Programming kills MSP430F2274?

If I understand, you made a custom board with 2274 you are connecting to your pc through the ez430 "USB to SBW"?

On the eZ430, gnd and Vcc are pin 5 and pin 2 on the SBW connector, are you sure you are connecting your board to the true signal? It is easy to connect gnd on vcc and inversely. Are youpowering the 2274 board directly or with the ez430?

However I found yesterday that the BSL interface did'nt work anymore after a fuse blown on few devices (I'm not using SBW), didn't have the opportunities to spent time to work on this issue yet... perhaps a production problem with a batch?

Hello and thanks for reply.

Yes, you understand correctly. This is a custom board. I'm pretty sure that connection between eZ430 and board is correct I double checked it . Processor died during programming when it was powered from eZ430. What is stranger one of the boards I was able to program  normaly but it died when i started the debugger. 

Regards

The programming outputs of the FET are normal MSP port pins, perhaps with additional series resistor. I doubt they'll destroy the MSP.

One Szenario I can imagine is that the MSP is powered through the port pins. All port pins have clamp diodes which route overvolteage (or undervoltage) to VCC (or GND). This means, if the MSP isn't powered at all, current will flow form the port pins to VCC, and the MSP is powered from there. These clamp diodes, however, only endure up to 2mA continuous current. If the current is higher, they may melt, either disrupting the emergency connection from port pin to VCC, or, more likely, shortcutting the port pin to VCC (GND).

You should check your board once again whether the MSP really gets its VCC from the FET power out. (some FETs also have an external power-in, and if you mix up these two, things will go wrong.) A normal multimeter will only tell you that the MSP has VCC, btu not from where it comes.

the ez430 programmer has an LDO that regulates to 3.6V at some maximum current. I think it should protect itself from overcurrent but it could be damaged. DId you try another ez430 programmer?

I've dealt with many shorts on boards I've made, usually because the process I used didn't have any soldermask and that's not always good. I'vGo over each pin of your board
that's connected to the MSP430. Test every pin with respect to GND, especially VCC. Look at your schematics and verify that no pin is grounded where it shouldn't.

The F2274 is a small part and you can easily short things if you're not careful

It would help if you showed us your schematic. It might be a problem with that design.

GL

Ok, it seems that the problem is solved. After hours of experimenting the problem was tracked down to be the flux that I was using for reworking. The flux seems to be conducting some current which resulted in excessive current consumption an fried chips:)

Can anybody from TI recommend a flux for reworking because now I'm afraid to use anything new without recommendation.

Thanks for all replies.

Regards.

I'm not from TI, but I can tell you that there are three distinct categories of solder containing flux and standalone flux out there:

"Water soluble" which generally MUST be washed off of a PCB after it is used with hot water.

"No clean" which is formulated so that cleaning the flux residues from the PCB is not necessary in most applications due to it not being corrosive / conductive.

and general flux which is not marked as either of those categories which generally must be removed due to its corrosive and or conductive residual properties; typically these sorts of fluxes will need some kind of organic solvent or soap solution to effectively clean them such as 99% isopropyl alcohol or some purpose formulated flux cleaner/remover of whatever the manufacturer recommends.

So for ease of cleaning, if you're going to clean it, using lead free solder and water soluble flux is a good combination.  If you don't want to clean the unit and don't mind some tacky or spotty flux residues on the PCB, use "no clean" solder+flux and don't worry about it.

If you're dealing with optics or extremely low current picoamp/nanoamp signal metrology, it may not be a bad idea to clean residues from the PCB with a suitable cleaning process to avoid contamination and very small leakage currents.  Otherwise no clean should be all you need for a flux that shouldn't get conductive / corrosive on you.

Thanks for reply.

So it seems that the flux from no clean category is the way to go in my case. As sometimes the flux was getting underneath the chip which made cleaning it almost impossible.

And i really don't mind a bit  of spotty residue on pcb as long as it does not destroy it:)

Regards