I put together a site with all the useless and useful info i've gained about the beloved nintendo top loader over the last several years.
Let me know what you think.
http://nes-101.wikispaces.com/
Nitpick 1: "This is still an original Nintendo so there are no compatibility issues with any games or any accessories" except for one game that requires the lockout chip and Xious's coming adapter for Famicom audio and Famicom expansion controllers. It's not nearly as widespread as on SA-1 and SDD-1 games for Super NES, as only the original (not
reproduction) version of NWC 1990 is known to need a host lockout chip to set up the mapper.
Nitpick 2: A lot of Europe-exclusive games, such as Asterix, don't work on any 60 Hz machine no matter what lockout chip they have or lack. They rely on the fact that the PAL PPU has a vertical blank significantly longer than the 2270 or so cycles of the vertical blank on the Famicom or NTSC NES.
tepples wrote:
Nitpick 2: A lot of Europe-exclusive games, such as Asterix, don't work on any 60 Hz machine no matter what lockout chip they have or lack. They rely on the fact that the PAL PPU has a vertical blank significantly longer than the 2270 or so cycles of the vertical blank on the Famicom or NTSC NES.
Hmm, is there a comprehensive list of what euro games do and do not work on the top loader?
Would you care to share how to add the RCA connectors to the top-loader and avoid any of the vertical lines?
I'm sure you didn't want this post to simply be an advertisement for your services...
Could Asterix be hacked to make it NTSC compatible? I see it uses an unrolled LDA (nn),Y \ STA 2007 \ INY loop to do drawing, and also has blank tiles at the bottom of the screen.
Maybe throw in a DMC IRQ or something to turn off the screen early. Seems like the game needs 18 or 19 extra scanlines to pull it off.
Dwedit wrote:
Maybe throw in a DMC IRQ or something to turn off the screen early. Seems like the game needs 18 or 19 extra scanlines to pull it off.
Do you know how to make a DMC IRQ fire at exactly the same time every frame? Or do you mean use a DMC IRQ to trigger some lines above sprite 0?
So the IRQ trigger time may vary by a couple scanlines. But you can sync to the end of a scanline by reading OAM data, so you know a safe time to turn the screen off.
leonk wrote:
Would you care to share how to add the RCA connectors to the top-loader and avoid any of the vertical lines?
I'm sure you didn't want this post to simply be an advertisement for your services...
There's way more on my website than my mod services.
Heh you're the guy that got this all going, all the answers are in the thread you started in this forum, you have my gratitude!
akaviolence wrote:
leonk wrote:
Would you care to share how to add the RCA connectors to the top-loader and avoid any of the vertical lines?
I'm sure you didn't want this post to simply be an advertisement for your services...
There's way more on my website than my mod services.
I spent some time this afternoon reading your site. No, there really isn't much else on the site you linked in your first post other than advertising your mod services (ex. how to re-brighten plastic and clean cart connectors). I don't have a problem with that, but still. It would be quite cool of you if did a thorough walk-through of the mod you do -- specifically any PCB re-wiring or PCB part replacement -- with regards to addressing the vertical line problem. I think most folks care about solving that problem and not so much about a power switch LED, over-clocking (...), or faux-stereo.
The
thread in question is long and difficult to follow, but there are evidence/statements in there which indicate the RCA mod doesn't solve the problem. If your mod is different than what's discussed in that thread, then maybe combining the two sets of information will clue folks in to where the real problem with the output lies.
Basically what I'm saying is this: I'd love to send you my top-loader and pay for the modification, except there's absolutely no indication in the above thread that the problem is solved with an RCA mod. Hopefully you understand my confusion. :-)
Also, regarding the user/customer who's banging on gongs over heat emissions -- I know nothing about EE, but I've seen this problem when there's all sorts of power-related voodoo going on within parts on a PCB. What I'm getting at: does the "updated" version of the NES top loader (meaning the version Nintendo came out with, not an RCA-modded unit) emit the same amount of heat as the non-updated version? My gut feeling says it probably doesn't.
Thanks!
I didn't say anything from the beginning but it seems other people felt the same way. For now, until I can get information that explain how to mod a nes, it feels more like advertising a service than anything else, especially since it was your first post.
At the least it still related to the nes and not spamming, which is a good thing but more explanation on how to mod it would make more sense since you say that your website contain useful information but that's very subjective.
To be fair, even if he's advertising a service, there is a lot more information there than what someone would normally have up if they really were only trying to sell something. He's narrating his process, which helps provide information. He's making an effort to explain what's going on, but as others have said, it could use more details in certain sections.
My advice is less about the content (I'll let everyone else handle that!), and more about the display-- Make it easier on the eyes! Walls of text are always much trickier to read through. I suggest spaces between each paragraph, and short titles / headers for each section. Something like this:
Quote:
Introduction:The NES-101 or "top loader" was released around 1993 and few were made, making it very collectible and one of the greatest reasons to get one, no one else will have one. They are also superior to the original front loading (toaster) unit as the cartridge connector never wears out and with little cleaning the game loads the first time every time - no more blinking or blowing or frustration!!! This is still an original Nintendo so there are no compatibility issues with any games or any accessories, if it works on a front loader it will work on this unit as well. I have heard many complaints that newer clones like yobo and messiah won’t play every game released for Nintendo, but you get what you pay for...
The top loader was redesigned to be a cheap console for people to buy instead of the new and relatively expensive super Nintendo. I've heard Nintendo did some research and thought not many people were using the composite output on the original front loader so didn't bother putting it on the new top loader. Maybe they were right, but that’s just not the case these days and that’s why I mod them for composite output.
The Composite Video Mod:About the video output upgrade I do: From the factory the top loader only had an r/f hookup and its video quality is poor, that is why I mod these units with rca jack type composite outputs. You will never worry about the r/f switch going bad and being able to replace it, and history already speaks for itself! I have been modding these units since 2003!
I've had many people ask if the lines on the tv screen, common with the top loaders, are improved with this mod and YES they are nonexistent Also, the overall picture quality is as near to a front loaders video output as it gets, they're hard to distinguish in side by side comparison. The composite output is better than the original r/f as it doesn't have to go through rf modulation or a crappy rf switch, just straight from the top loader's source video output to the tv input or whatever input you hook it up to!
The r/f output no longer works due to this mod, but believe me you won't care! Outputs are color coded rca jacks for easy hookup. Audio right and left are wired mono since the machine is not capable of true stereo without another serious mod (read below), but both were installed for easy hook up to a home theater.
I've also found an alternate method of audio/video output, instead of 3 rca jacks, I found a headphone style connector that will do the same job and fit into the r/f jack hole. This means the shell will not have to have the sticker removed and 3 holes drilled in it! Plus you will not be able to tell right away that it’s even been modded. It does require a special cord, but I will supply it if this is the kind of mod you desire.
etc
Also just make sure the sections are well organized overall. Your pricing is bolded somewhere in the middle, which seems like a very odd spot for it. Perhaps you should have a separate section of the site which goes into more detail on services and contact.
You've got the foundations of a handy resource, keep it up!
koitsu wrote:
I spent some time this afternoon reading your site. No, there really isn't much else on the site you linked in your first post other than advertising your mod services (ex. how to re-brighten plastic and clean cart connectors).
Less than 1/3 of the content is directly related to my mod services, I did a word count.
koitsu wrote:
The
thread in question is long and difficult to follow, but there are evidence/statements in there which indicate the RCA mod doesn't solve the problem. .... Hopefully you understand my confusion.
Yes its a long thread, but the info is there and I dont care to repeat it and act like I had anything to do with it other than testing it on my own and finding out what works.
koitsu wrote:
What I'm getting at: does the "updated" version of the NES top loader (meaning the version Nintendo came out with, not an RCA-modded unit) emit the same amount of heat as the non-updated version? My gut feeling says it probably doesn't.
You lost me somewhat, do you mean the redesigned nes-101 compared to an original nes-101? If thats what you mean its irrelevant since no one can get a hold of one. Also i'm just now starting to research heat inside the top loader, hence the whole section at the bottom of my page.
Agreed, the whole site (page?) is filled with pitches for the products/services, but it's also filled with lots of interesting information, mainly an apparently large sampling of NES units and the variations between them. Those heat pictures were awesome! I've recently been trying various wall warts for my consoles and network gear, trying to decide which is the best. I'd been thinking about this very thing, being able to see exactly what is heating up with various configurations; too bad such a camera is probably quite expensive.
I second requests for information on the mods. I only think it could improve business, at least for people who are skeptical about the build quality. Some pictures of the internal mod work would go a long way. I think it's great that these kinds of modifications are in enough demand that it's a marketable thing.
From my own personal experience of the RCA mod, it does improve the video quality, but it's nowhere near the quality of the front loader.
If the front loader is the goal, then here's how I see it:
- nes-101 no mod - 75% video quality (due to vertical bars)
- nes-101 with mod - 92% (vertical bars very difficult to see, except if you look for them, like on games like SMB1 or 3 in starting screen where everything is bright). But the mod does have a flaw, it's about 1-2 levels darker than front loader in brightness.
In comparison, I also recently got the FC Twin. It's as bright as the front loader, but it too has terrible vertical bars.
When I saw your web site stating that there are no vertical bars, that's why I was under the impression that you got a new circuit board design. Also, the original thread I started required you to remove and reuse some parts from a front loader (i.e. rebuilding the front loader circuit board on PCB). But it seams that you didn't do that (300+ mods suggests this).
Lastly, I assume you have a rare Nintendo RCA modded system, or one with no vertical bars (due to the video you produced, and images). The thing that I found interesting was the redone bridge rectifier area. It seems that Nintendo switched from a single bridge rectifier IC to using diodes to reproduce one. Can you confirm? Can you produce high quality images of that section? This might be the source of all the vertical bars.
leonk wrote:
From my own personal experience of the RCA mod, it does improve the video quality, but it's nowhere near the quality of the front loader.
If the front loader is the goal, then here's how I see it:
- nes-101 no mod - 75% video quality (due to vertical bars)
- nes-101 with mod - 92% (vertical bars very difficult to see, except if you look for them, like on games like SMB1 or 3 in starting screen where everything is bright). But the mod does have a flaw, it's about 1-2 levels darker than front loader in brightness.
Not been my experience, I actually had some great video captured, but I couldn't get it to load with that crap youtube editor, i'll get back to it and get it posted. I'm no expert visual comparator, but I couldn't see any lines and minimal brightness difference.
leonk wrote:
In comparison, I also recently got the FC Twin. It's as bright as the front loader, but it too has terrible vertical bars.
Never had one, but good to know.
leonk wrote:
When I saw your web site stating that there are no vertical bars, that's why I was under the impression that you got a new circuit board design. Also, the original thread I started required you to remove and reuse some parts from a front loader (i.e. rebuilding the front loader circuit board on PCB). But it seams that you didn't do that (300+ mods suggests this).
If you read the thread closely the parts are available in the top loader, but I saw no difference with or without the choke (fc2). The transistor was easy to cross reference and find a cheap replacement.
leonk wrote:
Lastly, I assume you have a rare Nintendo RCA modded system, or one with no vertical bars (due to the video you produced, and images). The thing that I found interesting was the redone bridge rectifier area. It seems that Nintendo switched from a single bridge rectifier IC to using diodes to reproduce one. Can you confirm? Can you produce high quality images of that section? This might be the source of all the vertical bars.
The one I have is not a rca or a/v version, but it is a redesigned board just with r/f out. Its in the video description. Yes the power circuits are different, but I still whole heartedly believe its all about the circuit board design running traces in the wrong places picking up interference, as stated in the thread by I forget who. After trying all the variations, like cutting traces close to the ppu (didn't work), lifting the pin worked, it worked well.
The thermal camera heat research portion has been updated and moved to its own page.
Lots of new pics with the shell removed.
http://nes-101.wikispaces.com/Heat+Research
Just added thermal camera results from over-clocking the top loader by 32%.
Check the bottom of the heat research page.
Is there any way to actually verify the temperatures? It's hard to believe that a device can show temperature like that. It's really cool.
I've lately been changing power supplies out on various things, thinking about heat buildup. Most of the old game systems use linear regulators, so the higher the input voltage the more heat it dissipates. So I've been trying lower-voltage power supplies. One problem is that the rectifier for AC adds a voltage drop, and hence heat, that you don't need if you run it off DC. If you can get the loaded transformer output as close to around 7V as possible, you should have the least heat. Of course I run my NES with the top cover and RF shielding removed, so it has a lot of air flow. I suppose your pictures can tell us really how much heat the rectifier and regulator are giving off, as compared to the other components, to get an idea of how much you could remove if you somehow eliminated the former entirely.
The thermal cam is a Fluke and even though its a few years old I still had to pay almost 4 grand for it! So anyway, i'm pretty sure its accurate.
May dig out a front loader and check it out next, but i'm in no rush and have other stuff to get done first.
Just added a TON more pics to the main page.
You could have used PNG for the image comparison shots, the JPG artifacts make both images look like complete crap.
tokumaru wrote:
You could have used PNG for the image comparison shots, the JPG artifacts make both images look like complete crap.
I do believe the originals are in fact png, but I guess my habit is to save pics as jpg after alterations. I'll go back and try to make a better one because I agree the end pic looked crappy on both sides.
I just noticed that you can view the images larger if you use the browser's "view image" popup.
As a test of the device itself, if you just point it at the same objects when they've been off for hours, does it show a uniform temperature on all surfaces and materials? That'd pretty much confirm that it really is measuring heat and not the surface reflectivity or something incidental. It's still voodoo to me how this thing can measure temperature like that. I find it incredible, thus these attempts to make it credible.
blargg, look up "IR thermometer" on Google and be amazed. :-)
OK, another great new addition to my site!
This one is called "power research" and is actually full of applicable info to many other consoles and electronics.
http://nes-101.wikispaces.com/Power+Research
Excellent! It's like you read my mind for that page.
Quote:
I am not knowledgeable enough to how to fix this yet, but maybe someone will step up. The issue is now that I modified the power circuit, but still have the same 5.5mm x 2.1mm power port, it could easily be fried if someone hooked it up to an original NES a/c adapter! Origianlly you could hook up a cheap adapter that outputs DC to the NES-101 and it wouldn't hurt anything, but with this mod anything BUT 5vdc will damage something.
Yes, this was the problem I came to when I was considering using a 5VDC adaptor for a console-related thing, and removing its 7805 regulator. With the 7805 and a diode on input, the thing can withstand almost anything short of being hooked up to 120V via tha DC power connector. Without it, it's very susceptible to the wrong adaptor. My first Zip drive used a 5V power brick, and one time I accidentally hooked up a 9V one or something. Fried the drive, melted inside. You could probably put a voltage crowbar in there, which shorts to ground if the input exceeds the specified voltage.
It seems the best would be to put a
buck-type switching regulator, like much networking gear has these days. They can accept something like 3V to 30V, and put out very little heat, since they very their current draw based on voltage. I wonder whether someone makes a small switching board with three terminals, for direct replacement of 78xx regulators. It'd have the filter capacitor, switching control IC, pass diode, and inductor all on a small board, with just the in, ground, and out terminals coming out the bottom.
And like you mention, I imagine weight is one reason they've changed to switcher power bricks for devices these days. Shipping cost is significant, and the old-style transformers added significant weight compared to the device itself.
blargg wrote:
I wonder whether someone makes a small switching board with three terminals, for direct replacement of 78xx regulators.
You're not going to believe this! I just got a reply on another forum and he directed me to an
ebay auction with just what we're looking for - a direct 7805 switching regulator replacement! Excellent! I contacted the seller and asked him to check out the webpage on power research and see what he thought. Maybe he can help even further, he also may know how to integrate it better than just simply replacing the 7805, like remove this component and/or add this one...
Very cool. If you put one of these in a NES, for example, you could then power it from from something like 3V to 40V. When powered by the original NES adaptor, it'd draw considerably less current, and thus generate less heat in the power supply as well as very little in the NES. You shouldn't need to remove any components, as these really are drop-in replacements. Leaving the rectifier in the NES, for example, allows AC or DC of either polarity to work. The rectifier probably doesn't generate much heat either. There's nothing keeping you from using a switching wall adaptor as well, further reducing power usage and shipping weight.
I found a couple more:
DE-SW050 7803SR
I wonder if it'd be possible to power an NES off a Wii's USB port the way I can with a DS. How many amps does the NES draw?
koitsu wrote:
500A.
I hope you mean milliamps.
blargg wrote:
When powered by the original NES adaptor, it'd draw considerably less current, and thus generate less heat in the power supply
I'd have to see it with the thermal cam and Kill A Watt to believe it, but that would be nice.
blargg wrote:
The rectifier probably doesn't generate much heat either.
Didn't you see the thermal cam images of the rectifier? It was the hottest part in the power supply circuit!
I'm also investigating higher efficiency rectifiers, less heat (wasted power).
blargg wrote:
There's nothing keeping you from using a switching wall adaptor as well, further reducing power usage and shipping weight.
Probably true, though it may have to be of high enough output, there's a 1.1V drop accross the rectifier, but i'm not sure if thats a/c only?
blargg wrote:
I found a couple more:
DE-SW050 7803SR
Except the digikey one isn't in stock and the other one costs more than the ebay one...
That seller hasn't replied yet, but its possible (now that I see that one from digikey) that he didn't make them and has no answers for us...
...Wait, 5V and 500mA? I could run it off a USB port?
Hmm, I just found the 7805SR-C on mouser after searching for the part number AND its cheapest i've found yet, $9.90 for one. Will get it ordered and tried out asap.
lidnariq wrote:
...Wait, 5V and 500mA? I could run it off a USB port?
Amp meter said the top loader was using .39 A, with the switching wall adapter, so maybe if the usb will output this much power then its possible. You may have to take out and jumper accross components like I did in order to avoid losses though.
akaviolence wrote:
blargg wrote:
When powered by the original NES adaptor, it'd draw considerably less current, and thus generate less heat in the power supply
I'd have to see it with the thermal cam and Kill A Watt to believe it, but that would be nice.
I tried running my cable modem and router off 12V and 5V (they both use switching convertors), and the lower voltage had around twice the current draw, as measured by a multimeter. I even did the same with a tiny LED flashlight module which also uses a buck-type converter, and it was the same. These converters are quite impressive, around 90% or more efficiency. Thermal pictures of the difference would be good. BTW, when comparing two schemes, try to take the thermal pictures from the same angle, so that the only difference is the heat.
akaviolence wrote:
blargg wrote:
The rectifier probably doesn't generate much heat either.
Didn't you see the thermal cam images of the rectifier? It was the hottest part in the power supply circuit!
Ahhh. Then it would make sense to eliminate it, unless you want to preserve the polarity-independence it provides. You can still get protection from reverse polarity with a single diode. You could put it in series, which would cause reverse polarity to be ignored. This would still dissipate some heat due to the voltage drop across the diode. You could put it in parallel, with a fuse in series before the diode, so that it crowbars the circuit if polarity is reversed, and blows the fuse. This wouldn't dissipate any power with proper polarity.
akaviolence wrote:
blargg wrote:
There's nothing keeping you from using a switching wall adaptor as well, further reducing power usage and shipping weight.
Probably true, though it may have to be of high enough output, there's a 1.1V drop accross the rectifier, but i'm not sure if thats a/c only?
The drop is always there, since there are always two diodes in series, giving double the diode drop. If you do leave the rectifier in, you should get less heat if you use a higher-voltage adaptor, since there will be less current used by the switching regulator. The heat dissipation based on the product of the voltage drop and the current. The voltage drop is fixed (0.6V for silicon diodes). Given that the NES adaptor puts out at least 10 or 11 volts when not fully loaded, it would serve well with a switching regulator. Shipping say a 12V switching wall supply (as opposed to a 9V one) with a NES would be confusing if the person ever tried to use it with an unmodified NES.
lidnariq wrote:
...Wait, 5V and 500mA? I could run it off a USB port?
I haven't done much with USB yet, but I believe you get much less than that normally, but there a command to send to the USB host to tell it the current required. But since USB has sort of turned into a power plug, I'd bet that a lot of stuff out there might ignore that anyways.
blargg wrote:
BTW, when comparing two schemes, try to take the thermal pictures from the same angle, so that the only difference is the heat.
As long as I avoid sensing reflected heat (noticeable on shiny stuff like screw heads and the exposed ground around the outside of the pcb) then it won't matter what angle I take the picture at, it will still give me correct readings. Unless you just want it to be aesthetically pleasing?
akaviolence wrote:
blargg wrote:
The rectifier probably doesn't generate much heat either.
Didn't you see the thermal cam images of the rectifier? It was the hottest part in the power supply circuit!
blargg wrote:
Ahhh. Then it would make sense to eliminate it, unless you want to preserve the polarity-independence it provides.
Shipping say a 12V switching wall supply (as opposed to a 9V one) with a NES would be confusing if the person ever tried to use it with an unmodified NES.
Here is the real issue with taking out the rectifier, some one will try to plug an original 9V NES a/c adapter into it and it will burn...
Memblers wrote:
But since USB has sort of turned into a power plug, I'd bet that a lot of stuff out there might ignore that anyways.
And that's even official since USBIF adopted the
Battery Charging Spec. A port can pull up to 1500 mA before trying to communicate.
This is really great stuff! I'd like to see information on how to replace the AC port with a USB one for charging. One can buy a USB charging brick (e.g. one for iPod) for close to nothing, and use that to power the top loader.
Also, has anyone looked into getting a CopyNES installed in a toploader?
akaviolence wrote:
blargg wrote:
BTW, when comparing two schemes, try to take the thermal pictures from the same angle, so that the only difference is the heat.
As long as I avoid sensing reflected heat (noticeable on shiny stuff like screw heads and the exposed ground around the outside of the pcb) then it won't matter what angle I take the picture at, it will still give me correct readings.
It's easier to compare when they're from the same angle, and it eliminates the possibility of error due to different angles. It may work fine, but it introduces uncertainty, because one should never fully trust his tools, or trust them more than necessary. When doing science, it's best to control everything not being tested.
blargg wrote:
As tepples has posted elsewhere, the standard NES composite video signal has enough information to fully recover the palette indicies of every pixel, which is enough to create an RGB signal
@ Blargg or Tepples, found this post elsewhere, I hate when I see a post with a snippet of info that blows my mind and I can't find anything else.
How far along did this thought get?
Previous topic |
Other previous topic
One would need a PLL to sync to the color subcarrier, clocking a 42.95 MHz ADC. The maximum and minimum value of the composite signal during each pixel period would give the upper 2 bits; the location of the transitions between minimum and maximum would give the lower 4 bits. But ADCs this fast are expensive, even if you do need only 8 distinct levels. It might be possible to use a slower DAC with more gray levels if Viterbi-type techniques are used.
tepples wrote:
One would need a PLL to sync to the color subcarrier, clocking a 42.95 MHz ADC. The maximum and minimum value of the composite signal during each pixel period would give the upper 2 bits; the location of the transitions between minimum and maximum would give the lower 4 bits. But ADCs this fast are expensive, even if you do need only 8 distinct levels. It might be possible to use a slower DAC with more gray levels if Viterbi-type techniques are used.
Do you have a diagram etc? For proof of concept I dont really care about cost, I just want to try it!
So far it's mostly theory. Going by the
NTSC video page, we see the color phases:
Code:
111111------
22222------2
3333------33
444------444
55------5555
6------66666
------777777
-----888888-
----999999--
---AAAAAA---
--BBBBBB----
-CCCCCC-----
PPPPPPPP <-- pixel width
Each pixel lasts for 8 units of this. As long as a pixel contains at least one transition, the phase can be recovered. It seems so, since each cycle lasts 12 units, thus there are transitions every 6 units. Thus, an 8-unit pixel must contain at least one. If it's gray, of course, it contains zero transitions, and you would detect that as well. I think you could even recover the tint bits, since you'd always get at least one transition of that waveform as well.
If anyone wanted to seriously make a device to do this (it would literally just connect to the RCA video out, nothing more), we'd need to first write a simulator that generated the composite waveform, then ensured that everything could be recovered in all cases. This would be pretty straight-forward. It's just that the hardware to implement it would be costly, and unless someone stepped up to purchase it AND had the expertise to build it and write the code for the microcontroller, it's going to remain just talked about here.
tepples wrote:
But ADCs this fast are expensive,
The analog devices AD9283 and AD9057 are $2.60 in quantity (looks like $8 in one-offs) and can sample at 50MHz and higher.
For output, 6MHz DACs are almost free.
blargg wrote:
I think you could even recover the tint bits, since you'd always get at least one transition of that waveform as well.
I sat down and wrote the state machine at some point, and found that there's barely not enough information for that for all combinations of color and tint. But you could assume they didn't change tint from the previous pixel and get a correct value almost all of the time.
Quote:
write the code for the microcontroller
Being able to deal with 43Msample/sec is faster than any micros can deal, so we're into FPGA area now, further increasing the cost.
One other snag about the tint bits is that they aren't even synchronous; they can change
between pixels (yes, this means the NES really has more than 256 pixels across the screen). This is more noticeable on PAL, where a CPU cycle isn't an integral number of pixels wide. Here's a series of video captures from PAL where the tint write moves fractions of a pixel (1/5) each time. The three lines on top are done using normal pixels, as a reference of where pixels fall. Note how the tint-drawn line moves very smoothly horizontally, at fractions of a pixel. Presumably the tint write directly sets a latch that drives a resistive adder for the tints. Makes sense.
Multiple updates:
The heat research page has been almost totally redone with better initial tests, drop in replacement switching regulator tests, and cpu and ppu heat sink tests.
The power research page also has additional info regarding the switching regulator and switching wall cubes, plus better initial comparison of power consumption of different wall adapters.
Modding has its own page now, with a few teasers added, but mostly the same stuff.
I remember hearing that the top loader is missing the 21.47727 MHz cart connector pin? Can anybody confirm that?
E: Okay, so it's like that in top loader and Famicom. I was wondering how can the reset timing (for saving) work in PowerPak without it but I forgot it has a 21MHzish oscillator on-board.
So nothing to see here.
So today I got a modded NES 101 from akaviolence, but it doesn't work. I contacted him an hour ago, but so far no reply.
Maybe you guys can help me out.
I'm using an NES 001 power supply. This system has been modded with AV, stereo (either amplified mono or simulated stereo by dividing the sound channels) and has an overclock crystal installed. The pot for the stereo/mono mix is where the RF input was and the clock speed switch is where the channel 3-4 was.
When I turn the system on, I get a black screen with very very faint gray outline of what should appear on the far right of the screen. Additionally, audio is extremely distorted and garbled. The problem occurs at either clockspeed or stereo setting.
When the composite yellow video plug is not inserted into the system, the system won't output sound, meaning the video line carries at least part of the audio signal. Not sure if this significant or by design.
I only have a vague knowledge of the NES 2 and this modded system is the first I've ever owned. Nevertheless I have the means to take out the security screws and take some high res pictures of the PCB.
Maybe doing so will allow you guys to understand what is wrong and address some earlier questions in this thread about just how akaviolence does this remarkable mod.
Even though I payed for a system that arrives not working, I don't think that it's his fault. Probably just something came loose during shipping, or so I hope.
So anyway, anyone have any theories for what could be the problem? I'm not just getting a bad picture, I'm getting almost no picture and extremely distorted audio, like what would happen if you crossed the audio and video jacks on an NES toaster. That high pitched whine.
I'll take some pics if it helps. What should I take pics of specifically?
Just took a video of the problem for you guys to see, as its difficult to explain in words.
http://www.youtube.com/watch?v=FuZBHCvDhTI
Sounds like ground is disconnected -- I can hear the 60Hz from the video signal in the audio. Also, the video looks like it's some kind of highpassed with the sharp edges showing up, which can also indicate a disconnected ground.