RAREWARE N64DEV V3 - N64 Development Cartridge

Hi all...

I've also been looking into this particular "Rareware" N64 Development cartridge.
Entitled the "N64DEV V3".

From what I've read so far... this cart is pretty much an impossible attempt to get going. Mainly due to the non existent software/Drivers for it...and also all the security features RARE had implemented for it to even just boot up.

For starters, the CIC dongle adapters that connect to the top DE9 connector "P5", and also the custom Flash boards that are required... which are storing extra validation data. The Flash boards simply plug into the Top cartridge slot on the back side.... again all of this was simply put in place to facilitate booting of the cartridge. MAN!.......


















Now.........
I have also just recently come into possession of one of these Dev. cartridges.
However, the one I have has been damaged, so I need to repair it. Unfortunately I don't have the special Flash board, or the CIC dongles. Which sucks....

I'm planning on at least attempting some interfacing and just seeing If I can do anything with this cart at all. If I can get it powered up, just doing some logic scans using the Logic Analyzer will be interesting.......if nothing else.


Heck, I'll be happy to even just backup some of the Flash content using the JTAG interface. I'm also planning on backing up the Altera configuration device using my Altera Master Programming Unit if need be.
Preservation if anything right?

Perhaps in the future something will be found...like a Windows driver or possibly some application software.

Anyhow,
I'm hoping someone here also has this cartridge in their collection, as I am hoping to get some High-resolution photos of the Top side of the PCBA. I'm trying to identify the Inductor "L1" part number. The inductor on my board has broken of.... and I wish to replace it with the exact part number if I can identify it. I want to make sure I get the right one before power up the cartridge to begin testing. Most of the photos online are not clear enough to identify if there is a part number visible.

The inductor is the circular component on the top left side of the board, directly under the Red GND3 probe test point.... and is also right beside "D1". (Indicated in the image below)



If anyone here can help me out that would be Great!
I'm actually also hoping to get hi-res photos of the Dongle PCBA's...to possibly figure out the simple Pinout connections from the DE9 connector to the CIC chip.

Anyhow,
Let me know.

Cheers! Folks,
-Gerry

 

Now, for those of you that are hardware designers.......(Marshallh....ahemm.....)....

I'm hoping that a good best guess could be postulated using the other onboard components that are in the circuit, and are all clearly identifiable. So now, inductors are usually part of Power regulation circuits, which I believe is the role of "L1" in this particular circuit.

So I have identified and collected datasheets on almost all of the components on this board. At least the primary ones. Most of the passive components I have not bothered with, except for those around the Inductor.

So here is what I have derived as the Power regulator circuit on the board which the Inductor is part of. :
Datasheet links for the active components involved are provided below. (Passive....who cares....lol )

Active Components....
U2 = MAX770 - Step-up DC/DC controller
http://www.digital-circuitry.com/FI...N64DEV/Datasheets/max770-max773_DATASHEET.pdf

Q1 = 2N02L - Power MOSFET
http://www.digital-circuitry.com/FI...ARE_N64DEV/Datasheets/motorola-mmft2n02el.pdf

Passive Components....
C9 = 220uF - Capacitor
C17 = 330uF capacitor
D1 = Diode -?

Trace #1: GND-------->C9 220uFCAP-------->3.3Vcc
I'm assuming this is a decoupling capacitor for noise on the main 3.3v supply rail.

Simple continuity checks show the 3.3Vcc supply passes through the Inductor L1 as follows:
Trace #2: 3.3Vcc------->L1-------->D1-------->C17(330uF)--------->2N02L Power MOSFET (Source Pin).

So now, my plan is to start looking at some of the datashseets and try to make a best guess at the inductors value. If there are any of you here that have Analog circuit design experience.... that could toss in your input on this.......that would be really great.

I'm also thinking that once I get the unit to power up successfully, the onboard Altera configuration chip will load the bootloader into the FPGA if anything......in turn if I connect the SCSI interface correctly to the PC, I should be able to get Windows to detect the hardware and this will identify the driver name for the unit....which we could then begin to search for online. This is my plan of attack anyhow.

So..........moving forward!

Cheers! Folks,
-Gerry
http://www.digital-circuitry.com

 

Here is a rough schematic.....forgive the crudity of this drawing....I'm in a coffee shop and not at my PC with my OrCAD design software. So I had no other option but to slap it together in Paint.... (*@#(*@!!! Uuuugghhrrrr!!

Kinda hurts a bit to post it..
Edit: Just made a quick correction to the schematic....had the Drain connected to the wrong side of the diode... corrected.



Cheers!
-Gerry

 

Another thing that has come to mind regarding these cartridges and their Cases being used....

Let's keep in mind that these cartridges are CMOS RAM based Dev. cartridges. So no ROM data remains after the power is removed from the cart. The onboard FLASH memory chip on this cart is only 4MBytes in size...therefore I'm thinking this is only used for storing the SCSI interface Firmware. All of this is similar to how the IS-VIEWER64 functions.

So, In regards to RAREWARE Cartridges like these that have Game titles written on a Sticker or Decal....
It could very well have been that RARE only provided developers with the DEV Cartridge PCBA only, and users would have to rig up their own enclosure for it. This would explain how many of the existing images of these Dev carts only show the PCBA itself...or show the boards inside of cartridge cases that have been modified by hand.
As shown below:

You can see how the holes for the JTAG IDE headers have been cut open with a knife...not exactly a pro cutting Job.





Also shown here in this photo below...
You can see the top of the cart has been cut and is not exactly a straight edge...sort of a rough cutting Job as well.
(Especially near the center.)



So with that in mind....
it is likely that any sticker with a Game Title designation on the Cartridge case.... it has nothing to do with the RAREWARE Cart at all.
It is just an old sticker from when the case was used with a Nintendo Flash cart. Again...these are RAM based Dev carts.

Anyhow,
This is not fact by any means....just my speculation. Hopefully someday someone who actually used these carts for development can come forward with some info.

Cheers!
-Gerry

 

So, I've been gong over the datasheets a little more tonight, also bounced the schematic around to a few contacts to get second, third, fourth opinions.

So after bouncing opinions around for 30 minutes....I'm basing a "best guess" on the fact that a "MAX770" DC/DC controller is in the circuit. The datasheet for this component U2 = MAX770 - Step-up DC/DC controller, makes reference to "standard operating circuits for the MAX770" when using an N-Channel Power MOSFET...
(in our case the Q1 = 2N02L)..... it indicates that a 22uH inductor is normally used. Also, inductors with ferrite cores are recommended.

Shown below is an excerpt from PG.16 of the component "U2" the MAX770 - Step-up DC/DC controller Datasheet:

(Notice the similar circuit with the N-channel MOSFET.)



So I think we are safe to say that component "L1" is a 22uH Ferrite core inductor.

I'll be repairing this this week, and will continue to post my progress as I move forward.

Cheers! Folks,
-Gerry

 

@Gerry_MAN, you really need to PM me when you need something as I don't scour these forums all day everyday!

I'll try and get you some nice crisp high res photos of the part in question when I get home from work.

 

@level42
Great Thanks!

I have a feeling that there may not be any written external markings on the inductor.
It could be that yours will, however I'm thinking not.

Regardless, based on the MAX770 datasheet I'm pretty confident it's a 22uH.
However, some extra shots won't hurt to see.

-Gerry

 

https://imgur.com/a/xAIfv

Let me know if these help

 

Amazing!
Thanks level42....

Your Camera has super amazing resolution.
Good to have those on file.

Cheers!
-Gerry

 

I have one of these but I don't have a flash board or cic dongle. Would be really neat to have them just as collectibles but I'm not getting my hopes up...

 

I really hope at some point I'll be able to get this little guy going.

Yeah well....eBay is your friend.
The odd time youI'll see Dev gear pop up on Yahoo Japan...so that's always something to keep an eye on.

-Gerry

 

I wish I knew how to order from Yahoo.jp, lately I've seen cooler stuff there, then in North America.

 

It *should* not be such a big effort to remove a CIC from a kart (Say Mario64?) and solder it on a dedicated board. The inputs can probably be traced back from the pictures / board?

 

@kammedo,

Yeah, I thought the same thing.....
It does look like the CIC's DATA I/O Pins.... CIC14 and CIC15, their traces are fed through to Pins 2 & 3 on the DE9 connector.

Vcc and GND are also shown to connect to DE9 Pin#1 (Vcc) and Pin#5 (GND).
The CLK and RESET are the last two on the opposite side of the board and that are not shown.

I guess it's easy enough to check the traces for the CLK line with a scope.
Getting an actual dongle PCBA for the sake of being consistent would be nice.... and just clone it.
However, you're absolutely right. You can probably trace it out fairly easily.

I guess I was just being lazy.... lol ^_^
Cheers!
-Gerry

 

Ok I've taken apart my dongle, desoldered the 6105 and taken pics. I have gone ahead and written the pin mapping if you're curious. All you need is a male DB9 and a 6105. Take note of pin 1 which is square. Bypass cap isn't fitted but it certainly woudn't hurt.





DB9 pin 1: 3v3
DB9 pin 2: P2.1
DB9 pin 3: /COLDRESET
DB9 pin 4: NC
DB9 pin 5: GND
DB9 pin 6: P2.0
DB9 pin 7: CLK (~1.6mhz)
DB9 pin 8: NC
DB9 pin 9: NC

However, the dongle is not necessary, because it's electrically identical to the CIC socket that is already inside the cart board anyway.

Some other details on the Rareware cart that might not be explained anywhere:

• EEPROM socket supports 4k/16k like usual
• Cart connector on top seems to be used only for the FlashRAM plugin boards. This is simply to allow them to debug saving onto the 128k flash (Used by Jet Force Gemini, etc).
• SCSI Micro DB68 port includes an active terminator onboard. Seems to work fine with a AHA-2940 board on Win32 that also works with the SN64 cart.
• Motorola 68340 is present, which is a 68020 core with DMA and some other stuff to make it more useful. It definitely will be talking to the SCSI interface chip and marshalling transfers with the fpga.
• 32mbits NOR flash is present and probably holds firmware for the onboard 68340. It could also contain a N64 upload stub, I haven't dumped it yet.
• 256kbyte DRAM is present and this is definitely the work ram for the 68k.
• 4Kx16 dual port ram is present, probably used as a FIFO between the 68k and/or SCSI, FPGA.
• Power is supplied by the N64, though it seems possible that it could be actually powered from SCSI termination supply, I'm not sure.
• Lower 2x5 header is JTAG for the fpga, upper port is probably 68k jtag.
• 64mbyte DRAM is populated, but the board has footprints for up to 128mbyte. The N64 would work fine with 128mbyte roms but they would've been hugely expensive to produce.

In any case I haven't been able to get it to do anything on N64, though I recall it did identfiy as "Rare N64 debug" or something when plugged into a PC. Of course, useless without drivers and software. If someone was to disassemble the 68kb binary that presumably exists in flash it may be possible to repurpose it.

As far as the Rare dev workflow, it seems it'd be something like:
1. Load bootstrap from PC into cart
2. Boot N64 to a stub
3. Load debug binary and start it
4. Running binary fetches assets at runtime
5. Running binary inevitably crashes and spews debug log to PC, rinse and repeat

If you've looked at the Conker's debug build you'll see that is designed to pull down game assets on demand, so a full rom build is unnecessary. Remember that packing all the assets via LZ takes time (definitely more so back in 1998) and so it wasn't practical to compress and link everything for each debug cycle.
The conker's build (along with Perfect Dark debug) checked a memory mapped area in cart space for the existence of a magic word. This is why the builds when they were originally released were not runnable by any emulator or hardware until I patched them. My theory is that these were designed to be loaded onto the dev cart and it would talk to the cart as it booted, checking for newer copies of assets on the host PC.
Whether the Rare cart dongles that were supposedly able to let these dumps run did so for copy protection, I'm not sure. I think it was just a way to make builds start without needing to talk to the dev cart. In any case you can read the write up I did on those in the release nfo.

 

Holy Sweet Je*su5 Marshall !!!

This is some Crazy in depth info you've provided here. Right On!
Amazing.... Thanks for doing all that leg work.

- Conker's build debug (along with Perfect Dark debug)
I'll definitely be going and checking those out and your write ups on them.

Thanks again for the Photos as well.
Sweet.

Cheers!
-Gerry