thats the price I was quoted for the S-RGB A chip when I was looking for quotes a few month ago. I looked into it originally because an early French N64 is the one PAL N64 model which has a VDC-NUS chip, followed by an S-RGB A chip instead of an ENC-NUS as in the NTSC systems. You can see it in this thread:-
http://nfggames.com/forum2/index.php?topic=3083.0

I was curious if I could experiment with adding it to an NTSC N64 with the VDC NUS chip. Its interesting to note aswell that the 'S-RGB A' chip was also used in the Ultra 64 development boards aswell.

They are listed as 'BA6596F' but I asked and they do have 'S-RGB A' printed on the chips. The reason I didnt order any is because shipping to the UK is apparently $30-40 and I only wanted to buy 10, but if anybody else is interested in some we could do some sort of group order to help split the cost of shipping (It wouldnt cost that much to re-ship them on to each person) Maybe ordering 50 for $140?

They offered it that cheap as they had no datasheets or information about the chip they could give me (I knew they wouldnt, but I thought id ask!) so no companies are interested in buying them. Apparently they have thousands in stock!

l_oliveira

A bit unrelated post, but not totally off topic...
I like the "perfboards" too (The perfboard type with individual copper islands for each hole), like viletim. :)

About CXA series of chips, from my personal experience with them, I found out they will sync with any voltage level from 0.7v to 2v but varying the voltage has impact on the "image position" on the screen and also affects the quality of the video signal.
I got best results with the Yamaha VDP (MSX computers) when I set sync voltage somewhere between 1.5v and 2v. I think the Mega Drive also uses it with 2v. 
Anything more than that will cause the chip to lose sync and then the screen will be black.

<off topic> (or maybe not)
some pics of my adventures with the video encoder while trying to make a MSX computer into a MSX2+  The old VDP had no RGB output so I had to modify the machine video output circuit completely...
That's why I had to make a RGB encoder with a CXAchip, for it. And fixing the proper sync level was a *big* headache...

link: http://picasaweb.google.com/leonard.s.oliveira/DropBox?authkey=Iouw4fpGQMs#


viletim

Link83,

It's probably not worthwile, anyone who wants one can just rip it out of a surplus SNES and I've never heard of anybody doing that.

BTW, when buying from chinese suppliers, you will probably have to pay by bank transfer which has high fees (at both ends!). They will only send parts by courier because if they used the postal system your warrenty might expire before you receive the parts, if you receive the parts.


RARusk

I also noticed that the U4 chip uses 3.3V. Would it make any difference if it was replaced with 5V or would Bad Things happen?

In addition, did you measure the signal strength before you did anything? If you did, what did it measure?
Console hacking is like sex. For best results you got to know where to poke.....

Link83

#14
Quote from: viletim on September 25, 2008, 01:00:19 PM
It's probably not worthwile, anyone who wants one can just rip it out of a surplus SNES and I've never heard of anybody doing that.
The only problem is I hate destroying a perfectly working/repairable system just for one part, and I have yet to come across a broken SNES that couldnt be easily fixed.

Quote from: viletim on September 25, 2008, 01:00:19 PM
BTW, when buying from chinese suppliers, you will probably have to pay by bank transfer which has high fees (at both ends!). They will only send parts by courier because if they used the postal system your warrenty might expire before you receive the parts, if you receive the parts.
Thanks for the info  :)  I have never bought anything from Chinese suppliers before so didnt know all that. Probably for the best that it never went any further than a quote and a few questions then.

Quote from: RARusk on September 25, 2008, 01:36:32 PM
In addition, did you measure the signal strength before you did anything? If you did, what did it measure?
Id be interested in knowing the signal strength measurements too if you still have them? Also if you have the measurements from 'before and after' removing the board resistors and capacitors that would be great to know :) (I read the percentages in the Wiki - 35% and 60% -  but am just curious as to what the actual values measured  ;))

viletim

Quote from: RARusk on September 25, 2008, 01:36:32 PM
I also noticed that the U4 chip uses 3.3V. Would it make any difference if it was replaced with 5V or would Bad Things happen?
Yup, bad things.
Quote
In addition, did you measure the signal strength before you did anything? If you did, what did it measure?
Strength? Not sure what you mean... The peak to peak voltage (that is the highest voltage that the video signal will acheive (white) minus the lowest voltage (black)) is 0.7Vpp in its original state.

Quote
Id be interested in knowing the signal strength measurements too if you still have them? Also if you have the measurements from 'before and after' removing the board resistors and capacitors that would be great to know  (I read the percentages in the Wiki - 35% and 60% -  but am just curious as to what the actual values measured  )

The capacitors are only for coupling (DC levels for U4 output and U5 input are different).

The three measurements I made are:

  • IC loaded with 110 ohms (original), 0.7Vpp
  • IC loaded with 45 ohms (original with TV connected directly), 0.7*0.35= 0.25Vpp
  • IC loaded with 75 ohms (110 ohm resistors removed, TV connected directly), 0.7*0.60= 0.42Vpp
With the resistor removed and no TV connected, the signal is completly distorted. This IC needs a load resistor of less than 500 ohms to work properly.

RARusk

Okay, so the voltage of the signal is correct, right? But if that is the case why is the video signal weak?

*scratches head*
Console hacking is like sex. For best results you got to know where to poke.....

viletim

Quote from: RARusk on September 26, 2008, 01:27:27 PM
Okay, so the voltage of the signal is correct, right? But if that is the case why is the video signal weak?

*scratches head*

Connecting a TV/Monitor modifies the circuit, it's no longer original.

QuoteIC loaded with 45 ohms (original with TV connected directly), 0.7*0.35= 0.25Vpp

Link83

#18
So can I just check i have this right - even though an RGB signal needs to be 0.7vpp, it actually has to be 1.4vpp before entering the TV. Is it only 0.7vpp inside the TV then?

May I ask how you learn all of this stuff Viletim? I have had a look online but there doesnt seem to be much information out there on vpp and video signal levels. I wish there was some sort of 'beginners guide to video signals' or similar.

One other question, how does Waltarzars old N64 amp compare?
http://www.eurasia.nu/wiki/index.php?pagename=N64RgbBooster
It is directly connected to the VDC-NUS chip with no components removed - obviously this makes your amp more accurate (No offence intended to Waltarzar) but im just curious. Knowing what parts he used would it be possible for you to say what Waltarzars amp would do to the RGB signal levels?

I cant wait for the transistors I ordered to turn up so I can try this amp  ;D

viletim

#19
Quote from: Link83 on September 26, 2008, 09:29:45 PM
So can I just check i have this right - even though an RGB signal needs to be 0.7vpp, it actually has to be 1.4vpp before entering the TV. Is it only 0.7vpp inside the TV then?

The video signal should be 0.7Vpp at the point in enters the TV.

Idealy, the source is 75 ohms, the load is 75 ohms, the voltage across the load 0.7Vpp. If you remove the load, the voltage will rise to 1.4Vpp. Kirchoff's law.

In the real world, the source isn't always 75 ohms (like the video amp you mention below) and the load isn't always 75 ohms (a few crap TVs use a higher value termination resistor to get a higher voltage signal). The only reason that this works is because the vast majority of devices do stick to the standards so the resault of doing something differently is predictable.

Quote
May I ask how you learn all of this stuff Viletim? I have had a look online but there doesnt seem to be much information out there on vpp and video signal levels. I wish there was some sort of 'beginners guide to video signals' or similar.

I'd like a explain this further, but I'm short on time at the moment. To adequately explain everything will take some preperation... and pictures!
Quote
One other question, how does Waltarzars old N64 amp compare?
http://www.eurasia.nu/wiki/index.php?pagename=N64RgbBooster
It is directly connected to the VDC-NUS chip with no components removed - obviously this makes your amp more accurate (No offence intended to Waltarzar) but im just curious  Knowing what parts he used would it be possible for you to say what Waltarzars amp would do to the RGB signal levels?

Voltage gain of this amplifier is something like 0.95. The Vpp would be unchanged, there would be an additional DC offset added. The exact value of this offset would depend on both the average brightness of the picture, and gain of the individual transistor used. Makes it very difficult to predict.

Basically, the idea is to reduce the impedance of the circuit to as low as possible then stick it into a TV. Then the 75 ohm load of the TV harly makes any difference to the amplitude.

This circuit relies on tolerances which vary too much...it's a poor design.

Link83

#20
Wow, I have so much to learn  :o

Quote from: viletim on September 27, 2008, 12:22:06 AM
I'd like a explain this further, but I'm short on time at the moment. To adequately explain everything will take some preperation... and pictures!
That sounds great  ;D
With all your knowledge maybe you should consider writing a book on video signals?

Quote from: viletim on September 27, 2008, 12:22:06 AM
Voltage gain of this amplifier is something like 0.95. The Vpp would be unchanged, there would be an additional DC offset added. The exact value of this offset would depend on both the average brightness of the picture, and gain of the individual transistor used. Makes it very difficult to predict.

Basically, the idea is to reduce the impedance of the circuit to as low as possible then stick it into a TV. Then the 75 ohm load of the TV harly makes any difference to the amplitude.

This circuit relies on tolerances which vary too much...it's a poor design.

If it would help at all here is the datasheet for the BF494 used in Waltarzars N64 amp:-
http://www.datasheetcatalog.org/datasheet/MicroElectronics/mXqwswz.pdf
It shows the transistors gain, although that might not be much use without knowing the brightness of the picture I guess?

On another note all my 1% tolerence metal film resistors turned up today, so I think I will make a start on the board so ill be ready when the transistors turn up - will post pics when im done  :)

viletim

Quote from: Link83 on September 28, 2008, 12:14:50 AM
Wow, I have so much to learn  :o

Quote from: viletim on September 27, 2008, 12:22:06 AM
I'd like a explain this further, but I'm short on time at the moment. To adequately explain everything will take some preperation... and pictures!
That sounds great  ;D
With all your knowledge maybe you should consider writing a book on video signals?

As far as the electrical characteristics go, that's about all there is to it. Things like source impedance, input imedance, etc apply to other signals too, such as audio. It doesn't matter so much in this case and it's even expected that the source impedance and signal amplitude will vary.

For an RGB video system with near non-existant standards then just look at arcade boards/monitors. There's a reason for R/G/B signal strength pots on any widget designed to work with arcade PCBs.

Quote
If it would help at all here is the datasheet for the BF494 used in Waltarzars N64 amp:-
http://www.datasheetcatalog.org/datasheet/MicroElectronics/mXqwswz.pdf
It shows the transistors gain, although that might not be much use without knowing the brightness of the picture I guess?

Here's why the design is crap:

The transistor is biased into its linear region by a voltage divider between the 1.5k resistor to Vcc and the resistance seen looking into the base of the transistor, which is 75 * Hfe (transistor current gain). According to the datasheet, this may be anything from 67 to 220 for any given BF494 transistor. And this only is when the collector current is 1mA and the temperature is 25 deg C.

Right, so what's the collector current? It's the voltage at the base minus 0.7v then divided by the TV's 75 ohm termination resistor.

So what's the voltage at the base? Well, that depends Hfe.

Then what's the Hfe??? That depends on the collector current! As well as the transistor and the temperature.

This transistor has a maximum collector current of just 30mA (why was it chosen?, a BC548 is good till 100mA). Even if the base voltage was carefully adjusted by means of selecting the perfect resistor value in place of the 1.5k R then the collector current would still be about 20mA. That's pretty close to the edge!

I'd go as far as to say that it's impossible for this amp to work without either distorting the signal (Hfe too low) or destroying itself (Hfe too high) 90% of the time.

And if you have the misfortune to short the output to ground then... you'll have all the current that the +5v regulator can muster flowing through the collector. It'll be a lot more than 30mA.


Link83

#22
Thank you so much for all the info in your previous post Viletim - I cant honestly say I understood everything  :D but the parts I did understand were really interesting  :)

I had been holding off replying until I got the BC548 and BC558 transistors so I could attempt this amp myself - and they have just arrived today. I decided to follow the veroboard design I made previosly, as I couldnt find anywhere locally that had even heard of one hole 'prototype board' - although I did manage to find some on eBay from Hong Kong, but I didint want to wait that long.

Anyway, I put the amp together and I just cant seem to get it to work  :( When I turn the N64 on I get a black picture with sound, and on the TV it says its 'RGB', but theres nothing visable on screen. I reconnected the VDC-NUS RGB wires to the output just to check I hadnt fried it and it is still working fine, so it must be my amp.

I dont know what I have done wrong though - I have checked and double checked all the connections on the amp and still cant see my error. There is no accidental bridges (Scored all breaks with a knife and checked with a multimeter) So I can only assume either my veroboard design is wrong to begin with, or I have put the transistors in the wrong way round, or something else?

I have marked up on the original circuit diagram the collector, base and emitter of the BC548 and BC558 transistors - can I just check I have got them in the right order?

If so, shouldnt this (slightly revised) veroboard layout be right?:-


Thanks for any advice/help you can give me, i'll post some pictures of my amp later.

viletim

Link83,

Veroboard is designed to have the tracks cut by drilling out the holes with a drill bit bigger than the track width. I've even seen a small drill bit with a handle made specialy for this purpose. You don't need to drill right through, just enough to get rid of the copper. Much easier than using a knife.

But asside from that your layout looks fine, I can't think what could be wrong...

What is the DC voltage on each transistor leg with everything connected and switched on but no game in the system?

Link83

#24
Thanks for your help Viletim  :) I actually have a Veroboard Track Cutter, but I prefer using a knife as with a track cutter you 'lose' an extra row of holes making the overall circuit bigger - another disadvantage of veroboard.

I left this amp for a few days as I have been so busy with other things, but just came back to it this morning and its funny how with a fresh pair of eyes it became immediately obvious what I had done wrong - and I feel like such a fool. The amp itself was perfect - in my haste to test it out I had connected the +5v and Ground wires the wrong way round!  :o a few minutes later and the amp was working perfectly despite my previous error  :)...im sorry for wasting your time with my previous reply - I really do feel abit of an idiot for not noticing what i'd done ::)

The picture quality is just fantastic, I was already quite happy with 'legs lifted' method, but with this amp the picture is just so much more vibrant - that 40% difference really does show. I tried taking pictures but its hard to capture the difference on photos.

Originally when performing this mod I just lifted the legs on the VDC-NUS chip off the board and then connected the RGB through the amp to the MultiAV out socket. I didnt even bother to connect up the ENC-NUS chip - so I did lose the RF, Composite and S-Video output. With this the RGB picture is just perfect  ;D

I would imagine with this method the ENC-NUS chip would still be generating a C-Sync signal but without Composite video? My TV was fine with this, but im curious to know if all Scart TV's would be happy with just a C-Sync signal on the Composite line - and not being able to strip it first?

For anybody interested here a quick pic of the way I decided to build my amp, and one of it on top of PAL Mario 64 for size comparison:-

I decided to make my amp as a long strip so I could keep the +5v and Ground connections all on the same line. I could have just built the amp as a square shape and linked up all the +5v and Ground lines but the long strip method seemed the most logical to me and it fits easily in the centre of the N64's heatsink, or down the right hand side of the casing (I know my 'Red output' wire uses a yellow wire - its only because I ran out of red wire to use)

-----------------------------------------------------------------------------------------------------------------------------
Please disregard the following paragraphs about the screen jumping - in the end I discovered the problem was caused by my TV when another SCART device was connected, the RGB amp was not at fault. I only leave the below text so the rest of the thread still makes sense.
-----------------------------------------------------------------------------------------------------------------------------
Then I thought I would connect up the ENC-NUS chip to restore the RF, Composite and S-Video output - and this is when I started to have a slight problem.

I removed the surface mounted 110ohm resistors off the backside of the motherboard, and then connected wires from the R8, R9 and R10 vias to the ENC-NUS output points on the amp, leaving the 1uf capacitors on the board as I had purposefully left them off my amp. Once I did this I again had RF, Composite and S-Video output as expected - the RGB also still worked - so far so good.

However there was one side effect - when displaying in RGB every so often when there is a bright intense picture displayed the screen 'jumps' - its like the sync signal has been disrupted for a split second. The place where I was able to see this everytime was on Zelda Ocarina of Time when you leave it to run without pressing start you see a short segment with Kotake and Koume (The two witches) throwing bright 'magic balls' at Link. I recorded it on my camera so you can see what I mean. It isnt the best quality video as my cameras not very good - so please dont judge the RGB quality by this video. You can see at about 16-17 seconds in just as the second 'magic ball' is thrown the screen jumps up and there is blackness on about 25% off the screen momentarily:-
Video Deleted - problem resolved as mentioned above

Now although this is a rare occurance I figured id rather have perfect RGB with no other video output methods, than RGB with a minor niggle with the other video outputs intact. So what I decided to do was remove the ENC-NUS wires from the amp. I did this but now with the 110ohm resistors still missing from the board the picture 'jitters' up and down very quickly all the time, its like the sync signal is messed up  :(

Any idea why this would be happening? I wouldnt have thought the 110ohm resistors still missing from the board would have caused any problems if theres no RGB signals connected up to the ENC-NUS?

viletim

Link83,

Good to hear you got it working.

Quote from: Link83 on October 11, 2008, 01:07:03 PM


I would imagine with this method the ENC-NUS chip would still be generating a C-Sync signal but without Composite video? My TV was fine with this, but im curious to know if all Scart TV's would be happy with just a C-Sync signal on the Composite line - and not being able to strip it first?

It's no problem at all. The Dreamcast does exactly this when it detects an RGB cable is plugged in.



To be clear, you've lifted the R,G,B pins of the VDC-NUS and connected them to the amp?

I don't know why you'd be loosing sync on bright images... there's no problem with the composite video output on the console I'm working on. The only thing I can think of is that you might be using a PAL N64 RGB cable with the extra 75 ohm resistor on the composite line.

It appears that the ENC-NUS doesn't like to have a high (in this case infinate) impedance source connected to its input (the CXA1645's datasheet warns about this too). What you normaly do with this type of input when you're not using it is connect a capacitor to ground. So just put some shorting links where the 110 ohm resistors used to be.

Link83

#26
Quote from: viletim on October 12, 2008, 11:58:35 AM
Good to hear you got it working.
Thanks  :) Im just sorry I made such a sillly mistake (although im only human!)

Quote from: viletim on October 12, 2008, 11:58:35 AM
It's no problem at all. The Dreamcast does exactly this when it detects an RGB cable is plugged in.
Really? How interesting. I wonder if that was this reason behind the issue this forum member was experiencing?:-
http://nfggames.com/forum2/index.php?topic=3047.0
Slightly OT but does the Dreamcast still generate C-Sync when in VGA mode?

Quote from: viletim on October 12, 2008, 11:58:35 AM
To be clear, you've lifted the R,G,B pins of the VDC-NUS and connected them to the amp?
Thats correct, heres a pic of all the connections there currently are:-

I used the +5v and Ground from the MultiAV output - I assume thats ok?

Quote from: viletim on October 12, 2008, 11:58:35 AM
I don't know why you'd be loosing sync on bright images... there's no problem with the composite video output on the console I'm working on. The only thing I can think of is that you might be using a PAL N64 RGB cable with the extra 75 ohm resistor on the composite line.
Sorry if I wasnt very clear - its only the RGB ouput that it seems to be having problems with the sync on bright images (Im assuming its the sync) the Composite output appears fine  ??? Its very strange and only happens for a moment.

I am using an Official Gamecube Scart cable on which I have removed all four capacitors (R, G, B, and Composite lines) and also removed the 'hidden' 75ohm resistor to ground on the Composite line. The only thing I have left in place is the 100ohm resistor on the +5v line.

Quote from: viletim on October 12, 2008, 11:58:35 AM
It appears that the ENC-NUS doesn't like to have a high (in this case infinate) impedance source connected to its input (the CXA1645's datasheet warns about this too). What you normaly do with this type of input when you're not using it is connect a capacitor to ground. So just put some shorting links where the 110 ohm resistors used to be.
Phew! I was not looking forward to trying to resolder those 110ohm resistors (If they even still work after being stuck to the tip of my soldering iron) - I really hate surface mount stuff. I may not mind it so much if I had the right equipment, but that seems to be abit out of my price range  :'(

------------------------------------------------------------------------------------------------------------------------------------------

I was thinking about the best way of making this N64 'auto-switch' to the AV1/RGB channel using Scart. Usually when making a Scart cable for an NTSC SNES/N64 I just alter the wiring so that the +5v line powers both pin 8, and pin 16 (through a 100ohm resistor) making sure to cut the +12 wire (if its even present) so that the +5v cant return back down the cable and along the C-sync line. This is the only easy way I have found to make the NTSC SNES auto switch (without using batteries) as it doesnt have a +12v line internally you can use. The side effect of this of course is the the Scart cable now auto switches to AV1/RGB channel in 'Widescreen mode'.

I was wondering though - as the NTSC N64 does use +12v internally would it be safe to just directly link up the +12v line to the MultiAV output Pin 3? (making sure to isolate Pin 3 first by cutting the C-Sync trace on the board) This way it would auto switch to AV1/RGB in the correct 4:3 aspect ratio. Would this method be ok, or would I need to add a resistor/diode of some sort to make it 'safe' for both the N64 and TV?
------------------------------------------------------------------------------------------------------------------------------------------

Lastly, out of curiosity you mentioned previously that to build an N64 amp without removing/lifting components would be "possible but would require an amp with voltage gain. It would be fairly complex with just transistors, a high speed opamp or similar would reduce the parts count but the're not always easy to obtain"

I was just curious to know how many transistors would we be talking about? and how complex? I hope you dont mind me asking.

viletim

Quote from: Link83 on October 13, 2008, 01:43:06 AM
Quote from: viletim on October 12, 2008, 11:58:35 AM
It's no problem at all. The Dreamcast does exactly this when it detects an RGB cable is plugged in.
Really? How interesting. I wonder if that was this reason behind the issue this forum member was experiencing?:-
http://nfggames.com/forum2/index.php?topic=3047.0
Slightly OT but does the Dreamcast still generate C-Sync when in VGA mode?
Probably not. There's nothing wrong with a composite video signal with no video information on it.

QuoteSorry if I wasnt very clear - its only the RGB ouput that it seems to be having problems with the sync on bright images (Im assuming its the sync) the Composite output appears fine  ??? Its very strange and only happens for a moment.

Right... When you wire up the RGB amp and connect RGB (from amp) + composite video (from normal output) you have the problem. It goes away if you either switch to composite video for the video source or disconnect the amp input to the ENC-NUS. That's a bit odd... can you try it on another TV?


Quote
I was thinking about the best way of making this N64 'auto-switch' to the AV1/RGB channel using Scart. Usually when making a Scart cable for an NTSC SNES/N64 I just alter the wiring so that the +5v line powers both pin 8, and pin 16 (through a 100ohm resistor) making sure to cut the +12 wire (if its even present) so that the +5v cant return back down the cable and along the C-sync line. This is the only easy way I have found to make the NTSC SNES auto switch (without using batteries) as it doesnt have a +12v line internally you can use. The side effect of this of course is the the Scart cable now auto switches to AV1/RGB channel in 'Widescreen mode'.

I was wondering though - as the NTSC N64 does use +12v internally would it be safe to just directly link up the +12v line to the MultiAV output Pin 3? (making sure to isolate Pin 3 first by cutting the C-Sync trace on the board) This way it would auto switch to AV1/RGB in the correct 4:3 aspect ratio. Would this method be ok, or would I need to add a resistor/diode of some sort to make it 'safe' for both the N64 and TV?

I like to put a 220 ohm resistor in series with this SCART line. It protects the power supply in case of a short circuit.

Quote
Lastly, out of curiosity you mentioned previously that to build an N64 amp without removing/lifting components would be "possible but would require an amp with voltage gain. It would be fairly complex with just transistors, a high speed opamp or similar would reduce the parts count but the're not always easy to obtain"

I was just curious to know how many transistors would we be talking about? and how complex? I hope you dont mind me asking.

Depends how it's done... a textbook example whould be to use the existing circuit and place a two common emitter (need two cause they invert the signal) gain stages between the two existing emitter follower stages. Maybe 4 transistors, 10 resistors per amp.

Link83

#28
Please disregard the following paragraphs about the screen jumping - in the end I discovered the problem was caused by my TV when another SCART device was connected, the RGB amp was not at fault. I only leave the below text so the rest of the thread still makes sense.

Quote from: viletim on October 13, 2008, 11:43:07 PM
Right... When you wire up the RGB amp and connect RGB (from amp) + composite video (from normal output) you have the problem. It goes away if you either switch to composite video for the video source or disconnect the amp input to the ENC-NUS. That's a bit odd... can you try it on another TV?

I have tried it on another TV (both are CRT's if that makes any difference) and still the same problem - when using RGB on very bright images the screen 'jumps' momentarily - thats the only way I can describe it. Still only happens on RGB though - the Composite appears fine - which doesnt really make sense considering its using the same sync signal. Could the ENC-NUS possibly be affecting the amp in some way?

On another note I disconnected the wires between the ENC-NUS and the amp, and then linked the points where the resistors were to stop the 'jittering image' I was getting without them.

Unfortunately this seems to have added a problem all of its own. The picture no longer jitters but now there appears to be horizontal thin grey bands/lines appearing momentarily across the screen on bright images  ??? Im a positive they werent there before I removed the resistors. I can only assume that the ENC-NUS does actually need the 110ohm resistors? Also when I felt the ENC-NUS after a few minutes of the console being turned on it was quite warm/hot, especially underneath. I dont remember it being warm before - the VDC-NUS isnt warm at all....Maybe im worrying about nothing.

Quote from: viletim on October 13, 2008, 11:43:07 PM
I like to put a 220 ohm resistor in series with this SCART line. It protects the power supply in case of a short circuit.
Thanks  :) So forgive if im being stupid here, but when you connect up a 220ohm reistor on the +12v line, whats the voltage level afterwards? I tried working it out myself but all the equations I found said I need to know the current level aswell, which I dont. Also would a standard 1/4 watt 220ohm resistor be fine or should I use 1/2 watt?, and does Carbon/Metal Film make a difference when working with power lines? Sorry for so many questions (Can you tell I have only been modding for 8 months!)

Quote from: viletim on October 13, 2008, 11:43:07 PM
Depends how it's done... a textbook example whould be to use the existing circuit and place a two common emitter (need two cause they invert the signal) gain stages between the two existing emitter follower stages. Maybe 4 transistors, 10 resistors per amp.
I dont suppose theres any chance of a quick circuit diagram for this method aswell? (Or one using a common op amp if its easier, although I realise this would make it harder for some people to build) Im just curious as to any method that wouldnt require lifting/removing components.

Many thanks Viletim, I really appreciate all your hard work and help :) and I hope you dont think im being ungrateful by asking about an alternate amp design.

<UPDATE 17/01/09> I just wanted to update that since I posted about having interference in the picture saying "there appears to be horizontal thin grey bands/lines appearing momentarily across the screen on bright images" I managed to resolve the issue  :)

A few weeks ago I turned the console on and noticed a very slight buzzing/humming noise coming from the N64 - and I thought that cant be right coming from the silent N64! Moving closer I discovered the noise was actually coming fom the power supply. I swapped it for a spare N64 PSU I had and instantly the intereference was gone! I guess I somehow damaged the PSU during the modification (possibly when I accidentally put the expansion pak in backwards for a moment  ::))

Anyway, I am now once again getting a nice clear RGB picture, so thank again Viletim  ;D

Just wanted to let future readers know in case they ever encounter the same/similar problem.

Cyrix

so, could anyone from this topic make a final guide for good rgb booster/console modification.
at first i read everything and messed up what to do... :)

Cyrix


Cyrix

here was a reply with a link for china site with n64 mod before my reply. now i could not find that link someone deleted it...

venchia3

#32
Thank you all so much for this really interesting topic that gave me many informations! I'm from Italy and really interested and passionate about Nintendo 64 scene, even if it's considered a dead console, but I prefer playing with my roku-yon more than my Playstation 3  ;) Please forgive me for my stupid english ;)

I was about to mod my Japanese Nintendo 64 with RGB using the classic method of the three wires, but after seeing this forum, even if the topic is a bit old, I registered just to say to all of you how much you did! Thanks again!

I would like to ask something, if someone expert can tell me. As I told I have a Japanese N64 RGB enable board, I would like to do a definitive mod to my Nintendo 64, so what I'm asking is, the good steps for the definitive mod are:

- Lifting the 3 legs of the NUS chip
- Building the circuit board as shown by Viletim and Link83
- Remove from the Nintendo board what Viletim pointed as not necessary anymore
- Connect everything as Viletim shown in his driver schema

Something else? I have an original European GameCube RGB scart cable, should I have to modify it aswell? Remove capacitors or something particular? Or I can enjoy finally pure RGB without any modification of this cable?

Thank you so much anyway, even if no one will reply to me, I subscribed first of all just to make my congratulations ^^


Cheers from Italy!

viletim

venchia3,
I think the problem that Link83 reported can be avoided by changing some resistor values.
R1=300, R2=360. If you choose to lift the legs on the video DAC IC then you need only remove R8, R9, R10 and leave C124, C125, C126 in the board to be used as a coupling capacitor (ie. C1).

And yes, if you use a Gamecube cable you have to to remove the capacitors.