Trouble Flashing 5272

[tod]

For the last couple of days I've been having trouble loading a new program over Ethernet. It will generate a bad checksum and there is no app and I'll boot right into the monitor. I'll power off/on and try a serial port reload and that will fail repeatedly. Often the program looks like it loads you get all the way through the process with the program copied message but then it generates an error and boots into the monitor. Then if I remove the Ethernet cable first, and do a a power cycle, I can reload via serial. Then I will typically get a few loads to work over Ethernet and then the process repeats. The programs themselves are never crashing.

I'm using a MOD5272 with a Mod-Dev-70 Rev 1.7 and powering and using serial via the USB. I've tried other Ethernet cables. Does this indicate a bad dev board, a bad module or something else?

[rnixon]

I wonder if it could be something your app is doing. A good test might be:
- Find the MOD5272 factory app.s19 file
- Run autoupdate as a separate utility
- Repeatedly do the update to see if you can duplicate the issue.

What is the module plugged into? What is the volt and curr capability of the power supply?

[tod]

Ha! You know I write bug-free code, so how could it be my app???

Power is coming from the USB connector. I looked at J2 pins 1 and 2 and I get a reading of 3.26 to 3.33.

I built the mod5272 factory demo (because now my confidence is shaken) and downloaded it and on the serial report it reports "Link Status: DOWN" which I don't think it should do. Then it reboots erratically. Sometimes as fast as 10 seconds other times every minute or so.

I got the factory demo to download over ethernet twice and then the third time I got the "No Valid app" message after the attempt.

I was then able to flash it over serial with the ethernet cable connected and it programmed and started the crashing loop again.

Its possible my other test program was also crashing and I just didn't notice it since I was just running quick test code for that lroundf() problem that was reported.

[tod]

Tried a second Mod5272 and I get the same behavior. If I remove the Ethernet cable the rebooting stops. I tried a different Ethernet cable but it behaves the same. I tried an external power supply but it also didn't help. Can some kind of ethernet traffic be causing this?

[tod]

Eliminate the impossible, whatever's left no matter how improbable....
I dug out my old (but trustworthy as it turns out) MOD-DEV 100 V1.04 board. Blew off the dust (literally) stuck in the original MOD5272 that was giving me problems, connected the same Ethernet cable and haven't had a problem for over 600 seconds. I'm willing to say there's something wrong with the MOD-Dev-70 board. Kind of strange that it only happens when the ethernet cable is plugged in.

[ecasey]

Hi Tod,

I have seen this behavior before, some suggestions:

1. Do you have anything connected to the dev board that might interfere with a signal path?

2. Are all the jumpers in factory position?

3. Take out any SD cards. (although that shouldn't matter, but sometimes a bad card can disrupt things).

4. Give the board a good cleaning (bush and compressed air) just in case some metallic dust got on it.

[tod]

1. Do you have anything connected to the dev board that might interfere with a signal path?
2. Are all the jumpers in factory position?
3. Take out any SD cards. (although that shouldn't matter, but sometimes a bad card can disrupt things).
4. Give the board a good cleaning (bush and compressed air) just in case some metallic dust got on it.

1. No, only the USB and Ethernet cables are connected, no external devices and nothing on the UARTs.
2. As far as I know. JP1 through JP3 are all on the "1". The SD Card Wp jumper is on. The dip switches are all in the OFF position (although now I tested with them all in the ON position)
3. No SD cards are installed
4. I brushed (lightly) and used compressed air. I did this with no MOD attached.

Behavior is the same. I also went back to the full factory config and put the MOD54415 in there, but it behaves like the MOD5272.
The relationship to the Ethernet cable has me perplexed.

[ecasey]

I once had a switch between the Module and the router that messed up the signals and kept the board from booting properly. I recall that it behaved differently with the 100 board and the 70 board, but I don't recall which was the problem. I replaced the switch and everything worked.

[dciliske]

Perhaps check the power rail on the Dev70? When you plug in the ethernet cable, the PHY will power up and draw an extra ~100-150mA. The supply is designed for a 1.5A (I think), but perhaps something is squirrely.

Edit: I should say, check the power rail after plugging in the ethernet.

Edit2: Perhaps the USB port you are connecting to is current limiting the board and you're encountering a brownout when you plug in the Ethernet and the PHY comes up. Try connecting the board to a standalone power supply (wall wart, bench supply, etc.) and see if the problem persists.

[pbreed]

I vote for the USB power limit.. ethernet comes up burns 40% more power...