Xbox AVIP break out board

[professor_jonny]

That was the idea for the pull down to try to pull the mode pin down to 3v.
ill try to explain the logic.

If the system fired up and detected no pack it would try to set a avpack type and pull down a mode or mode pins to logic low (say 0.5v) and no lower set via divider .

If you plugged in a pack as they bridge the mode pins to ground it would change the voltage of the mode pin to 0v if the micro was already driving the mode pin low, this would enable me to detect if a pack were plugged in as it would drop the voltage lower on the mode pin and I could detect the change in pack.

same way the other way around as the pull down will try to pull the nominal 5v logic high down to say 3v (still logic high) so if a pack were plugged in with the unselected mode pin tied high it would detect the change there also, (we may not need this) but who knows what pack makers hav been up to.

I was going to try it out when I got my samples, it is sort of a ballancing act and it probally wont work worst case you will get no picture until you reboot if you plug a pack that does not drive a mode pin low that was not previously set.
One would just have to reset the console and it would all be fine and dandy.


I thaught some zeners in the circuit could be used to supply the pull down reference of 0.5v and a pull up reference of 3v as a more reliable idea, but it is getting complicated it requires a whole lot of components to do that and is it really that much of an issue? if you plug in a pack you may have to reboot.

on a slightly different tangent there is a possibility of 7 different pack types available by our 3 binary mode switches, but only so many are defined,

I came across an article browsing around that sugested that there was is a pal and a ntsc mono xbox av pack type:

http://www.gamesx.com/avpinouts/xbox.htm

Would this be a mode for the RF modulator packs? as all the other info I have found on the net don't list this many modes, I just thaught they were spare unused I have not coded for mono output modes, it could be handy for arcade cabinet conversions and people with broken head phones?.

Would someone from the northern world have one of these an be able to check the mode pin connections for me please?, I have a pal one I will test (well I asume it is pal)

[professor_jonny]

So I have purchased a whole lot of av packs of different brands and models and I have found not all are made the same and it seems the info on the net is not all that correct or pack manufactures seem to not follow the rule.

I have a few standard av pack's from china and mad cats ones that set advanced AV mode so they enable Svideo and coax output but only have output cables for the red yellow and white composite output.

I also have a cable here that is scart and sets s-video output mode be default then when you connect it to the tv it bridges the other mode pin to Change to rgm scart mode but there is no svideo output on the cable? I guess they have done this so when you plug it in it resets the xbox or they just chopped off the svideo cable from the dual cable to make a SCART only one?

The PAL RF modulator I tested sets sets NTSC mode going by all the info on the net.

Would someone be able to test a NTSC RF modulator and tell me which pins are connected in the mode select pins?

On another note I have some proto type boards and they all fit great apart from the locating/ mounting pins on the avip socket need to be enlarged but for a proto type not to bad I guess they will work well for testing.

I have reamed out the holes with a very fine drill bit by hand and now everything lines up perfectly and is flush seated with the motherboard they dont actually need plating on them as they are not connected to any thing any way.

They are very good quality for what I paid I'm impressed but they missed the solder mask on the vias on the bottom I specifically asked them to make sure they do so there is a potential for shorts if there is exposed copper or solder on the motherboard, also Few issues with the text size being illegable but I'm happy with the outcome.

SoIi have learned from this exercise you get what you draw even if I ask for something if it is not in the gerbers it is not going to be done, so a dab of solder mask will have to do the trick or some kapton tape or something.

It will also require a small change for a production board as the vias cover the landings of the mounting/locating pins for the socket and as the vias are not solder masked it will require insulation manually with me applying a solder mask or sticker of some sort.

I would also like to move the H amd V sync out put terminals just above the digital audio U.FL socket so the vias will be clear of the landing zone for the mounting pins of the socket.

So there will be a few minor changes nothing too serious but these ones will do for now.

Also I don't really deal with too much smd stuff and I now know that 0402 is not universally recognised as metric size for an smd pad, as I went to stick on some capacitors and they are twice as long as they are meant to be, oh well I know for next time

[sharpfork]

Super cool project, subscribed!

[professor_jonny]

I have been busy with other projects non xbox related but I finally received some sample u.fl connectors from overseas and I have done a little bit of testing but I need to get my scope out and do some signal integrity checks as I'm as per say not using the proper impedance plugs on the cable as they are 50 ohm not 75 ohm but I got them made with 75 ohm cable.

I did not see this being a problem as rca plugs are not 75ohm either and they are used for video signals and it all seems to work ok on every component or coax audio device on the planet but I want to be 100% certain I get no reflections or ghosting artifacts etc...

The other option is to use H.fl connectors that are rated for 75 ohm but they are several magnitudes greater in price compared to the little brother U.fl ones, and trying to keep the price in check is the issue.

So below is my current play thing and I have modified an ebay sourced component to hdmi board with coax audio slightly to include U.fl plugs wiping of the rca's in the process, for all purposes it is working great but I need to do some further testing and some mods to the program.

I think one of the av pack types is muddled up I have got lost in the negative binary logic, but I have now implemented every possible binary pack combo even the unused ones. This will allow me to test the unused modes to see if I get any signal outputs, there is no indication of what they do if anything on the net.

I will have to work out a method of labeling or marking the connectors or cables as every time I unplug them I never know which one is witch and end up pulling the board out so I think a small band or heat shrink may be on the cards or colored cables or something.

Note 0408 smd components suck never try to hand solder them, I'm going mad squinting at the pcb's soldering them, kids bumping the table and loosing them etc.... so there will be a revision to increase the sizes of the resistors footprint's for sure

[spicemuseum]

0408 smd components suck never try to hand solder them, I'm going mad squinting at the pcb's soldering them, kids bumping the table and loosing them etc.... so there will be a revision to increase the sizes

Great!

[sharpfork]

I built a reflow oven out of a toaster oven from goodwill and a zallus reflow kit. I bake my smd protos, much easier.

[professor_jonny]

$95 is kind of expencive I have a ss relay at work I could botch up something with an arduino and a pc fan might give it a try.

problem might be putting it in the oven with all the resistors in place.

[psyko_chewbacca]

I don't have a reflow oven yet(Using hot air gun from my soldering station) but I agree that solder paste soldering is just so much better! I'm using one of those chinese solder paste dispenser you hook an air compressor to it and I just cannot believe how easy and fast soldering narrow-pitched components has become. It has certainly cut down the time to solder a board in half (if not more) compared to traditionnal soldering.

Of course it becomes quite useless when soldering through-hole components.

If you do a lot of SMD soldering, investing in gear to use solder paste is really a good move, unless you have a lot of time to spare!

[professor_jonny]

This post is more to do with picture quality I have been having a play around and I want to capture some picture to show you all before and after but the only method I have is to put a camera in front of the tv, would any one have a suggestion ?

I do get screen flickering (bright/ dark) just on the threshold of noticing it, if I turn on the digital noise reduction on the tv the problem goes away, I suspect the converter quality is the issue, as I see no problems with the component integrity up to the hdmi board.

would any one have a hdmi converter and experience a similar issue.

on different resolutions in stead of being the whole screen it is like a comb effect cross the display and it seems to be worst a 567i mode and is almost invisible at 1080i.

on some resolutions I get a sparkle like snow effect but it is allways in the same position on the screen.

I could put component outputs on the back of the xbox and use a direct component connection but I don't want to cut the plugs of the sample u.fl cables.

I forgot to mention the sound seems to do the same thing via digital audio connection (quiet loud) from the motherboard and that really stumps me as it is a digital connection not analogue!!!

does any one have more insight on what I may be experiencing and what may be the cause of the issue ?

[xman]

I have a couple of suggestions PJ. I'm sure you have tried some or all of them but I'll list them all just in case, can't hurt can it?.
1- The converter is powered by the Xbox power supply?. Try powering it off the DVDrive connector's 5 or 12volt pins rather than the power supply to mother board connector. Two reasons, I don't know how noisy the 5 or 12volt from the Xbox Power supply actually is. The board caps may be there to filter out the noise. The other reason, you won't have supply wires running over the board that could be picking up noise simply by the way they run over the board getting to the converter to power it.
2- Connect the video ground to the audio ground and ground that to the cage in the Xbox. It should be already but just a jumper wire will confirm all is actually good.
3- Make sure the power supply mount screws are actually tight and the surfaces on the bottom of the Xbox power supply touching the metal cage in the Xbox are clean.
4- Check your cables from the machine to the T.V. itself. Simply wrap the cables with aluminum foil and connect a wire from the foil to the Xbox metal cage. If you notice any changes, your cables or there grounding is at fault.
5- Open the converter box and see if there is any type of power supply inside it. It may have one that is isolating the supply from the Xbox. You may need to bypass this, if there is one, and power it direct from the Xbox's supply so all the grounding will work as designed in the Xbox.
Without seeing any parts 1st hand, I really am clutching at straws PJ but maybe something here may have you looking at something you may have not thought of. I wish you good luck and hope you have these issues sorted soon. I need a couple of these boards when you have them sorted.

[professor_jonny]

I have a few small challenges more to do with the internal adaptor as you are terminating the signals into the 75ohm load of the adaptor so an external cable that uses the same video lines can or could degrade the quality as you are effectively splitting the output.

I have looked into cutting the video output to the board or using an amp chip but it will push it out of the price range of cheap cheap .

If you are using an internal adaptor it is likely going to be vga or component so the standard cable will work with no issuies, and really if you have an adaptor likely it is going to be outputting hdmi so it is not an issue.

[psyko_chewbacca]

This post is more to do with picture quality I have been having a play around and I want to capture some picture to show you all before and after but the only method I have is to put a camera in front of the tv, would any one have a suggestion ?

I do get screen flickering (bright/ dark) just on the threshold of noticing it, if I turn on the digital noise reduction on the tv the problem goes away, I suspect the converter quality is the issue, as I see no problems with the component integrity up to the hdmi board.

would any one have a hdmi converter and experience a similar issue.

on different resolutions in stead of being the whole screen it is like a comb effect cross the display and it seems to be worst a 567i mode and is almost invisible at 1080i.

on some resolutions I get a sparkle like snow effect but it is allways in the same position on the screen.

I could put component outputs on the back of the xbox and use a direct component connection but I don't want to cut the plugs of the sample u.fl cables.

I forgot to mention the sound seems to do the same thing via digital audio connection (quiet loud) from the motherboard and that really stumps me as it is a digital connection not analogue!!!

does any one have more insight on what I may be experiencing and what may be the cause of the issue ?

If you have an oscilloscope, you could try to probe DC voltage supply lines of the HDMI converter unit. Cheap chinese electronic devices includes cheap DC power supplies most of the time. These are generally made to supply a high amount of current but with lot of ripple. Video converter circuits require clean and sufficient power source to properly operate, especially when analog signals are involved!

For personnal education, Texas Instrument has set up a very comprehensive yet concise document to explain the phenomenon of parasitic noise on DC power supply lines. A potential solution for a noisy supply can be found in sectin 3.1:
https://e2e.ti.com/support/power_manage ... age-ripple

If the input voltage is an issue here, you can either replace your power supply with a higher quality one(like the internal DC supply of the Xbox) or try adding an array of capacitors at the power input of the converter unit. Depending on Volts/Amps requirement of your HDMI converter unit, you could try to add a high capacitance cap (like 100uF) and work your way down to 1uF in 3-4 steps. If the input voltage of the HDMI converter board is 5V, there's a high probability that the converter board itself has a step down converter circuit to bring DC voltage to 3.3V and even 1.8V, you might want to check ripple on these supply lines too.

Could you post hi-res pictures of the HDMI converter board so we could have a look at it?

Other than that, I would suggest to try a simpler setup and hook your interface board to your TV via component... There's still the impedance mismatch issue that could be part of the problem.

[professor_jonny]

I'm using the internal Xbox power supply the original adaptor it has massive ripple problems.

I have tested the output on the scope meter I have a fluke 199c 200mhz scope I have checked common and differential noise problems and nada, I think the converter is just plain bluntly (shit).

The board it self seems to inject ripple into the Xbox as with it connected I get the same problem on the other outputs when it is powered connected to the xbox, maybe it relys on its case for screening or is not suitable for use inside the Xbox?

The output over component direct from the board is perfect no visible difference between that of the official (china copy) component cables I have.

I do admit the hdmi board does a good job at blanking the black level giving you more contrast than the component cables.

I have put a large cap filter on the hdmi board and it seems to have done nothing, I'm going to purchase another adaptor I was hoping to use this one as it has digital audio input, so I can transmit digital audio through a amp to the tv.

[psyko_chewbacca]

Maybe the noise from the HDMI board's power input going back into the Xbox 5V power rail is causing issue with the video output of the Xbox itself; especially if you selected a power source near the AVIP port. Re-introduced noise could disturb one of the video output/amplification stage and screw up component output.

I'm not sure if I read you correctly but when you tested video output via component, was the HDMI board still powered ON via the Xbox but video input signals were disconnected from it?

Do you have any other sufficient, clean 5V source to power ON the HDMI board to separate it from the Xbox generating the video feed?

Also, a single big capacitor will not filter all noise. Bigger cap will only filter out low freq noise. That's why you need to put a small one in parallel of your big cap. You could try to put a 1uF or even a 0.1uF and see if that helps. It's possible it won't help much. In that case, we could assume the design of the power conversion stage of the HDMI board is just plain bad. Ferrite around the 5V wire between the Xbox and HDMI board can't do any harm too.

[professor_jonny]

I have put some smaller caps in parallel with the filtering caps i just thaught that goes without saying , I have some ferrites but only really big ones for mains, I think I will put the converter back together and try it the way it was intended see if I get the same results.
In hind sight I should of tested the adapter before jacking it all in and expecting it to be all good.

[psyko_chewbacca]

I think I will put the converter back together and try it the way it was intended see if I get the same results.
In hind sight I should of tested the adapter before jacking it all in and expecting it to be all good.

Ah! I can relate to that. So many times have I skipped testing units the way they are before modding them! I still have to pace myself into not doing this anymore. With all the trouble it brought me in the past, it should come instinctively to test before disassembling it but sadly no...

Anyway, from what you told me via PM, the artifact on the screen are most likely due to improper analog inputs filtering (maybe also amplified by impedance mismatch ). Do the artifacts change in density depending on the input resolution?

[professor_jonny]

Yip different resolutions yield a different pattern worst at 1080i.

The pattern is static for that resolution but I did not notice if it changed with picture change I just used a static image.

Filtering fixed the brightness flicker and sound problems but that was made worse by leaving component Av plugged in at the same time.

[psyko_chewbacca]

Yip different resolutions yield a different pattern worst at 1080i.
The pattern is static for that resolution but I did not notice if it changed with picture change I just used a static image.

That makes sense. Interlaced video signals with artifact will always look worst then it's progressive counterpart. Of course, the picture displayed on screen affects the visibility of the artifacts. Brighter/whiter images will look less worst than darker/blacker picture. This is due to the fact that the signal level ratio between the desired signal and the injected noise is far greater when image is darker as signal voltage levels are closer to 0V than when they are white and signal voltage is higher(0.7V or 1V, I cannot remember).

If you're still up for experimentation, I would look closely at the input filtering circuit of the 3 video lines. There should be at least a basic filtering circuit on every line comprised of a 75ohm resistor going to the video ground plane and a coupling capacitor in serie(I don't think these cheap devices have multi stage filters!). Value should range between 47nF and 56nF if max input resolution of the HDMI board is 720p/1080i. This cap could be of lower value if HDMI board accepts 1080p input(like half of that, so 15nF to 22nF). You could measure the values of these resistances and capacitor to check if they match; maybe replace them with high quality and high precision caps (like 51nF-52nF ideally but 47nF should be OK too) to be right below the cutoff frequency of the RC circuit. Cutoff frequency can be calculated with 1/(2*Pi*R*C) where R is the 75ohm resistance and C is the coupling capacitor. For accurate measuring, always probe component soldered on circuit and account for PCB trace impedance and inductive parasitic impedance added up by the components and solder; both are minimal but not negligible. That's why a 56nF might be too close to the max frequency of a 720p/1080i signal(37.13MHz). Of course a good quality LCR meter is in order here but from what I read, you are fairly well equipped!

Of course, the impedance mismatch is an issue that can just add up to the visual artifacts but I guess we'll have to live with it unless you're ready to pay for 75 ohm cable and connectors!

For anyone wondering what's this about impedance mismatch and have about 30 minutes to spare, there's an excellent video made in the 50s by AT&T/Bell that illustrate so well what is impedance and why it matters so much in the transmission of a high frequency signal. No special prior knowledge is really required to understand except maybe basic algebra. Recommended for everyone who wishes to understand more what's involved in the transmission of electrical signals, wether it's in a cable or wirelessly!
https://www.youtube.com/watch?v=DovunOxlY1k

Filtering fixed the brightness flicker and sound problems but that was made worse by leaving component Av plugged in at the same time.

Of course if you are both feeding the HDMI converter board with video/audio signal and have a Xbox video cable plugged in the AVIP port you'll see and hear some weird shit. Is that what you meant by this?

I'll be waiting on the photo shots of the board when you have the time. Maybe something else will come to mind once I see them.

[professor_jonny]

So I have done some futher tests at all resolutions and I only start to get a problem at 720p and 1080i all other resolutions offer perfect display over hdmi.

Below is the sample pictures at 1080i and 720p

You will see slightly off vertical dots in packs of three in the 1080i image, following a diagonal pattern

You will see vertical dots up the page in a vertical pattern in the 720p image.

To be honest I'm nit picking unless you really looked you would not notice the dots.
where the picture is dark these dots are not present.

These dots are static across all displays even moving pictures.

of note turning on the soften function makes the dots more noticeable.

These dots do not show on component cables when outputting the signal via an avpack with the converter hooked up and powered.

[professor_jonny]

the important chips on the board are:

CS8416 digital audio decoder
cat6613 hdmi transmitter
cat9883c Triple 8-bit Video A/D Data Converter
LM1117 regulator *4

there is also the micro but it is really hard to read the info on it

all video lines have a dual stage filter with a resistor in series with the signal and a cap to ground then it has a resistor in series into the a/d converter and a cap to Ground on the input.

as you see the filtering on the regulators is not all that flash I have restored the whole board to before I played with it.