CC2540 custom board not being recognized by CC Debugger

Hi Frederik, how are you?

I write here not to open a new thread. I've developed a custom board based on the CC2540 and the CC debugger is not detecting my board either. We have the latest firmware on the debugger and it works fine with the keyfob and sensortag. But not with our board.

We tried powering the board either with  the CC debugger or via a battery and we have no response.

We also measured pin 40 obtaining 1.8V and pin 30 obtaining 1.22V as said on other posts.

We also measured the 32MHz crystal with an oscilloscope obtaining a 32MHz signal, as supposed.

We measured the DD DC GND VCC and RESET pins and there is no short circuit nor open circuits within the lines.

Is there a need to keep a length matching between RESET_N and DC or DD??

Do we need any specific firmware on the debugger or on the device??

I guess with this done we out out of options so I attach here the gerber files for you to make a quick check, if possible.

8512.Gerbers.zip

Thank you very much and I hope we can figure this out.

Cheers,

Matias

Hi Matias,

I cannot see anything that should prevent the debugger from connecting properly to the CC2540 in your design. The length of the reset trace compared to DD and DC is not critical.

Measure the voltage on all the VDD supply pins. Measure Reset, DD and DC with an oscilloscope to see if you have signals, are you sure the cable from the debugger is connected the right way?

Cheers,
Fredrik

Hi Fredrik,

Yes, we tested the boards a lot and everything seems to be okay. We took a few images with the osciloscope from the communication between the debugger and the SoC and we found a few strange  things.

When the debugger sends the CHIP_ID  command, the SoC responds with a 0xFF code instead with the 0x8D code corresponding to a CC2540 chip.

Actually the debugger makes 2 attempts to find out the chip ID and the chip answers both time a different thing. The first time answers with 0x00 and the second with 0xFF. I think this is why the debugger does not identify the boards.

Here I attache a few pictures showing this.

0317.Debug communication.pdf

The picture in this document shows the CHIP_ID command byte followed by the two response bytes.

I attach also a picture of the chip to see if you can identify where the chip comes from and if there is anything wrong.

Thanks for your time and have a good day.

Hi Matias,

Have you tested more than one board/chip?

Cheers,

Fredrik

Hi Fredrik,

Yes we built 10 boards and we have the same problem in all of them. I guess the manufacturer didn't use good chips.

We will try changing a few of them and see what happens.

Cheers

Matias.

Hi Fredrik,

We finally changed the CC2540 Chips and everything is working fine now. So it seems that there are some chips around there which are not working. 

I already sent a picture with the complete chip code, maybe you can track them and find the problem. I have no idea where the manufacturer took those chips from.

Hope no one else faces this problem.

Cheers,

Matias.

Hi Matias,

Great that you solved this issue. I will run the chip code through the tracking system and let you know if I find anything suspicious.

Cheers,

Fredrik

Hello,

Like everybody else, i am also facing problem with the detection of my cc2540 chip. With respect to the points above,i am sure i have taken care of the things but still i am in the problem. If i explain a bit more, what i am facing is.. In initial stage(i.e. just after board bringup), cc-debugger was able to detect, and we could debug the chip, after some iterations, we got problem of "sometime detection-sometimes not", and after then,cc-debugger doesnt detect at all. We built 3 boards, and faced the same problem one by one.

Thanks,

Gaurish

Facing a situation where you have a non-responding board including a CC2540 device can be due to a multitude of reasons. Using a non-optimized solder flow profile is one possibility potentially causing poor connection; too much/too little solder paste e.g. causing poor/no contact at one or more pins; damage - including ESD/EOS - to the device during assembly, test, handling or potential rework.

So how do you deal with such a situation? One method that is often effective is to probe the pin impedance of all pins and compare that of a good board to that of a non-working board. A multimeter is all that is needed. If you only have non-working boards you could use a CC2540EM if you have this available. Often a short or open can be detected this way. If the problem is due to the soldering process it can be remedied by proper device cleaning and resoldering; if the device was somehow damaged during the assembly, test or handling the device will have to changed. Doing a swap test, i.e. soldering a failing device onto a known good board will normally give the definite answer as to whether the issue follows the chip or not.

All CC2540 devices sold have been production tested and passed as good units prior to shipping. The likelihood of such units being damaged at arrival is very low. Good luck with the debug bringing your boards up. If you still struggle let us know and we can help out initiating an FA.

Happy holidays!

Regards
Chips

Hi Matias,

I checked the chip trace code and based on this I cannot see anything unusual with the device.

Have a nice Christmas,

Fredrik