N8 EverDrive Vs. NESRGB (glitches) *now with video*

[Kyle]

I'm going to let Jason borrow my Famicom Everdrive to test for you guys. I am going to mail it to him tomorrow so hopefully you will know soon if the NES version is the problem or not.

You are good people! Where are you in Texas?

[evil_ash_xero]

Awesome!

[mickcris]

I'm going to let Jason borrow my Famicom Everdrive to test for you guys. I am going to mail it to him tomorrow so hopefully you will know soon if the NES version is the problem or not.

You are good people! Where are you in Texas?

Dallas

[game-tech.us]

Haha. I really don't want to either but probably would if Jason could show it made them go away. I tried to offer him my black Sharp NES controllers lol.

Uhh I never said I didn't want em, just that you were a luck SOB to have them at all!

[game-tech.us]

I also just ruled out the 74HC373 theory, got them right away, badly in fact, different patterns than I had seen before as well. Could be another clue...
One other item I forgot about was my adapter board, it's another thing that is specific to the tests i've been doing. I really don't see it being the problem and if i'm not mistaken some of the other ppl with the snow prolly are using Tim's adapter board and not mine, yes, no, anyone?

[mickcris]

I also just ruled out the 74HC373 theory, got them right away, badly in fact, different patterns than I had seen before as well. Could be another clue...
One other item I forgot about was my adapter board, it's another thing that is specific to the tests i've been doing. I really don't see it being the problem and if i'm not mistaken some of the other ppl with the snow prolly are using Tim's adapter board and not mine, yes, no, anyone?

Have you tried to replace the reset switch yet?

[Kyle]

Once you figure this out those Sharp controllers are yours.

[game-tech.us]

It's more involved than just replacing the switch, the theory is the whole reset circuit needs to be converted to act like the front loader's circuit.

[mickcris]

It's more involved than just replacing the switch, the theory is the whole reset circuit needs to be converted to act like the front loader's circuit.

I was thinking more along the line of a worn switch causing some interference. Could be some component in the reset circuit starts going out and then causes this issue maybe? Just thought it might be something fairly easy to replace to try.

[Pasky]

Why not scope the reset line and see what noise it's introducing (if any) and filter it out? If you got a dual channel scope, scope the one with what you think is the correct reset circuitry, and the culprit, see what's different.

Also, try scoping the video line and the reset switch at the same time. See if any noise spikes correlate to video interference. An additional small ceramic or larger aluminum cap just may be all it needs.

[viletim]

The problem is that the NESRGB is failing to sync correctly to the PPU's digital video output. This syncronisation happens at start up only. Only a power cycle can make it resync. The question isn't what the problem is but why it's happening.

One other item I forgot about was my adapter board, it's another thing that is specific to the tests i've been doing. I really don't see it being the problem and if i'm not mistaken some of the other ppl with the snow prolly are using Tim's adapter board and not mine, yes, no, anyone?

This is significant. The Everdrive may be generating noise (especially right after power on) which is being picked up by the NESRGB. If the adapter board isn't well designed it could contribute to the noise picked up. Could you send me a photo showing the layout of your adapter board?

[Kyle]

Thanks for chiming in Tim.

Jason - Was your adapter board used to substitute the one Tim was selling for top loaders or is this something different entirely?

[game-tech.us]

Yes this is used instead of Tim's as this centers the nesrgb in a us top loader better than the one he designed for the av fami but also mostly works in the US top loader as well.

[Pasky]

Those trace sizes look rather dinky, especially the 5V trace, but I can only judge by that poor rendering, so it's hard to tell.

[Skips]

Those trace sizes look rather dinky, especially the 5V trace, but I can only judge by that poor rendering, so it's hard to tell.

I agree with Pasky. When I was originally creating an adapter for the original Famicom my traces on 5v and ground were too small. It caused it to not sync correctly at startup (different problem from the faulty kit issue I had). Try using the official adapter and see if it fixes the issue.

[game-tech.us]

The 5 volt trace is 40 mils (.040"), which is only like 10mil less than the trace on the nes feeding the power pin. According to circuitcalculator.com the trace only needs to be like 4 mils to supply 200mA. I don't remember the exact draw of the nesrgb so I input 200mA and that may be high. Even if I reduce temp rise, raise ambient, and lengthen the trace it still only says 5.2mil needed. That's sounds crazy low, but I have little experience with trace width calcs. For fun I put in 1Amp, 47.8 mils, same specs, wow...
Because of where I needed the sockets the rest of the traces ended up being 8 mils or they wouldn't route nicely.
The top and bottom are filled with ground plane so that shouldn't be a problem.

[game-tech.us]

Those trace sizes look rather dinky, especially the 5V trace, but I can only judge by that poor rendering, so it's hard to tell.

I agree with Pasky. When I was originally creating an adapter for the original Famicom my traces on 5v and ground were too small. It caused it to not sync correctly at startup (different problem from the faulty kit issue I had). Try using the official adapter and see if it fixes the issue.

Yeah next time I do an av fami i'll put the everdrive on it and see if I can get the snow.
That might rule out or add blame to the adapter board.

[ApolloBoy]

Yes this is used instead of Tim's as this centers the nesrgb in a us top loader better than the one he designed for the av fami but also mostly works in the US top loader as well.

I still don't quite understand why you needed to make a different adapter board just for the top loader. Tim's adapter board does the job just fine and I've never had a single problem with it not fitting inside the system.

[game-tech.us]

It was a bad rendering, this is how it looks without the ground planes showing, higher dpi too so bigger pic.
I made it nearly the same as Tim made his where pins next to each other were routed on the opposite side.
The only difference is the traces go in the opposite direction and may actually have more space between them than the traces on the av fami adapter pcb, 12 mils.

[game-tech.us]

I still don't quite understand why you needed to make a different adapter board just for the top loader. Tim's adapter board does the job just fine and I've never had a single problem with it not fitting inside the system.

I wanted better internal fitment as I showed in one of my vids Tims does fit but with only a few mm's between it and the cart throat.
Maybe mostly it was because I wanted to and was capable of doing it and there's no law against it.
If I find out it is causing these problems i'll do it a different way.

[Skips]

Those trace sizes look rather dinky, especially the 5V trace, but I can only judge by that poor rendering, so it's hard to tell.

I agree with Pasky. When I was originally creating an adapter for the original Famicom my traces on 5v and ground were too small. It caused it to not sync correctly at startup (different problem from the faulty kit issue I had). Try using the official adapter and see if it fixes the issue.

Yeah next time I do an av fami i'll put the everdrive on it and see if I can get the snow.
That might rule out or add blame to the adapter board.

A proper test would be to do it on the system having the problem. Try the official adapter on the system in question. I've used the AV Famicom with the original adapter with both the Famicom and NES versions of the N8 with no issue. Try it on the system giving you the problems so you can 100% rule out your adapter is the cause.

[game-tech.us]

I got an av fami testing now, but I think I may have found the issue and it is in fact on my adapter board.
The clock trace on Tim's pcb is kind of off by itself but on mine it runs right next to two other traces, /INT and /CS!!!
CRAP! I wasn't even thinking about the clock trace needing extra space!
This could be it and I hope so just for the sake of getting this issue figured out.
I think i'll try to bypass that trace on a problem console's adapter pcb and retest.
Here is the trace highlighted in yellow. A redesign of simply sending this trace or the /INT trace to the bottom layer may be all that is needed to fix this issue.

[CkRtech]

Wow. That is a frustrating and yet high-five moment at the same time if that ends up being the problem.

[Pasky]

Welp, that's solved.

[viletim]

The placement of the clock track could make it marginal. To improve the layout I would move the INT# track close to the edge of the board and move the clock away from CS#. It's best to have a solid ground copper fill on the blue layer underneath it.

Other that that I think the adapter board is ok. The video signal (pin 21) is spaced apart from the data bus, which is good.

[evil_ash_xero]

If this fixes it, I'm going to cry with joy.

[viletim]

The 5 volt trace is 40 mils (.040"), which is only like 10mil less than the trace on the nes feeding the power pin. According to circuitcalculator.com the trace only needs to be like 4 mils to supply 200mA. I don't remember the exact draw of the nesrgb so I input 200mA and that may be high. Even if I reduce temp rise, raise ambient, and lengthen the trace it still only says 5.2mil needed. That's sounds crazy low, but I have little experience with trace width calcs. For fun I put in 1Amp, 47.8 mils, same specs, wow...
Because of where I needed the sockets the rest of the traces ended up being 8 mils or they wouldn't route nicely.
The top and bottom are filled with ground plane so that shouldn't be a problem.

This is only for working out the temperature of the copper to prevent the track for burning. The main concern here is the track's inductance which increases with length.

[game-tech.us]

It's best to have a solid ground copper fill on the blue layer underneath it.

It does, I usually put them top and bottom, is that not ok?

[game-tech.us]

FAIL again! I bypassed the trace and just ran a short wire from the nes clock trace to the nesrgb pin 18.
Major snow right off the bat and I can get it to come back with power cycling.
The reset cap is still off the nes-101 but games were loading normally, I was sure without the cap I had to hit the reset button to get them to load, weird..
I guess next i'll just put one of Tim's pcbs in it and if no snow leave it in there and call it fixed.

[eightbitminiboss]

Probably not applicable to this thread's glitching issues. But there is a new OS version out now.

http://krikzz.com/forum/index.php?topic=1877.new#new