3DO 240p

[mario64]

Don't have Wolf 3D, but on Escape From Monster Manor, it runs a lot faster in official 240p mode for me. Makes the game a lot more fun imo, though some of the animations look silly.

Great thank you. That confirms it behaves identical to the 240p mod

[Taiji_Mods]

I was curious so I opened up my A-B Mode 240p FZ-1, and it seems citrus is right, it seems to be a wired in switch. I can't tell if the switch itself has its own things going on so that it automatically takes care of line placement.

At least now I know for sure both of my 3DOs can be RGB modded.

Spoiler

Just thought I'd add that not all Japanese Panasonic FZ-1 3DO's are equal when it comes to the MODE SELECT switch.

As you can see from my model on the outside it looks like a 480i/240p model but on the inside it is a different story.

I have tried Composite and S-Video on my capture card and it just stays in 480i A or B

It just seems like a redundant RF channel switch on my model at least.

@Leonprotos could you provide more detailed shots of the wired switch in your 3DO like removing that metal plating cover above the switch. I might be able to replicate it on my model.

EDIT: I did some more experimenting and it seems the switch does do something but it doesn't appear to operate as intended.

My capture card reports both modes as 480i which I know is incorrect. Mode B should be 240p. Also Another World doesn't speed up either which is expected in true 240p.

I tried the 3DO on one of my TV's and the Mode B 240p doesn't look as good as Mode A 480i

Mode B just looks very blurry. So I guess it is possible they altered this option in some way or my 3DO has a fault which is not
making the switch work correctly.

I would just like to add some information to the discussion. The 'Official' 240p Japanese FZ-1 is simply an after production factory modification using some wires and an empty RF box. Your 3DO seems to have had its wires removed. A 5V wire to the switch and a wire from the switch to the pad next to the interlace select pin (Pin 52) is all that's needed. The VP536 pulls the pin internally low to GND so an external GND wire is not needed. That would explain why you are not getting 240p when u use the switch.






You might be able to get 5V from inside the RF box instead. When I removed my PAL 3DO encoder (BT9103) I found a trace leading from the Interlace select pin to the CLIO (Pin 197) https://assemblergames.com/threads/new- ... 3do.69804/. To get reliable 240p switching, the resistor between the CLIO (Pin 197) and the encoder needs to be removed. Some models can't support 240p (US models with BT9101 encoder) and some have simply had it deactivated in production (CAN 3DO with BT9103, JAP FZ-1 with VP536?)

The issue with 3DO being unable to boot in 240p may be specific to FZ-10 bios as I haven't seen evidence of this problem with my PAL FZ-1.

I also found that the encoders on all 3DO (with the exception of the ANVIL chip) feed Hsync and Vsync signals back to the CLIO as a sort of timing function. This would explain why some games run at a different speed when progressive scan is activated. The only solution I can see to this would be to install a second RGB encoder which can generate its own progressive sync signal alongside the native encoder (BT856 for example). The original encoder could be left untouched to feed back interlaced Vsync to the CLIO while the new encoder could display progressive scan to the display.

[Syntax]

The one I just worked on had the 240p rf switch and no wires.
It connects via the 3 rf input prongs. Worked fine to switch between video modes.

[Taiji_Mods]

That's interesting. Do you have pics of the RF box and video encoder? Maybe it has an extra trace that the other wired 3DO don't.

[Whitesnake]

Thanks for the informative replies. @Taiji_Mods @Syntax

What would make me reluctant to add the wire modification to my 3DO is the switch is doing something but seemingly not correctly
as seen in the images I posted. It seems to do something when rapidly switching.

@Syntax seems to suggest the switch should work and is connected. Looking @Taiji_Mods picture it seems to be a different
RF Box compared to mine which may be the reason for the wiring.

However my thinking is if the box is just a defunct RF Modulator then those 3 pins are usually Ground, Composite, Power. right?

[the Goat]

However my thinking is if the box is just a defunct RF Modulator then those 3 pins are usually Ground, Composite, Power. right?

Correct. I do not see how the RF box could be involved in choosing between 480i and 240p output.

[Syntax]

Cant you just track the traces on the switch instead of speculating or do I have to ask my customer to bring his one back next time hes over so i can?

[Taiji_Mods]

Thanks to Syntax reply, I noticed an extra trace in your picture compared to the picture I had.



I would check for continuity between the 3 points in the RF box and Pin 52 on the VP536, and also 5V. It may be a problem with the switch machanism, do you have any pictures of the switch?

[Syntax]

You have some damn good eyes!
I had a quick look over the pictures in this thread and didn't spot that trace. Nice one

[CZroe]

However my thinking is if the box is just a defunct RF Modulator then those 3 pins are usually Ground, Composite, Power. right?

Correct. I do not see how the RF box could be involved in choosing between 480i and 240p output.

I would expect mono audio + composite video + 5v power (ground typically goes through the mounting tabs; no dedicated pin).

[Whitesnake]

I was able to get some spare time and dig the 3DO out and test some of the suggestions.

@Taiji_Mods I found no continuity on the trace coming from the VIA next to R170 and the two pads next to the GPS chip.
There is no continuity from GPS PIN 52 to CLIO PIN 219. There is continuity from the outer solder pad as your image suggests.

Here is my terrible attempt at shedding some more light on this.

Preview image below.
Red/Green = Traces Topside
Blue/Yellow = Traces Underside



Complete high resolution image here http://images2.imagebam.com/3f/c0/cc/0b ... 180014.JPG

I eventually gave up as it felt like I was just chasing grounds and wasn't really shedding anymore light on the situation.
I did manage to lead to some pins on chips that seem connected to the one connection coming from the RF Box.
Any idea what those pins do? I couldn't find any pinouts for the CLIO.

@Taiji_Mods would you be able to post a picture of your 3DO with the top of the RF Box exposed so I can see exactly how it is wired?
That would be really helpful. Or even a complete shot of the whole PCB exposed.

Other differences I have noticed between your 3DO and mine are two missing resistors on your PCB next to the RF Box.
If I'm going to replicate the switch on your 3DO I may have to remove them on mine. Removing L520 on my 3DO would seem to
disconnect a chunk of the circuit I traced

[Taiji_Mods]

You have some damn good eyes!
I had a quick look over the pictures in this thread and didn't spot that trace. Nice one

Thanks. I'm finding that I need to use loops more and more for soldering as I get older.

@Whitesnake It looks like you have been on a bit of an adventure. Though you have traced mainly 5V and GND signals so forget about them. The signal tracing at the RF box is useful though. I have illustrated on this picture where to check next. You need to check that the resistors are working properly and are connected to pin 52 on the VP536. Please check the resistance between the points. The 2 RF connections should read 1k each through their respective resistors and there should be continuity between the resistor at LC520 and pin 52 of the VP536. Please post your readings.



The FZ-1 service manual can be found on RetroRGB's website http://s3.amazonaws.com/retrorgb/assets ... NIC_EN.pdf It is the PAL/CAN edition but the RF circuitry remains the same. This might be useful with future probing.

I don't own the FZ-1 with the wire install so I can't get better pictures of the RF box, sorry. The components on your FZ-1 are 1K resistors where filters are originally used (L520, LC520). This bridges 5V to the RF box and back out again. On the FZ-1 i posted, these are not present and wires are used instead.

[Whitesnake]

Yesterday I actually did the mod based on what you had posted before.



I also removed the two resistors.

Now the interesting thing about this mod is it performs exactly the same as it was before without the wires.

So I started thinking that the real issue here is that my devices are not properly accepting 240p over S-Video.

I tried 3 TV's that have S-Video input and I couldn't see any improvement other than some kind of disruption to the image
when switching.

My StarTech USB3HDCAP keeps saying 480i regardless of the switch position, I tried my Amiga CD32 in NTSC with S-Video
and that would only display 480i on the capture card too. Even though games on that should run at 240p or 288p (PAL) when in gameplay.

I think my equipment for testing this stuff has had me chasing ghosts and the switch worked all along.

Sadly because all my devices are interpreting 240p as 480i this makes the 240p option on the 3DO a bit redundant for me.
I wonder how many other people are experiencing this issue?

The disappointing thing is my StarTech USB3HDCAP should work with 240p as I tried the Gamecube with S-Video and it recognises
240/288p Maybe some kind of alteration to the driver's resolution modes is required? (beyond my skills unfortunately).

Seems like the only solution now is something like the RetroTink 2X but I feel like that device is a little on the high side price wise
for what it does, I'd expect at least 1080p output on a £100 device. 480p is starting to lose it's appeal on modern TV's due to interpolation and scaling.
I already have an OSSC but that of course doesn't do S-Video.

Just as a side note I did notice the 3DO BIOS graphics move a little faster when in 240p mode so I guess that is another confirmation that the mode is working.

Any thanks for the help and feedback guys.

Maybe someone else can trace the non-wired version and see how it is switching differently?

[NoAffinity]

The USB3HDCAP seems to not handle 240p well. From another post on this forum, this may be helfpul (it was in response to a question about the USB3HDCAP):

240p is interpreted as 480i over composite and S-video, but no actual visual information is lost. To make it look right, you need to ignore the deinterlacing options in the capture card's setting itself (keep it on "Weave (None)") and deinterlace the video with a "double frame rate, top field first" method instead (or maybe "bottom field first" in some weird exceptions that I might not be aware of). I know that VirtualDub has an option for this. I'm not sure about Amarec.

This info might be helpful as well, if you weren't already aware of it: https://thethrillness.blogspot.com/2015 ... eview.html

I was having the same chasing-my-tail experience with the 3DO, with my devices treating the progressive output as 480i. My problem was compounded by installing the mod when I had a plasma that would accept it as 240p, but soon after I replaced that plasma, and didn't realize what was going on with my other devices. Just recently got a 27" trinitron which handles 240p gloriously, and my FZ-10 mod definitely works and looks great on the CRT.

[Whitesnake]

The USB3HDCAP seems to not handle 240p well. From another post on this forum, this may be helfpul (it was in response to a question about the USB3HDCAP):

240p is interpreted as 480i over composite and S-video, but no actual visual information is lost. To make it look right, you need to ignore the deinterlacing options in the capture card's setting itself (keep it on "Weave (None)") and deinterlace the video with a "double frame rate, top field first" method instead (or maybe "bottom field first" in some weird exceptions that I might not be aware of). I know that VirtualDub has an option for this. I'm not sure about Amarec.

This info might be helpful as well, if you weren't already aware of it: https://thethrillness.blogspot.com/2015 ... eview.html

I was having the same chasing-my-tail experience with the 3DO, with my devices treating the progressive output as 480i. My problem was compounded by installing the mod when I had a plasma that would accept it as 240p, but soon after I replaced that plasma, and didn't realize what was going on with my other devices. Just recently got a 27" trinitron which handles 240p gloriously, and my FZ-10 mod definitely works and looks great on the CRT.

Yep, I've tried deinterlacing options on various different pieces of software and I find the sharpness and decreased input latency that a proper 240p interpreted signal would have given is still lost. I think I'll just wait and see if an S-Video option for the OSSC comes out.
It is sad there are no open community drivers for the USB3HDCAP as I'm sure it could support a lot of modes and I doubt there will be any other capture cards that will even come close to the USB3HDCAP in the future either

As a side note it seems like the benefits the USB3HDCAP once brought are slowly being erased by newer drivers, such as 240p RGB no longer working and blue screen crashes. The only real benefit it still has to me these days the ability to support quite a few of the out of spec HDMI modes the OSSC puts out.

[nmalinoski]

Please forgive my ignorance about the technical layout of the 3DO. In the models that have discrete GPUs and encoders (Basically anything that doesn't have an Anvil chip, right?), do the GPUs send 240p to the encoders?

If they do, I think it would make sense to develop an aftermarket encoder, a la the UltraHDMI, that can take that tap those digital RGB lines and output analog RGB and/or HDMI, with or without scaling/framerate normalization. I imagine that would get you proper 240p RGB/HDMI output without affecting the rendering speed, and any HDMI option (even with simple line-doubling in lieu of a scaler as with UltraHDMI) should be sufficiently forwards-compatible, at least for the foreseeable future.

[Whitesnake]

Please forgive my ignorance about the technical layout of the 3DO. In the models that have discrete GPUs and encoders (Basically anything that doesn't have an Anvil chip, right?), do the GPUs send 240p to the encoders?

If they do, I think it would make sense to develop an aftermarket encoder, a la the UltraHDMI, that can take that tap those digital RGB lines and output analog RGB and/or HDMI, with or without scaling/framerate normalization. I imagine that would get you proper 240p RGB/HDMI output without affecting the rendering speed, and any HDMI option (even with simple line-doubling in lieu of a scaler as with UltraHDMI) should be sufficiently forwards-compatible, at least for the foreseeable future.

There are a few variations of RGB encoders for the 3DO that tap the digital signalling but they are a mess to install (Like PS2 modchip mess) I'm hoping someone comes up with a neater RGB install that reduces the wiring.

I've seen some on Youtube but nothing for sale.

This mod https://www.youtube.com/watch?v=2pAw7p2mREk is already light years ahead of the ones that are currently on sale.

[nmalinoski]

Please forgive my ignorance about the technical layout of the 3DO. In the models that have discrete GPUs and encoders (Basically anything that doesn't have an Anvil chip, right?), do the GPUs send 240p to the encoders?

If they do, I think it would make sense to develop an aftermarket encoder, a la the UltraHDMI, that can take that tap those digital RGB lines and output analog RGB and/or HDMI, with or without scaling/framerate normalization. I imagine that would get you proper 240p RGB/HDMI output without affecting the rendering speed, and any HDMI option (even with simple line-doubling in lieu of a scaler as with UltraHDMI) should be sufficiently forwards-compatible, at least for the foreseeable future.

There are a few variations of RGB encoders for the 3DO that tap the digital signalling but they are a mess to install (Like PS2 modchip mess) I'm hoping someone comes up with a neater RGB install that reduces the wiring.

I've seen some on Youtube but nothing for sale.

This mod https://www.youtube.com/watch?v=2pAw7p2mREk is already light years ahead of the ones that are currently on sale.

I was thinking a flex cable or QSB, like what we've seen with the UltraHDMI or GC/WiiDual; but that Doujindance mod is clever--I like how it floats above the GPU--and looks straightforward to install; however, it's still only an analogue RGB mod. It's absolutely an improvement over S-Video, but it doesn't do anything for compatibility with contemporary displays or video equipment.

Also, I doubt that design could properly accommodate HDMI; IANAEE, but I'm pretty sure the video proc and HDMI transmitter need to be fairly close, if not adjacent to each other, which means locating that mod at the rear of the console and tapping the GPU with a 3-sided QSB or flex cable (which is precisely what the WiiDual does). As a bonus, a dual-output, RGB/HDMI solution would reduce the amount of wire needed to hook up an 8-pin DIN, and it could theoretically also provide line-doubled or scaled RGB on the analogue output.

[Taiji_Mods]

Please forgive my ignorance about the technical layout of the 3DO. In the models that have discrete GPUs and encoders (Basically anything that doesn't have an Anvil chip, right?), do the GPUs send 240p to the encoders?

If they do, I think it would make sense to develop an aftermarket encoder, a la the UltraHDMI, that can take that tap those digital RGB lines and output analog RGB and/or HDMI, with or without scaling/framerate normalization. I imagine that would get you proper 240p RGB/HDMI output without affecting the rendering speed, and any HDMI option (even with simple line-doubling in lieu of a scaler as with UltraHDMI) should be sufficiently forwards-compatible, at least for the foreseeable future.

Excluding the combined chip 3DO (Anvil), the 3DO with the CLIO VDP sends 24bit digital RGB to the video encoder (BT9101, BT9103, VP536) for conversion to S-Video and CVBS. The video encoders are responsible for producing the VSYNC, HSYNC and CSYNC (on Luma). Some of the encoders can also switch between interlace and progressive scan.

It is possible to design a hdmi mod but it would need to be line doubled at least to 480p to meet minimum hdmi spec.

I've got a new version of my current 3DO RGB board due out in the new year which plugs onto the top of the current encoder so only the output wires need soldering. I may look at HDMI in the future but my current mod is geared towards use with OSSC for modern displays.

[mario64]

Please forgive my ignorance about the technical layout of the 3DO. In the models that have discrete GPUs and encoders (Basically anything that doesn't have an Anvil chip, right?), do the GPUs send 240p to the encoders?

If they do, I think it would make sense to develop an aftermarket encoder, a la the UltraHDMI, that can take that tap those digital RGB lines and output analog RGB and/or HDMI, with or without scaling/framerate normalization. I imagine that would get you proper 240p RGB/HDMI output without affecting the rendering speed, and any HDMI option (even with simple line-doubling in lieu of a scaler as with UltraHDMI) should be sufficiently forwards-compatible, at least for the foreseeable future.

Excluding the combined chip 3DO (Anvil), the 3DO with the CLIO VDP sends 24bit digital RGB to the video encoder (BT9101, BT9103, VP536) for conversion to S-Video and CVBS. The video encoders are responsible for producing the VSYNC, HSYNC and CSYNC (on Luma). Some of the encoders can also switch between interlace and progressive scan.

It is possible to design a hdmi mod but it would need to be line doubled at least to 480p to meet minimum hdmi spec.

I've got a new version of my current 3DO RGB board due out in the new year which plugs onto the top of the current encoder so only the output wires need soldering. I may look at HDMI in the future but my current mod is geared towards use with OSSC for modern displays.

This is great news. Will this mod work around the speed issue with the current 240p toggle switch? Games running too fast I mean

[Taiji_Mods]

The new version will be socketed, support 240p switching and have horizontal picture adjustment. Regarding the 240p game timing issue, I still need to do some more testing as it will require an encoder which can also generate its own sync.

[Whitesnake]

The new version will be socketed, support 240p switching and have horizontal picture adjustment. Regarding the 240p game timing issue, I still need to do some more testing as it will require an encoder which can also generate its own sync.

That is interesting. When you say socketed I assume you mean a socket that will push fit on top of the GPS chip? Any ETA and ball park pricing?

[Taiji_Mods]

The new version will be socketed, support 240p switching and have horizontal picture adjustment. Regarding the 240p game timing issue, I still need to do some more testing as it will require an encoder which can also generate its own sync.

That is interesting. When you say socketed I assume you mean a socket that will push fit on top of the GPS chip? Any ETA and ball park pricing?

Yes the idea is to use a 68 pin PLCC socket to push fit on top of the native video encoder which will remove the need to solder so many wires. Hopefully it will be ready early 2019. I need to design and test 3 different boards to cover the 3 different encoders (BT9101, BT9103 and VP536). The US models with BT9101 cannot be socketed but I'm also working on a separate method to restore 240p as well as RGB to those models.

[FriendofSonic]

From what I read, it seems like this was settled, but I did want to add to the discussion-- since I have a few JP 3DO units with the mode switch, I noticed 1 was fully wired for the mode switch, and two of them were not wired at all (but functioning fine, likely because of the additional trace(s) . Maybe the very early units that were assembled had the wiring before revised boards were fabricated?

EDIT: horrible bump on my part, but I think we all needed more 3DO love! Afterall, this is a board about shmups and what better system for shmups than the 3DO!