SparkFun Forums 

Where electronics enthusiasts find answers.

Have questions about a SparkFun product or board? This is the place to be.
User avatar
By mitchschaft
#33609
...........................
Last edited by mitchschaft on Fri Sep 19, 2008 12:28 am, edited 2 times in total.
By n1ist
#33669
U140 looks like it is a MAX202 or similar level shifter. Measure the voltage between pins 15 and 16 (the two closest to C222). If it is 5V, use a MAX202.
If it is 3.3V use a MAX3232. That looks like a wide-SOIC footprint so the
parts from SparkFun won't fit without some creative soldering.

It looks like the RS232 port connects to the 3-pin header just to the right of J11 in your picture (below the 4 capacitors). The leftmost pin is ground (pin 5 of the 9-pin D). The other two go to pins 2 and 3 (or possibly pins 3 and 2) of the D connector. You can tehn plug the D connector into one of those adapters that you have listed from TigerDirect.

If you want TTL-level serial to feed a USB-TTL Serial converter or the one you have listed from SparkFun, try the two pads just below the "1" in "07/22-17" in the silkscreen.

I doubt the thing circled in red (a via or fiducial) is of any interest.

This is just from what I can see from your pictures; I have never worked with this device.

Good luck.
/mike
By wiml
#33671
Yeah, I think the thing circled in red is actually a registration mark / fiducial.

Maybe the "debug port" is the three contacts labeled JP3 (not quite visible in the picture you posted). It has a few traces leading to the unpopulated wide-SOIC spot. On the other hand, the board seems to use "J" for external connections and "JP" for configuration jumpers, so maybe JP3 isn't a serial console after all.

Sometimes these kinds of products are very, very similar to the reference designs provided by the CPU manufacturer. You might see if you can find a reference design for this CPU and compare it to ths board.
User avatar
By mitchschaft
#33716
....................
Last edited by mitchschaft on Fri Sep 19, 2008 12:29 am, edited 1 time in total.
User avatar
By mitchschaft
#33719
Awesome, pins 15 and 16 both measure 3.3 volts! I feel like i'm getting somewhere now.

What should my next step be, aside from looking for a reference model?
I thought the part I linked from spark fun would work if I solder leads to it then connect those to the board. I guess I don't understand. Are you saying the pins on U140 are so close together that it would be difficult to play with?
I'm not familiar with the lingo either (MAX3232, wide-SOIC, etc.. :D ). This project should help me with that and that's another reason why I'm doing this.
By busonerd
#33721
The MAX3232 part can come in a variety of different sized packages. The package that we carry isn't the same one as is intended to be used on the board, so you'll have a difficult time fitting it on there!.


Digikey should have the right sized package so its easy to solder right up.

Cheers,

--David Carne
User avatar
By mitchschaft
#33722
Hey, thanks, David! I checked out digikey's site and I don't know which milliamp value I need. There's a lot to choose from.
User avatar
By mitchschaft
#33723
I've searched for the past 3 hours for reference diagrams and schematics. ATI and Linksys won't give 'em up. I even did a live chat with a Linksys tech and they aren't even allowed to have them. :(
User avatar
By mitchschaft
#33725
.......................
Last edited by mitchschaft on Fri Sep 19, 2008 12:29 am, edited 3 times in total.
By busonerd
#33747
The MAX3232 is for serial port level conversion. My guess is that that 3 pin port on the side is simply a serial port.

JTAG requres a whole lot more pins.

Cheers,

--David Carne
User avatar
By mitchschaft
#33750
I'm not sure what you're trying to tell me :lol:.
By busonerd
#33752
I'm saying its a different debug port then the one referenced in the text.

The debug port that snippet of text referenced is a JTAG port. You need at least 5 IO lines [TMS TDI TDO TCK SRST] for a JTAG port, not to mention ground.

I suspect that 3 pin header at the edge of the board there is a standard RS232 serial port. The MAX3232 assumption further supports that idea, as a MAX3232 is used to do the voltage level conversion between the CPU and that kind of serial port.

Cheers,

--David Carne
By Philba
#33759
Hacking existing hardware is fun and maddening all at the same time. You need to be flexible and try lots of random things. Look for any door the designers forgot to close.

start tracing lines. Find the ejtag pins on the chip, see if they go anywhere. They may lead to a jtag header. Also, find any pins called RX and TX. Trace those. it's likely that they go to U140. Since you now know gnd, +V, RX and TX, you should be able figure out if it's some sort of RS232 driver in the max2xx family. You could then solder one in place.

However, I would build a small board with a 232 driver and hook up the 4 wires (rx,tx,+v,gnd). Then hook it up to a PC running something like hyperterm or similar. Boot the system and see if anything happens. you might need to fuss with different baud rates.

Another attack is to look at any upgrade firmware files. they are probably some standard format. Once you figure that out, you might be able to create your own. Make sure you know how to de-brick the device when (not if) you screw it up.

If this a networked device? I'd poke at the ports on the device to see what happens. some times they leave something open and it could give you a back door.

Finally, if you can track down the jtag port pins, you could get a debugger. I have no idea if something is available. reasonable price or not.

For me, the goal would be to figure out if there is some sort of boot loader. If there is, I would try to figure out how to use it to boot a file that I supply. Then I'd look to put Linux on it and go from there. I'd definitely bag WinCE.

[edit] I got curious and looked at the manual. the chip is a BGA so you won't easily get at the pins (balls, actually). The serial pins are called TD and RD but there is probably no way to trace. Do you have an O'scope or logic analyzer or access to one? great tool for this sort of thing. Might be hard without [/3edit]
User avatar
By mitchschaft
#33819
Very nice reply! I think that guy I've been talking with gave up on me. I haven't heard from him in a week.
This device is networked with both an RJ45 and a wireless connection. I found a cheap oscope on ebay that I might pick up. I understand that a 20MHz model is considered low end, but will that help me get by some hurdles with my project?
There's a device similar to this made by Roku called the roku hd1000. It uses the same CPU as this one, the xilleon 225. There are people who are able to hack in to that thing and put new software in it, so maybe I will have some luck with some help.
Thanks for checking out the manual for me. That's for the CPU itself instead of the whole board. I'm still looking for that info. I doubt I'll find it unless I can get ahold of someone on the inside of ATI.

I'll need some help on how to use the oscope if/when it gets here. I'm not sure where I'll stick the probe or what reading I'm looking for.