amstan

MSP430FR5969 Wolverine Issues

18 posts in this topic

I recently designed a board that uses the MSP430FR5969 wolverine.

I got TI to send me a couple of samples for the first few boards.

 

The problem I'm having is getting the chip to respond in any shape or form.

 

I knew it would be hard to solder, being a QFN48 package. I used solder paste with a hot air gun. When i designed the board I made sure the pins were long so i could potentially touch them up with an iron after. I'm pretty sure all pins are connected right since i can see them from the side(the pin overhangs a little on the sides as well, it's not just under).

 

I'm using the MSP-FET430UIF to program it. I successfully managed to upgrade its firmware. I'm now using mspdebug with tilib.

 

The chip successfully gets programmed, I can even debug and step through the program. However it seems that the GPIO pins I tried so far just don't work, they just act like HIZ pins. The only pins I managed to get working were the PJ pins, the ones that jtag is also connected to.

 

I did make sure the watchdog was disabled.

 

Is there anything I'm missing? Perhaps a gpio mode I have to enable first? Did anyone else try the new chips, perhaps their design is broken or something, does this happen?

Share this post


Link to post
Share on other sites

I used solder paste with a hot air gun

 

Really, you should try the electric hot-plate method.

 

Moreover, I have never had much luck with solder paste, unless I also have a stencil.  It just gets everywhere and causes all kinds of problems.  QFN48 is an ultra-common package, so having a one-up stencil for it isn't a bad idea: http://www.stencilsunlimited.com/flip-up-stencils-p-362.html.  For example, I use QFN48 almost exclusively as an MCU package in my designs, and I use more Cortex-M than I do MSP430.  Everything comes in QFN48.

 

As a final note, make sure to clean the board thoroughly with a toothbrush you never plan on using, and high-concentration isopropanol (above 90%).  If you look at solder paste under a microscope, it is a combination of flux and thousands of tiny solder balls.  Those balls cause all kinds of problems if they are in places you don't want them, and they are impossible to spot with the naked eye.  Your board make look clean -- it isn't, unless you've cleaned it.  

ike and spirilis like this

Share this post


Link to post
Share on other sites

I've seen the "solder balls all over the place" crap cause issues... I don't do TSSOP ICs anymore with solder paste (reserving them for hand-soldering while I do all the time-consuming passives with paste).

Share this post


Link to post
Share on other sites

Sorry for OT, but you can always make simple stencil from soda can, using the same method as etching PCBs, but with hydrochloric acid and hydrogen peroxide. Takes 15 minutes (excluding scrubbing the internal layer of the can), Costs almost nothing and you have extra Coke to drink ;)

roadrunner84 likes this

Share this post


Link to post
Share on other sites

Sorry for OT, but you can always make simple stencil from soda can, using the same method as etching PCBs, but with hydrochloric acid and hydrogen peroxide. Takes 15 minutes (excluding scrubbing the internal layer of the can), Costs almost nothing and you have extra Coke to drink ;)

Cool tip.  I don't have any of those chemicals though.  :)

Share this post


Link to post
Share on other sites

hydrochloric acid - just google nearby supplier (sometimes it's called muriatic acid), hydrogen peroxide - any chemist nearby should have it. 

Just thermotransfer stencil image on the cleaned aluminium from the can (don't use sanding paper, just acetone + plastic scrubber), secure the area with tape, mix acid with hydrogen peroxide about 1:3,1:4. Etch for few minutes keeping eye on the process. When you see it's etched enough, rinse in the cold water. As I said - the whole process of etching takes maybe 10-15 minutes 

jpnorair likes this

Share this post


Link to post
Share on other sites

hydrochloric acid - just google nearby supplier (sometimes it's called muriatic acid), hydrogen peroxide - any chemist nearby should have it. 

Just thermotransfer stencil image on the cleaned aluminium from the can (don't use sanding paper, just acetone + plastic scrubber), secure the area with tape, mix acid with hydrogen peroxide about 1:3,1:4. Etch for few minutes keeping eye on the process. When you see it's etched enough, rinse in the cold water. As I said - the whole process of etching takes maybe 10-15 minutes 

Thanks again.  Muriatic acid is a combination of H2SO4 and HCl, but as an acid is freakin' strong.  That's probably all that matters.  I will google it to get more thorough description so I don't kill myself, but thanks again for the tip.

Share this post


Link to post
Share on other sites

Thanks again.  Muriatic acid is a combination of H2SO4 and HCl, but as an acid is freakin' strong.  That's probably all that matters.  I will google it to get more thorough description so I don't kill myself, but thanks again for the tip.

@@jpnorair Muriatic acid is HCl; should not contain H2SO4.

 

It's commonly sold as "swimming pool acid" to dissolve deposits on swimming pool walls, and in other places to bleach concrete and in others to remove rust / etch parts prior to priming for top coat paint applications.

 

When used as pool acid / treatment, it's often followed by sodium hypochlorite to neutralize the acid.

Share this post


Link to post
Share on other sites

The chip successfully gets programmed, I can even debug and step through the program. However it seems that the GPIO pins I tried so far just don't work, they just act like HIZ pins. The only pins I managed to get working were the PJ pins, the ones that jtag is also connected to.

 

 

Hi @@amstan,

 

I think I know what might be your problem.

 

In the FR5969 user's guide section 8.3.1 on p. 310, it mentions that on this device, all port pins are not only Hi-Z at startup, but that the pin initialization that you do does not take effect until you clear the LOCKLPM5 bit in the PM5CTL register. This is different than most other MSP430s, so it's an easy step to miss.

 

If you check the latest TI code examples for this device, this line has been added after GPIO initialization:

  PM5CTL0 &= ~LOCKLPM5;           // Disable the GPIO power-on default high-impedance mode
                                  // to activate previously configured port settings

Regards,

Katie

Edited by bluehash
[ADMIN] - Added @amstam to alert member.
jazz, amstan, bluehash and 3 others like this

Share this post


Link to post
Share on other sites

Not sure about FR5969, but I am pretty sure that my FR5739 (port output) is damaged. I use it for development of FRAM support for my SBW+ multi-programer. It survive lot of bad flashing (well it is FRAM, not flash, so this is not big problem), and many times it was connected wrong (mixed bad connection +3.3V - GND - PORT) by my mistake. After last time, when I made wrong connection, port is not working anymore. My SBW+ programer is using PX.0 pin for high speed IO transfer, and during flashing (under JTAG control) PX.0 on target device (FR5739) is blinking. After flashing is done successfully (write / verify), and PX.0 blinking example is started (released from JTAG control, free running mode) PX.0 is not blinking anymore. It was working OK with same PX.0 blinking example firmware, before, but not anymore.

 

For me not so important, because I still use it just for programer testing, because there is no problems until the device is under JTAG control. In my example, "PM5CTL0 &= ~LOCKLPM5;" didn't change anything. So if "PM5CTL0 &= ~LOCKLPM5;" don't help to @amstan, I'm afraid that this FR5969 is also demaged.

Share this post


Link to post
Share on other sites

hydrochloric acid - just google nearby supplier (sometimes it's called muriatic acid), hydrogen peroxide - any chemist nearby should have it. 

Just thermotransfer stencil image on the cleaned aluminium from the can (don't use sanding paper, just acetone + plastic scrubber), secure the area with tape, mix acid with hydrogen peroxide about 1:3,1:4. Etch for few minutes keeping eye on the process. When you see it's etched enough, rinse in the cold water. As I said - the whole process of etching takes maybe 10-15 minutes 

 

The acid + oxygenated water mix must be made with *extreme* care and with the proper safety equipment and procedures. Safety goggles, protective gloves, long sleeved shirt and rugged pants, a *very* well ventilated place and small quantities to test is the bare minimum. Don't even dream of smelling or being near the fumes.

HCl + H2O2 + H2O mix process is exothermic so unless done in the correct way, one can suffer painful and potentially handicapping accidents.

Not for the noob chemist to do unless the specifications (acid concentration, H2O2 dillution), proportions and procedures are detailed carefully. And proper warnings done.

 

Gast

Share this post


Link to post
Share on other sites

Not sure about FR5969, but I am pretty sure that my FR5739 (port output) is damaged. I use it for development of FRAM support for my SBW+ multi-programer. It survive lot of bad flashing (well it is FRAM, not flash, so this is not big problem), and many times it was connected wrong (mixed bad connection +3.3V - GND - PORT) by my mistake. After last time, when I made wrong connection, port is not working anymore. My SBW+ programer is using PX.0 pin for high speed IO transfer, and during flashing (under JTAG control) PX.0 on target device (FR5739) is blinking. After flashing is done successfully (write / verify), and PX.0 blinking example is started (released from JTAG control, free running mode) PX.0 is not blinking anymore. It was working OK with same PX.0 blinking example firmware, before, but not anymore.

 

For me not so important, because I still use it just for programer testing, because there is no problems until the device is under JTAG control. In my example, "PM5CTL0 &= ~LOCKLPM5;" didn't change anything. So if "PM5CTL0 &= ~LOCKLPM5;" don't help to @amstan, I'm afraid that this FR5969 is also demaged.

Hi Jazz,

 

Sorry to hear that about your FR5739 troubles. As a note, the PM5CTL0 &= ~LOCKLPM5 doesn't apply to FR57xx devices, it is something new on FR59/58xx Wolverine parts. So @@amstan was probably seeing a different issue than you are.

 

For your issue on FR57xx: you may have damaged your part as you said. If you are running exactly the same code and it doesn't work anymore (no changes to software at all) or if the same code runs correctly on a different FR57xx device, then it does sound like a hardware issue. If you are able to get other code to run, or if you had recently made a change to the FR57xx software, then it might still be some software problem though.

 

-Katie

Share this post


Link to post
Share on other sites

Hi @@amstan,

 

I think I know what might be your problem.

 

In the FR5969 user's guide section 8.3.1 on p. 310, it mentions that on this device, all port pins are not only Hi-Z at startup, but that the pin initialization that you do does not take effect until you clear the LOCKLPM5 bit in the PM5CTL register. This is different than most other MSP430s, so it's an easy step to miss.

 

If you check the latest TI code examples for this device, this line has been added after GPIO initialization:

  PM5CTL0 &= ~LOCKLPM5;           // Disable the GPIO power-on default high-impedance mode
                                  // to activate previously configured port settings

Regards,

Katie

Totally what it was!

I am very grateful!

 

All my pins work properly now!

KatiePier likes this

Share this post


Link to post
Share on other sites

If you check the latest TI code examples for this device, this line has been added after GPIO initialization:

  PM5CTL0 &= ~LOCKLPM5;           // Disable the GPIO power-on default high-impedance mode
                                  // to activate previously configured port settings

Regards,

Katie

 

Thank you @@KatiePier.  I've been going nuts trying to figure out what is going on with my LaunchPad.  This was it.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now