JTAG a Buffalo WZR-HP-G300NH2?

[LightworkerNaven]

The reason I ask you to focus on CPU ID is you get excited when you see DEBUG ON when your CPU ID is ffffffff.
If you get CPU ID 0 and even you do get DEBUG ON it does not mean anything. It will not help you get detect CPU ID nor read data properly.

I agree with you there might be missing trace but there is very little interest of this router and spend time to find out the missing trace is not what I would do if there is only one interest.

However if you do get CPU ID I would be happy to create the XML for you. (Very easy and not rocket science if you open the WRT160NL.xml). It is just a few lines of memory definition and possibly some init strings that can ge added basked on the information from Google.

The reason I guess that this router should work out of the box is
1. It has proper JTAG port populated.
2. It uses dd-wrt as its stock firmware. We all know dd-wrt is for hackers and hackers will mess around with it. So the JTAG port should be available for people to play.

This is a guess and I could be wrong but it is a reasonable guess.
I need several other people who show the interest before I invest the money on this router.
Even I do get the router and the JTAG port is not accessible, I might not be able to do much as there is really NO DEVELOPMENT needed for this chip. Development for this chip was done when WRT160NL was finished. Create a new XML with different flash size is not considered development in here.

I have already a huge pile of routers collected over the years. Most of them were donated by members. I remember the last router was M20 donated by Canadian member which created new release of software.

On example of router I bought off the ebay was WNR854T which needs a lot of development. I was challenged by the member since at the time NT simply could not do it.

Recently I have client who used Atheros 9XXX in his development in Taiwan. I helped him remotely to create the XML and he is happy. So I have no doubt that NT now does support Atheros SPI flash with the framework same as WRT160NL.

I agree with you that the XML isn't a problem and neither is USB JTAG NT's capabilities since the WRT160NL is very much like this router and even the boards are laid out much the same way. (In reference to the position of the components and what ones are used)

I only got excited about the "Debug On" because I thought I made a break through and I didn't realize I was just getting a binary 1. (Active high) Before I went to bed, I noticed that pin 3 on the JTAG header connects directly with VCC and that explains the active high state, so I'll need to work on that a bit more.

[Killswitch]

Yea, sure - got the Weller iron w/ a fine tip around ten (10) years ago at a Distributor here in Philly for arounde $29.99 at the time. That's the starting point for anything detailed where ya don't want to leave blobs of solder that are almost, if not in contact with a circuit next door to it. Funny - I started with Radio Shack irons that didn't have a fine tip and their solder and it was a chore to lay a spot of solder dead on target. And, those cute little suction bulbs to remove excess solder. Damn, that's purely laughable in retrospect. That's Stone Age at its best. We all have humble beginnings. And, like you say, a Dremel seems to be next on the list. That'll put me where I need to be. The Heat Gun was gotten from the Harbor Freight for $12.99 (cheep) and has 2 temps. Low is 750 degrees F and High is 1550 F . Damn if that's not a Solar Flare ! There are NO thru-holes for the Header and I rationalized that I needed a Heat Gun. I'll punch the holes out with the Dremel and take from there. Would seem to be the best approach. Thanx for the heads-up .

[LightworkerNaven]

Yea, sure - got the Weller iron w/ a fine tip around ten (10) years ago at a Distributor here in Philly for arounde $29.99 at the time. That's the starting point for anything detailed where ya don't want to leave blobs of solder that are almost, if not in contact with a circuit next door to it. Funny - I started with Radio Shack irons that didn't have a fine tip and their solder and it was a chore to lay a spot of solder dead on target. And, those cute little suction bulbs to remove excess solder. Damn, that's purely laughable in retrospect. That's Stone Age at its best. We all have humble beginnings. And, like you say, a Dremel seems to be next on the list. That'll put me where I need to be. The Heat Gun was gotten from the Harbor Freight for $12.99 (cheep) and has 2 temps. Low is 750 degrees F and High is 1550 F . Damn if that's not a Solar Flare ! There are NO thru-holes for the Header and I rationalized that I needed a Heat Gun. I'll punch the holes out with the Dremel and take from there. Would seem to be the best approach. Thanx for the heads-up .

I'm glad I could help. For electronics you want to keep your heat gun at 500 degrees Celsius (932 degrees Fahrenheit) tops. The lowest you can go and still be effective is what you should set it at. I have one of those 2 setting heat guns myself, but they ruin electronics like there's no tomorrow. After I used it to strip down part of a board, I realized just how terrible they are. I've never had a problem with my soldering station, though, and with the tip being pin point perfect and not able to get messed up with using it a whole lot, $75 is a drop in the bucket for what I got.

My old Weller had a chisel tip that after a while of using it, it had a chunk taken out of the middle of it. I also had my thicker tips get all grotty on me, but this tip I've never gone wrong with. It's like it's indestructible. lol

[merkin]

A CET that calls it "white ink"...where is Capone?

All the tap signals are connected to vcc, thru a pullup, you already confirmed this.

Also usbjtagnt can already be used as a universal spi programmer.

Please stop assuming, @usbbdm has given you the right info. In fact you should not even connect usbjtagnt until you know the tap signals are correctly routed on the pcb. This can be done in seconds if you know what you are doing.

Of course you could take my other advice and just connect usbjtagnt directly to the cpu.

A heat gun for a surface mount header, really goon? And drilling thru a multilayered pcb, is not the best idea.

[Skillet50]

and I thought the drill method sounded good. Guess I'll keep using my PACE to suck that solder out of there. I'll jump in with Goon and say stay with it, cause I've hit a few dead ends, but by reading this thread the way might have been cleared. I remember I could get an ID code from a Marvell Xscale and even tried to build the Flash xml. , but it's been sitting ever since. I'm off to hit new dead ends.

[LightworkerNaven]

A CET that calls it "white ink"...where is Capone?

All the tap signals are connected to vcc, thru a pullup, you already confirmed this.

Also usbjtagnt can already be used as a universal spi programmer.

Please stop assuming, @usbbdm has given you the right info. In fact you should not even connect usbjtagnt until you know the tap signals are correctly routed on the pcb. This can be done in seconds if you know what you are doing.

Of course you could take my other advice and just connect usbjtagnt directly to the cpu.

A heat gun for a surface mount header, really goon? And drilling thru a multilayered pcb, is not the best idea.

Unlike you, I don't feel the need to impress anyone else. If I see white printing on the board, then I'll call it whatever the fuck I want to call it. It's printed on there and it's white.

As for modifying the circuit to make sure all the pins are connected properly, I don't have that experience doing it from scratch, so no it won't take seconds. Unless you've played with this board or are willing to help mod it, then you might want to stop acting like a know it all and only speak on what you have knowledge of.

The only info I assumed @usbbdm was giving me that was correct is the knowledge of how his device and software works. If I didn't, I'd just assume my efforts were worthless and scratch the project since it didn't work by slapping a header onto the board like he claimed it should.

I have to connect the USB JTAG NT device to test the connection when I think everything is correct and that's why I was doing. If I didn't think it was right, I wouldn't have hooked it up in the first place.

I don't know how to hook USB JTAG NT up as a SPI programmer as that's not documented. (To my knowledge) I know the software supports doing that, though.

As for your last statement about the heat gun, he wasn't experienced in this stuff, so he did the best he can. As for my solution of drilling though plugged holes, if I can drill through just solder and not ruin any pads, what's the problem? I've done it on other boards and they turned out fine. As you didn't give any solution for this issue, I think this is just another case of "Let's bully people and see if I can look cool," but in reality, bullies just look desperate for attention because they feel week and powerless. They feel the need to show off for others so they can be seen as "cool." Way to go, Merkin!

and I thought the drill method sounded good. Guess I'll keep using my PACE to suck that solder out of there. I'll jump in with Goon and say stay with it, cause I've hit a few dead ends, but by reading this thread the way might have been cleared. I remember I could get an ID code from a Marvell Xscale and even tried to build the Flash xml. , but it's been sitting ever since. I'm off to hit new dead ends.

Nah, the drill works fine and it's better than risking burning out the irreplaceable pads on what you need to keep heat on to suck the solder out of.

[LightworkerNaven]

A CET that calls it "white ink"...where is Capone?

All the tap signals are connected to vcc, thru a pullup, you already confirmed this.

Also usbjtagnt can already be used as a universal spi programmer.

Please stop assuming, @usbbdm has given you the right info. In fact you should not even connect usbjtagnt until you know the tap signals are correctly routed on the pcb. This can be done in seconds if you know what you are doing.

Of course you could take my other advice and just connect usbjtagnt directly to the cpu.

A heat gun for a surface mount header, really goon? And drilling thru a multilayered pcb, is not the best idea.

Unlike you, I don't feel the need to impress anyone else. If I see white printing on the board, then I'll call it whatever the fuck I want to call it. It's printed on there and it's white.

As for modifying the circuit to make sure all the pins are connected properly, I don't have that experience doing it from scratch, so no it won't take seconds. Unless you've played with this board or are willing to help mod it, then you might want to stop acting like a know it all and only speak on what you have knowledge of.

The only info I assumed @usbbdm was giving me that was correct is the knowledge of how his device and software works. If I didn't, I'd just assume my efforts were worthless and scratch the project since it didn't work by slapping a header onto the board like he claimed it should.

I have to connect the USB JTAG NT device to test the connection when I think everything is correct and that's why I was doing. If I didn't think it was right, I wouldn't have hooked it up in the first place.

I don't know how to hook USB JTAG NT up as a SPI programmer as that's not documented. (To my knowledge) I know the software supports doing that, though.

As for your last statement about the heat gun, he wasn't experienced in this stuff, so he did the best he can. As for my solution of drilling though plugged holes, if I can drill through just solder and not ruin any pads, what's the problem? I've done it on other boards and they turned out fine. As you didn't give any solution for this issue, I'd imagine you don't know of one.

and I thought the drill method sounded good. Guess I'll keep using my PACE to suck that solder out of there. I'll jump in with Goon and say stay with it, cause I've hit a few dead ends, but by reading this thread the way might have been cleared. I remember I could get an ID code from a Marvell Xscale and even tried to build the Flash xml. , but it's been sitting ever since. I'm off to hit new dead ends.

Nah, the drill works fine and it's better than risking burning out the irreplaceable pads on what you need to keep heat on to suck the solder out of.

[merkin]

viewtopic.php?t=7454

The issue is either lack of understanding or not listening.

You can use your multimeter to make sure everything is connected CORRECTLY before connecting the usbjtagnt.

moongoon was not talking about drilling through a via that is filled with solder...he said drill through the empty pads for a surface mount connector on a multi-layered pcb!...thats unbelievable.

[justsomeguy]

Unlike you, I don't feel the need to impress anyone else. .

I would just like to add to this thread that Merkin is a VERY VALUABLE asset to this community!!! he had put in a helluva lot of time into testing and finding new and interesting things in many many aspects of testing and does NOT need to impress anyone, nor does he try..

please keep on topic..

and have fun testing !!

[Killswitch]

No, it's very believable. And, I ain't gonna qualify myself with any level of accountibility. I said it and stand by it. Save the reply - most of the peeps here got ya figured. You're far too Professional for the likes of myself. Go take a nap.

[LightworkerNaven]

I would just like to add to this thread that Merkin is a VERY VALUABLE asset to this community!!! he had put in a helluva lot of time into testing and finding new and interesting things in many many aspects of testing and does NOT need to impress anyone, nor does he try..

please keep on topic..

and have fun testing !!

I agree and he helped me in the past, but I was referencing his bit about Capone. I was irritable when I wrote what I did, so don't take it personally. Sorry if I stepped on anyone's toes.

[LightworkerNaven]

This is my plan on what to do next. So I don't damage my unit, I want to know if this sounds legit or not. Also, when programming a unit with 2 flash chips, I program each chip separately and give each half of the image I want to use, correct? Is there a way to program both of them at the same time? If so, which chip get what connections? (Like should TDO go to chip 1 or 2, TDI to chip 1 or 2, etc.)

The pinout for the MX25L12845EMI-10G (25L128) numbered counterclockwise is: ( http://html.alldatasheet.com/html-pdf/2 ... I-10G.html )

Code: Select all

1 - NC/SIO3
2 - VCC
3 - NC
4 - PO2
5 - PO1
6 - PO0
7 - CS#
8 - SO/SIO1/PO7
9 - WP#SIO2 (Write protection pin)
10 - GND
11 - PO3
12 - PO4
13 - PO5
14 - PO6
15 - SI/SIO0
16 - SCLK

USB JTAG NT Pinout:

Code: Select all

1  - TRST
2  - GND
3  - TDI (SI)
4  - GND
5  - TDO (SO)
6  - GND
7  - TMS
8  - GND
9  - TCK (SCLK)
10 - GND
11 - RESET (CS#)
12 - NC
13 - DEBUG
14 - NC

So, I plan to map the following pins together. Again, if I can do both chips at the same time, let me know how, please. Also, if I don't need to map something, let me know that as well.

Code: Select all

Chip Pin - USB JTAG NT Pin
15 - 3 (SI - TDI)
8 - 5 (SO - TDO)
7 - 11 (CS# - RESET)
16 - 9 (SCLK - TCK)
10 - 2 (GND - GND)

[usbbdm]

If you plan to program the flash while it is on board, be careful about the ISP issue.
If you find the ISP you can program it on board.
If you apply the VCC to the flash, it might not work on some boards.

[LightworkerNaven]

If you plan to program the flash while it is on board, be careful about the ISP issue.
If you find the ISP you can program it on board.
If you apply the VCC to the flash, it might not work on some boards.

By ISP I assume you mean In System Programming, but what issue would there be and how would I go about finding that?

As for VCC, if it doesn't work, I'll just pop that pin off the board while I work. I'll probably reconnect it with a wire so I can pull it out whenever I need to program it again and not risk breaking off the pin.

[usbbdm]

By ISP I assume you mean In System Programming, but what issue would there be and how would I go about finding that?

As for VCC, if it doesn't work, I'll just pop that pin off the board while I work. I'll probably reconnect it with a wire so I can pull it out whenever I need to program it again and not risk breaking off the pin.

Pop off the chip will always work. Follow this video.