N64 RGB mod Tim board, Borti firmware - Troubleshooting help

[Filurkatten]

Hi!

I installed the RGB board from Tim with Borti firmware a couple of days ago in my PAL N64, it had the DENC-NUS chip in it.
I'm using an NTSC style cable (with caps) and sync on Luma.

The sharpness of the picture is great and the colors also except it seems to have interference on Green and Sync problems.
The green colors are distorted, the most bright tones have distorted outlines.
Sync is also acting strange, it works great for a minute or two, sometimes longer sometimes shorter then it starts degrading.
Picture starts to roll, first just a bit then more until it's all gone.

Also, after the mod the composite video output don't seem to work like it should either.


Any directions I should take to troubleshoot the problem would be really appriciated!

Thanks in advance.

[kel]

Sounds like a short circuit between one or more of the data lines on the ribbon cable. Check both ends of the ribbon cable with a magnifying glass to make sure that there are no bridges as sometimes they can be difficult to see with the naked eye.

[borti4938]

I also got a mail that the recent version got some quirk on colors. It looked different from what I've seen there compared to your description, but it could be also a firmware issue. Could you try to use the recently uploaded firmware on GitHub. Maybe it helps.

Edit: we found out it was not the firmware on the other project; but it it a try nevertheless.

[Filurkatten]

Thanks for taking your time to answer.

Here a picture of the green distortion in a tree in sm64.
Theres a light green outline between the diffrent shades of green.
https://imgur.com/a/fcrU5L1

When the sync starts to act up red fuzzy vertical lines starts to show up. I'll try to get a picture of this later today, quite hard to get a good picture taken from the crt.


I got a magnifying fluorescens lamp and have gone throu all solder point, could not see a short anywhere.
Then I resoldered the cable once again and the problem still persists.

I bought the board allready flashed with your firmware, dont know if I have the tools to flash it myself. I'll look into it!

[Filurkatten]

Did some more tests;
removed the kit alltogether, composite video is fine.
Resoldered the kit once again, same problem, blue also get the outlines between the diffrent shades but only after a while.
Double checked for shorts, none.
I tied wiggle the cable to see if the problem was close to the kit or the board, nothing happend at kit end or console end but sometimes the picture flashed when bending the cable in the middle. Cant see anything wrong with the cable thou, looks fine. (using the colored ribboncable that came with the kit)

After trying testing for a while the picture starts to degrade more and more until the sync is completly lost, sometimes resetting the console helps against this sometimes not.

I did not have any tool avalible to try a reflash of the firmware.

I'm out of idéas.
Will try to change the cable for another one also got another N64 to try the kit on (not to keen on soldering in another console as this one seems to work fine when the kit is dissasembled and they got same motherboard revision and same DENC-NUS chip in them)

[borti4938]

Where are you from? Maybe you can send your N64 to someone who is familiar with the N64 kit installation and troubleshooting. At the moment it's quite difficult to help you beside of saying, double check connections and shorting installation wires.

[Filurkatten]

I live in Sweden.

Was hoping there where an common problem with an easy fix but it doesnt seem like it.
I tried the kit in an different N64 and with diffrent cables, same problem.

Either, I'm terrible at soldering (got a couple of years of soldering experience, I dont thinks thats the problem), both N64 are duds, the kit is a dud. or I killed the kit with a short when trying to install it the first time.

[kel]

Still sounds like a data line short to me taking into account the vertical rolling and that Vsync is sent on one of the data lines. If you've tried it in two consoles with different wiring then it seems like the fault is on the board, could even be a loose solder joint on the CPLD rather than a short which would account for the delay in the fault occurring while the board warms up.

[borti4938]

Filurkatten, you can try to reflow the input pins on the CPLD, which are pins 12 to 21.

What I can also offer is
a) to send an adapter board for teh RCP-NUS which the resistors on it. https://twitter.com/leonkiriliuk/status ... 8585103365
b) to look over it (shipping from Germany to Sweden is not that expensive at least if you send just the mainboard)

BR

[Filurkatten]

kel, borti4938, Ive reflowed the pins 12-21 on the CPLD the problem is still there.
I imagined that it worked for a bit longer this time, maybe, could just be becous I've done so many tests now.
The installations where the ones that worked the longest, when the console has been on for a little bit the video doesnt work at all, but when it has cooled of it starts to work again. I tried reflowing the pins going from RCP-NUS to the DENC-NUS, no difference.

borti4938, whats the advantages in using the adapter board with resistors?
You also got an N64RGB board aswell? Whats would your board cost to ship to sweden?

The seller of the my N64RGB board offered to do the installation for me, but shipping cost to and from England plus installation cost would be more expensive than just buying a new RGB board.

[CZroe]

With Borti’s latest FW on Tim’s board I noticed horizontal banding in the trees and some odd patterns in light colors that look almost like dithering. I had it installed with UltraHDMI using Borti’s flex PCB and that looked great until updating from 1.07A to 1.08. After that, I saw horrible visual artifacts everywhere in Super Mario 64 (my preferred test cart).

https://imgur.com/a/XwSwvjE

When it was artifacting that badly I couldn’t even downgrade it so I had to move the kit into another console.

While it seems that issue is caused by the UltraHDMI, the horizontal banding in the trees and the dithering-like patterns in RGB do not (disconnected UltraHDMI). I checked all my grounds to make sure it wasn’t jailbars or anything and I repositioned my analog lines but no dice.

[Filurkatten]

Thats exactly how it looks for me aswell!
I'm also using SM64 as a testcart.

[unmaker]

With Borti’s latest FW on Tim’s board I noticed horizontal banding in the trees and some odd patterns in light colors that look almost like dithering. I had it installed with UltraHDMI using Borti’s flex PCB and that looked great until updating from 1.07A to 1.08. After that, I saw horrible visual artifacts everywhere in Super Mario 64 (my preferred test cart).

https://imgur.com/a/XwSwvjE

When it was artifacting that badly I couldn’t even downgrade it so I had to move the kit into another console.

While it seems that issue is caused by the UltraHDMI, the horizontal banding in the trees and the dithering-like patterns in RGB do not (disconnected UltraHDMI). I checked all my grounds to make sure it wasn’t jailbars or anything and I repositioned my analog lines but no dice.

Same issue here. I've captured some footage so others can see the graphical issues more clearly:

https://streamable.com/tcp63

The graphical issues appear only on the HDMI output. The RGB output is crystal clear. I also was not able to downgrade to version 1.07A and had to swap out UltraHDMI kits. I tried repositioning wires and making new ground connections, including shortening the UltraHDMI ground with a short wire connected directly to the console's ground plane. No visible changes. Hopefully MarshallH can address these issues with a new firmware.

[Filurkatten]

My issues looks the same as yours aswell unmaker.
Red ring by Marios nose, red distortion at menu but my sync also goes bad after a while.

Got borti4938's adapter with resistors for the wireing, will give it a try this weekend and post my result!

[unmaker]

My issues looks the same as yours aswell unmaker.
Red ring by Marios nose, red distortion at menu but my sync also goes bad after a while.

Got borti4938's adapter with resistors for the wireing, will give it a try this weekend and post my result!

I look forward to your results, GL! Assuming that resolves your issues, that will mean my UltraHDMI issues can be resolved by adding resistors at the RCP-NUS outputs.

[Filurkatten]

So I got the adapter installed and it worked great, but one solder joint wasnt as good as I wanted it so when trying to fix it I acedently created a solderbride so I'm just gonna send my console to a modder to get it done.
I dont want to muck anything up.

But adapterboard seems to have done it.

[unmaker]

So I got the adapter installed and it worked great, but one solder joint wasnt as good as I wanted it so when trying to fix it I acedently created a solderbride so I'm just gonna send my console to a modder to get it done.
I dont want to muck anything up.

But adapterboard seems to have done it.

Good to hear that it worked!

Fixing a solder bridge isn't too difficult. You can use some solder wick or some flux (or both) to remove a bridge:

https://www.youtube.com/watch?v=OaBRak0HnQs

[arebokert]

I would like to add that these visual artifacts seem to be quite common with the N64RGB board from Tim. I have them as well. Some of them were removed by shortening the ribbon cable that came with the kit but all is not gone still.

[NoAffinity]

Gotta question regarding flashing Tim's board with Borti's firmware. I'm sure I'm doing something dumb, but can't seem to load the firmware file into Quartus and actually flash the board.

I solder on some dupont cables on the firmware header on the N64RGB v2.

I connected my USB blaster, have N64 powered and USB Blaster connected to PC USB.

I open Quartus 18.1, and it finds the Altera EPM240

I downloaded the latest firmware for the EPM240, from here, and follow Borti's instructions.

If you only want to flash the firmware, the configuration files are pre-compiled and located in firmware/output_files. You need the POF appropriate for the CPLD you have choosen (look for the _extension).

I open the programmer interface in Quartus, then choose "Add File". When attempting to add the pof file, I get the following error:

Code: Select all

Error (210001): Can't open programming file C:/Users/cglee/Desktop/n64rgb/n64rgbv2_EPM240T100C5.pof -- not valid SOF file

What am I doing wrong?

:edit: nevermind. it had to do with how I was downloadin the file from github. Just finished programming successfully.

:edit edit: So, it seems my N64RGB is a v1. It says N64RGB12 on it, but required the v1 Borti firmware to work properly. When I say work properly, flashing with v2 Borti firmware gave me incorrect colors and some serious noise in the image. I'm guessing all Tim boards are v1 only? Flashing with v1 firmware, all looks right. BUT, in game triggers don't seem to be working. Just to confirm, these firmwares should have IGR, yes?

[borti4938]

N64RGBv1 is my PCB design, which is similar to viletims, but differs a bit.
N64RGBv2 has an improved and sharper RGB output by utilizing a video DAC-IC instead of the R2R ladder.

What you need is the firmware build for viletims board. Obviously, the ‘_igr’- and ‘_sw’-extensions stand for IGR and for (mechanical) switch, respectively.

[NoAffinity]

Thanks for pointing me in the right direction, borti!

I flashed to the viletim -igr firmware, and in-game commands still don't seem to be working. I re-flashed to the v2 firmware, to make sure Quartus wasn't giving me a false positive. It was indeed screwed up again with the v2 firmware, and I then flashed back to the viletim -igr firmware, and still no in-game commands. Once the firmware is flashed, it's simply doing the button combo (z + start + R + c-up/down/left/right)? And 'R' is the right shoulder button?

I tested with ED64 - Mario Kart, Killer Instinct Gold, Fighting Force 64. Also tested with original carts - Mario Kart, Killer Instinct Gold.

I'm sure I'm still missing something...

[bobrocks95]

Thanks for pointing me in the right direction, borti!

I flashed to the viletim -igr firmware, and in-game commands still don't seem to be working. I re-flashed to the v2 firmware, to make sure Quartus wasn't giving me a false positive. It was indeed screwed up again with the v2 firmware, and I then flashed back to the viletim -igr firmware, and still no in-game commands. Once the firmware is flashed, it's simply doing the button combo (z + start + R + c-up/down/left/right)? And 'R' is the right shoulder button?

I tested with ED64 - Mario Kart, Killer Instinct Gold, Fighting Force 64. Also tested with original carts - Mario Kart, Killer Instinct Gold.

I'm sure I'm still missing something...

Do you have the additional wires soldered properly? Double-check their continuity.

[NoAffinity]

Thanks for pointing me in the right direction, borti!

I flashed to the viletim -igr firmware, and in-game commands still don't seem to be working. I re-flashed to the v2 firmware, to make sure Quartus wasn't giving me a false positive. It was indeed screwed up again with the v2 firmware, and I then flashed back to the viletim -igr firmware, and still no in-game commands. Once the firmware is flashed, it's simply doing the button combo (z + start + R + c-up/down/left/right)? And 'R' is the right shoulder button?

I tested with ED64 - Mario Kart, Killer Instinct Gold, Fighting Force 64. Also tested with original carts - Mario Kart, Killer Instinct Gold.

I'm sure I'm still missing something...

Do you have the additional wires soldered properly? Double-check their continuity.

Okay, that's probably what I'm missing. What additional wires?

[borti4938]

One wire for controller and one for the reset. I have to admit that it is a bit hidden in the readme. Read the first sentences of the linked section.

[NoAffinity]

One wire for controller and one for the reset. I have to admit that it is a bit hidden in the readme. Read the first sentences of the linked section.

Thank you again, borti! I read the other readme's probably 10 times each, but clearly overlooked the one that was actually needed...much like I was overlooking the "viletim" folder with the needed firmwares.

will report back once I get the wire in place.

[NoAffinity]

Okay, got the wire to pad A in place, and the IGR is working now. Thanks again for setting me straight, borti.

And sorry to the OP for hijacking the thread, but the title of this thread is a very good one for potentially a wide array of discussions within its scope. If what comes next is better placed elsewhere, someone please point me there.

My interest in the deblur toggling was triggered by a response to a post I made in the OSSC thread. My post was basically asking if the dancing pixels you can see quite prominently at the character select screen of Fighting Force 64, is normal. Notice it doesn't affect everything - character portraits, character models and other things are not affected.

https://www.youtube.com/watch?v=X2EVREA7wy0&t=14s

I have noticed this same effect elsewhere, particularly Killer Instinct Gold backgrounds. Foregrounds, characters, attack effects, are not affected, but the backgrounds consistently are.

It was present when playing on the stock firmware on the viletim board, and persists with borti's firmware. I'm not saying it's anything wrong, I'm just a bit curious about it. The response I got on the OSSC thread suggested that it could be perpetuated by the deblur feature.

So, I now have IGR working w/ borti's firmware, and below is what I've found, with OSSC set to 5x.

Toggling deblur on and off has a slight effect. It is noticeable - pixels, edges, etc getting sharper or softer as expected. It's cool to be able to see a comparison in real time. Deblur is definitely my preference.
Toggling 15-bit mode on and off doesn't have any effect on Fighting Force 64 and Killer Instinct Gold. So, then I tried Mario Kart, and found that toggling 15-bit mode on produces the dancing pixels on backgrounds. Turning it off elminates them, and the backgrounds are perfect. Toggling deblur on and off while in 15-bit mode has no affect on the dancing pixels, and also had no affect on their presence in FF64 of KIG.

So, I was out to try to eliminate the dancing pixels, but found that they can only be turned on, on games that otherwise don't suffer from them (or, better stated, they seem to be a function/result of 15-bit mode).

I'm sure these are very low-level findings but am curious as to the reason it occurs. I'm guessing that FF64 and KIG are natively 15-bit mode games?

[unmaker]

With a lot of help from borti, I was able to make progress with the UltraHDMI + firmware 1.08 + N64RGB (digital to analog type) installed into a single console. I tried various configurations with both Tim's N64RGB board and borti's N64RGBv2 board and the most effective one was as follows:

https://imgur.com/a/xTjwrog

-input resistors on the UltraHDMI replaced with 0 ohm resistors (using these: https://www.digikey.com/product-detail/ ... -ND/417424 )
-UltraHDMI FFC soldered onto borti's FPC adapter
-borti's N64RGBv2 with 47ohm input resistors soldered directly onto the encoder chip (DENC-NUS in this case) and mounted underneath the console


With this configuration all video noise is gone. I tried 10 firmware downgrades from 1.08 to 1.07a and each worked without issue. This requires going from 1.07a back to 1.08 each time so it takes a while. I also tried going into the firmware change prompt screen without proceeding with the actual firmware change. My reason for trying this is that you'll get 1 of 2 screens, either "You are on fw 1.xx, would you like to change to 1.xx?" or "sorry, there was an error please power down now". With this configuration, I got 1 error out of 151 attempts. For comparison, using a previous configuration with Tim's board + borti's FPC adapter + fw 1.08, I got 10 errors out of 10. So this new configuration is a massive improvement. With regards to that 1 error, it happened right after I plugged in my scart cable into the multi out port and I probably budged the hdmi cable out of place in the process.

I could not get it to work so well with Tim's N64RGB board no matter what configuration I tried. When trying different configurations the solder pads on the FPC adapter began coming off so I had to make jumpers to compensate. I thought I made mistakes with the jumpers so I double and tripled checked my connections but still no go. Maybe someone else can make it work but I wasn't able to. I prefer to use borti's N64RGBv2 anyway. It's an improvement over Tim's board and yields a sharper RGB output. I'm now wondering if borti's N64Advanced board will work with this configuration.

It's a bit of a pain replacing the input resistors on the UltraHDMI, especially if you don't have a microscope and just a soldering iron. There's not a lot of space to work with in that area by the FFC ZIF connector. I imagine it wouldn't be so bad if you're handy with a hot air station and that's probably the preferable method. When swapping the input resistors one should be wary about pads coming loose and causing shorts:

https://i.imgur.com/Lzh0vwj.jpg

Despite the massive improvement with 0 ohm resistors on the UltraHDMI, I'm not sure if it's preferable to use ferrite beads instead. I don't have the ability to determine this, as I'm not an EE by any means. I only know how to solder and soldering is just labor work. With borti's help I picked these out but have not tried them yet:

https://www.digikey.com/product-detail/ ... ND/2417265