shmups.system11.org

Shmups Forum
 
* FAQ    * Search
 * Register  * Login 
It is currently Sat Sep 18, 2021 8:13 am View unanswered posts
View active topics



Post new topic Reply to topic  [ 107 posts ]  Go to page Previous  1, 2, 3, 4
Author Message
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Sun Apr 04, 2021 10:01 am 



Joined: 07 Mar 2021
Posts: 1
Any news Opatus on how the project is coming along? :D


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Fri Apr 30, 2021 3:12 pm 


User avatar

Joined: 25 May 2020
Posts: 37
Pal360 wrote:
Any news Opatus on how the project is coming along? :D

Unfortunately the project was on hold since Christmas. I will try to find some time to continue working on it.


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Tue May 04, 2021 1:35 pm 



Joined: 05 Dec 2012
Posts: 535
I had committed to helping and completely dropped the ball :( Between the package delayed for months in shipping, getting sick and keeping up with the day-to-day, I just totally failed. If anyone wants to help, please DM me a Digikey cart that I can just click on and order, then I'll get the final parts. I'm really, really sorry for the delay.


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Wed Jun 02, 2021 7:24 am 


User avatar

Joined: 30 Jan 2012
Posts: 27
Opatus wrote:
Here is a video and an image:

https://youtu.be/n3JoIReIQJw

Yeah, that looks flawless. Awesome job! 8)


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Sun Jul 25, 2021 5:56 pm 



Joined: 21 Oct 2016
Posts: 48
:shock:


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Sun Jul 25, 2021 10:31 pm 


User avatar

Joined: 21 Jun 2019
Posts: 160
Location: Atlanta, GA
3-Chip isn't a common term. I get that it's a rare version of the 1-Chip SNES but, unlike other 1-Chips, has no Csync before modding.

Is 3-Chip in the SNES Jr. form only or is it in the normal form factor with native RGB?

Sorry if I missed it in the thread, but does the board have a Bill of Materials or is it closed source? Not against closed source for business reasons. I'm just suspicious of another SNES mod's addition with no proof that it does anything and the chip looks like a Chinese counterfeit.

Still surprises me how little testing or validation of a board is needed to get people to install it in their consoles. First version of NES and N64 RGB mods had flaws that got fixed in the next version. Pre-H revisions of SD2SNES have an audio flaw.


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Sun Jul 25, 2021 10:45 pm 


User avatar

Joined: 05 Mar 2018
Posts: 1612
3 chip refers to all SNES / SFC revisions prior to the 1CHIP

CPU
PPU-1
PPU-2

3 chips


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Mon Jul 26, 2021 9:41 am 


User avatar

Joined: 20 Jul 2017
Posts: 683
NewSchoolBoxer wrote:
3-Chip isn't a common term. I get that it's a rare version of the 1-Chip SNES but, unlike other 1-Chips, has no Csync before modding.

Is 3-Chip in the SNES Jr. form only or is it in the normal form factor with native RGB?

Sorry if I missed it in the thread, but does the board have a Bill of Materials or is it closed source? Not against closed source for business reasons. I'm just suspicious of another SNES mod's addition with no proof that it does anything and the chip looks like a Chinese counterfeit.

Still surprises me how little testing or validation of a board is needed to get people to install it in their consoles. First version of NES and N64 RGB mods had flaws that got fixed in the next version. Pre-H revisions of SD2SNES have an audio flaw.


Completion and widespread availability of this mod would be a total gamechanger for the SNES, it's a huge deal really. It would mean that pretty much any SNES can match the video output quality of the 1CHIP with none of the drawbacks. No more playing lottery with serial numbers on eBay, just get any old SNES and install this mod. It also lays a strong foundation for future HDMI mods.

I'm not sure your suspicions have any basis in reality tbh, as OP has had long technical discussions with a bunch of established and respected developers in this thread, it's obvious they know what they are doing.

Of course there may be some bumps in the road for early adopters, but that's just how it goes and it all just contributes to a better mod and more community knowledge in the long run.


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Mon Jul 26, 2021 3:31 pm 


User avatar

Joined: 25 May 2020
Posts: 37
NewSchoolBoxer wrote:
Sorry if I missed it in the thread, but does the board have a Bill of Materials or is it closed source? Not against closed source for business reasons. I'm just suspicious of another SNES mod's addition with no proof that it does anything and the chip looks like a Chinese counterfeit.

Still surprises me how little testing or validation of a board is needed to get people to install it in their consoles. First version of NES and N64 RGB mods had flaws that got fixed in the next version. Pre-H revisions of SD2SNES have an audio flaw.


The project won't be closed source, worst case it will be never quite ready and won't be open either. The basics of this project have been already investigated a few years ago. The point of my project was to further research and try to find a solution for the problems that were found. Progress has been very slow lately, I had absolute no time for it.

I'm wondering which chip do you refer to by "Chinese counterfeit". There's a PPU2 from a 3-Chip SNES (PAL) and an Altera CPLD. You should also understand that most mods are made by one person and not a team with limited resources, some mistakes are mostly inevitable. Even mainstream commercial products go through revisions, correcting some early mistakes. So I don't really get where your suspicion comes from. I also believe that for all your examples a solution was made available for the early adopters.


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Tue Jul 27, 2021 5:56 pm 



Joined: 27 Sep 2019
Posts: 48
I was thinking about this project again recently and wondered if there was any news about it. Have any new issues cropped up since your most recent prototype(s)? I might take another stab at this to help out where I can. Maybe play-testing and video captures if nothing else.

Inspired by this project, I successfully designed and installed a similar PCB adapter for my reverse LPF project, so I'm a little more confident in making something for the digital RGB outputs again. My SNES mobo is an NTSC GPM-02 unit, so all the chips are crammed close together and there is little clearance between S-PPU2 and the cartridge connector. I had to offset the position of S-PPU2 on the mod board, relative to its original footprint on the mobo, to get some decent clearance around pins 81-100. It was kind of an awkward design, but worked out well enough.

Is it possible to implement the mode7 fix easily? I got kinda lost in understanding how you accomplished it. If it can be done with some discrete logic gates/inverters, that would be nice.

From what I remember, some of our video output results differed and I didn't see a lot of the issues you described. I wonder if it's related to the S-PPU2 revision (I have revision C), or the region (NTSC vs PAL). In any case, everything I saw on the analog output looked good (except for the mode 7 graphical glitches I reported) when I did the following modifications:

- TST15 (pin 93) tied to VCC
- TST12, TEST13, and TST14 (pins 90, 91, and 92) floating
- OVER1 and OVER2 (pins 37 and 50) tied to VCC

I'll need to dust off the test unit to double-check for any other graphical problems.


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Thu Jul 29, 2021 1:36 pm 


User avatar

Joined: 21 Jun 2019
Posts: 160
Location: Atlanta, GA
maxtherabbit wrote:
3 chip refers to all SNES / SFC revisions prior to the 1CHIP

CPU
PPU-1
PPU-2

3 chips


I see, so in German language discussion, they call the 2CHIP the 3CHIP. I couldn't find a reference to what a 3CHIP was so assumed it was related to the rare 1CHIP-03 revision. Marketing a mod to the US and Canada, my advice is calling it the 2CHIP like in the popular (English) RetroRGB and My Life in Gaming content. Not saying English terms are better, we got the "composite sync" vs "composite video as sync" mess.

Kez wrote:
NewSchoolBoxer wrote:
3-Chip isn't a common term. I get that it's a rare version of the 1-Chip SNES but, unlike other 1-Chips, has no Csync before modding.

Is 3-Chip in the SNES Jr. form only or is it in the normal form factor with native RGB?

Sorry if I missed it in the thread, but does the board have a Bill of Materials or is it closed source? Not against closed source for business reasons. I'm just suspicious of another SNES mod's addition with no proof that it does anything and the chip looks like a Chinese counterfeit.

Still surprises me how little testing or validation of a board is needed to get people to install it in their consoles. First version of NES and N64 RGB mods had flaws that got fixed in the next version. Pre-H revisions of SD2SNES have an audio flaw.


Completion and widespread availability of this mod would be a total gamechanger for the SNES, it's a huge deal really. It would mean that pretty much any SNES can match the video output quality of the 1CHIP with none of the drawbacks. No more playing lottery with serial numbers on eBay, just get any old SNES and install this mod. It also lays a strong foundation for future HDMI mods.

I'm not sure your suspicions have any basis in reality tbh, as OP has had long technical discussions with a bunch of established and respected developers in this thread, it's obvious they know what they are doing.

Of course there may be some bumps in the road for early adopters, but that's just how it goes and it all just contributes to a better mod and more community knowledge in the long run.


I'm sorry, do you have modder friends who made design mistakes without performing due diligence that costumers paid another $50 + shipping to fix? Is being an expert modder the same as being an expert electrician or knowing you're supposed to terminate video with 75 ohm resistors?

This is a gamechanger? I bought a 1CHIP-02 SFC from Japan and I can't tell a difference in video quality from my 2CHIP just eyeballing on an L2. So we're going to perma destroy SNES consoles by removing their PPUs instead of fix them and put said chips on PCBs with FPGAs? I totally get that soldering the pins of a chip no longer made is too risky but how long is the aged PPU from a broken console going to last being run off-spec? After it dies, does the customer pays another $50 + shipping to revert back to umodded state?

I only mention FPGAs because they're expensive. Is this mod to cost less than the $90 / €76 I paid for 1CHIP + power supply? You know why I look over mods even though I don't do them? I credit Voultar and RetroRGB for educating modders to switch from THS7314 to THS7343 in 2015-2016. I joined this scene in 2019 so I had to imagine one person using the 14 then everyone copying the first solution without actually checking what other TI chips existed.

Meanwhile, every junior in electrical engineering learns not to chain two LPFs together without either calculating, simulating or measuring what the cutoff frequency reduces to. Phase and group delay would be considered, as well as ripples in the passband and the order of the filters for stability or load concerns. I'm not trying to sound opaque, these are fundamental topics.

Opatus wrote:
The project won't be closed source, worst case it will be never quite ready and won't be open either. The basics of this project have been already investigated a few years ago. The point of my project was to further research and try to find a solution for the problems that were found. Progress has been very slow lately, I had absolute no time for it.

I'm wondering which chip do you refer to by "Chinese counterfeit". There's a PPU2 from a 3-Chip SNES (PAL) and an Altera CPLD. You should also understand that most mods are made by one person and not a team with limited resources, some mistakes are mostly inevitable. Even mainstream commercial products go through revisions, correcting some early mistakes. So I don't really get where your suspicion comes from. I also believe that for all your examples a solution was made available for the early adopters.


I read through the https://circuit-board.de/ thread in rough machine translation. I didn't realize you had ikari, borti and rama in the scene. Real heavyweights.

Sorry, I didn't mean to allude to counterfeit chips applying to you because they don't. I'm fascinated by the technical detail and dedication on display for my favorite console. I believe in progress even if I disagree with end product. I should be helpful then.

You have to consider the electrical noise of the FPGA with its high switching speed and complexity. Could be causing much of your existing problems. I would not have placed it right next to the PPU and the PCB going under the steel EMI shield (or forcing its removal?) is unfortunate. Aluminum and copper EMI shields are better but cost more.


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Thu Jul 29, 2021 2:39 pm 


User avatar

Joined: 05 Mar 2018
Posts: 1612
NewSchoolBoxer wrote:
maxtherabbit wrote:
3 chip refers to all SNES / SFC revisions prior to the 1CHIP

CPU
PPU-1
PPU-2

3 chips


I see, so in German language discussion, they call the 2CHIP the 3CHIP. I couldn't find a reference to what a 3CHIP was so assumed it was related to the rare 1CHIP-03 revision. Marketing a mod to the US and Canada, my advice is calling it the 2CHIP like in the popular (English) RetroRGB and My Life in Gaming content. Not saying English terms are better, we got the "composite sync" vs "composite video as sync" mess.

What? I'm not german. The people who call them 2-chip simply can't count


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Thu Jul 29, 2021 4:25 pm 


User avatar

Joined: 20 Jul 2017
Posts: 683
NewSchoolBoxer wrote:
I'm sorry, do you have modder friends who made design mistakes without performing due diligence that costumers paid another $50 + shipping to fix? Is being an expert modder the same as being an expert electrician or knowing you're supposed to terminate video with 75 ohm resistors?

This is a gamechanger? I bought a 1CHIP-02 SFC from Japan and I can't tell a difference in video quality from my 2CHIP just eyeballing on an L2. So we're going to perma destroy SNES consoles by removing their PPUs instead of fix them and put said chips on PCBs with FPGAs? I totally get that soldering the pins of a chip no longer made is too risky but how long is the aged PPU from a broken console going to last being run off-spec? After it dies, does the customer pays another $50 + shipping to revert back to umodded state?

I only mention FPGAs because they're expensive. Is this mod to cost less than the $90 / €76 I paid for 1CHIP + power supply? You know why I look over mods even though I don't do them? I credit Voultar and RetroRGB for educating modders to switch from THS7314 to THS7343 in 2015-2016. I joined this scene in 2019 so I had to imagine one person using the 14 then everyone copying the first solution without actually checking what other TI chips existed.

Meanwhile, every junior in electrical engineering learns not to chain two LPFs together without either calculating, simulating or measuring what the cutoff frequency reduces to. Phase and group delay would be considered, as well as ripples in the passband and the order of the filters for stability or load concerns. I'm not trying to sound opaque, these are fundamental topics.


I'm not really sure what kind of point you're trying to make here. The 1CHIP SNES is way sharper than the older models, that is extremely well established. Congrats on getting one for less than half the going rate, shall we make a quick checklist of all the stuff you don't need and put the word out for everyone to abandon any related hobby projects?
Nobody is trying to sell this as a mod, it's basically a proof of concept at this point. They are just experimenting with the SNES and talking about the results. The community is built on the shoulders of people who do work like this in their free time, expecting very little in return. Your attitude just doesn't match this situation at all.


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Fri Jul 30, 2021 10:03 pm 


User avatar

Joined: 25 May 2020
Posts: 37
@yoshiyukiblade
Quote:
Is it possible to implement the mode7 fix easily? I got kinda lost in understanding how you accomplished it. If it can be done with some discrete logic gates/inverters, that would be nice.

From what I remember, some of our video output results differed and I didn't see a lot of the issues you described. I wonder if it's related to the S-PPU2 revision (I have revision C), or the region (NTSC vs PAL). In any case, everything I saw on the analog output looked good (except for the mode 7 graphical glitches I reported) when I did the following modifications:

It's nice reading that you got inspired. For the newer NTSC revisions I have created a different adapter board, with it it should be possible to fit the mod board upside down. But it was never actually tested.
I don't have any graphical issues anymore, at least I haven't seen any lately. Last time I have tied TST15 to VCC and I still didn't have any issues, but maybe at that time the PPU was running in NTSC mode.
You could do the mode7 fix with discrete logic. You have to check with the logic if you are in the problematic mode7 sub-mode. And then you have to pull down the data on the B data bus of the PPU2 if the OVER signal is 0. For the data on the B bus I use a SN74HCT541PW with pull-down resistors.

Here how it's done in the FPGA:

wire screen_over_data = DATA[7] & !DATA[6];
wire mode_data = DATA[0] & DATA[1] & DATA[2];
assign MD7PAT = over & !OVER1;
assign over = mode & screen_over;
if(!PAWR && (PADDRESS [7:0] == 8'b00000101)) mode <= mode_data;
if(!PAWR && (PADDRESS [7:0] == 8'b00011010)) screen_over <= screen_over_data;

@NewSchoolBoxer
Quote:
I see, so in German language discussion, they call the 2CHIP the 3CHIP. I couldn't find a reference to what a 3CHIP was so assumed it was related to the rare 1CHIP-03 revision. Marketing a mod to the US and Canada, my advice is calling it the 2CHIP like in the popular (English) RetroRGB and My Life in Gaming content. Not saying English terms are better, we got the "composite sync" vs "composite video as sync" mess.

Calling it 3-chip makes more sense as the 1-chip SNES combines 3 chips. But calling it 2-chip is also acceptable because the PPU1 and PPU2 can be considered as 1 chip.
Quote:
This is a gamechanger? I bought a 1CHIP-02 SFC from Japan and I can't tell a difference in video quality from my 2CHIP just eyeballing on an L2. So we're going to perma destroy SNES consoles by removing their PPUs instead of fix them and put said chips on PCBs with FPGAs? I totally get that soldering the pins of a chip no longer made is too risky but how long is the aged PPU from a broken console going to last being run off-spec? After it dies, does the customer pays another $50 + shipping to revert back to umodded state?

I only mention FPGAs because they're expensive. Is this mod to cost less than the $90 / €76 I paid for 1CHIP + power supply? You know why I look over mods even though I don't do them? I credit Voultar and RetroRGB for educating modders to switch from THS7314 to THS7343 in 2015-2016. I joined this scene in 2019 so I had to imagine one person using the 14 then everyone copying the first solution without actually checking what other TI chips existed.

The video quality of the 1-chip SNES is so much better. I can clearly see it on my 14-M4 with RGB and on a 4k flat panel it looks awful.
As for the price I think BOM cost should be around $60 without assembly and installation. The FPGA/CPLD used here should be around $10 (but I'm not entirely sure if all the planned features would fit), of course it is out of stock right now.


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Tue Aug 03, 2021 10:33 am 



Joined: 27 Sep 2019
Posts: 48
Opatus wrote:
You could do the mode7 fix with discrete logic. You have to check with the logic if you are in the problematic mode7 sub-mode. And then you have to pull down the data on the B data bus of the PPU2 if the OVER signal is 0. For the data on the B bus I use a SN74HCT541PW with pull-down resistors.

I was wondering if we can naïvely apply an AND gate to OVER and VDB and connect the output to the VDB pins. I poked at the OVER signal and saw that it is normally held high until mode7 related content shows up (it is very active during the problem scenes). So when OVER is high, the output will just follow whatever VDB is doing. When OVER is low, the output will be low. I've never dealt with logic stuff, so let me know if this is a dumb idea or not :P.


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Tue Aug 03, 2021 2:48 pm 


User avatar

Joined: 25 May 2020
Posts: 37
The SNES has 3 options in what to do if the mode7 layer is zoomed out and thus it does not cover the whole display (see register 211A). It can either repeat the mode 7 "map", draw the base color (backdrop color) or draw the first entry in the palette. If I remember correctly the PPU1 calculates the repeated mode 7 map in every modes, but keeps the OVER signal high for the first mode (mode 7 repetition). The OVER signal (when driven low) is used by the PPU1 to tell the PPU2 in the latter 2 modes that the processed pixel is outside of the memory map. Than the PPU2 decides based on the register setting to draw the base color or the first palette entry. But if TST15 is pulled high the PPU2 won't display the base color correctly outside the mode 7 area.

So I believe you cannot use the OVER directly to control the data on VDB0..VDB7. You most likely don't have to check for mode 7 (the OVER signal can only change in mode 7) but you definitely have to evaulate register 211A. And this is not that difficult by using some logic.


Top
 Offline Profile  
 
 Post subject: Re: Sharp analog RGB for the 3-Chip SNES using digital signa
PostPosted: Tue Aug 03, 2021 4:03 pm 



Joined: 27 Sep 2019
Posts: 48
Ah, okay. This is definitely outside my current skill level as I have no experience in probing the registers to get the necessary information. I'll start with something simple in the meantime and work my way up from there. I'll have to learn about it sooner or later to get the brightness information as well. Thanks for the detailed description on how mode 7 operates and the role of OVER in these situations.

The current plan is to get an output that looks pretty and capture it!


Top
 Offline Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 107 posts ]  Go to page Previous  1, 2, 3, 4

All times are UTC


Who is online

Users browsing this forum: SamIAm MkII, Syntax and 12 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum

Search for:
Jump to:  
Space Pilot 3K template by Jakob Persson
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group