CC2640R2L: Cc2640 reboot

Hi,

If I understood the issue correctly, you are powering the device through a LiIon battery and is experiencing problems when the system is immersed in low temperatures, right?

If so, I would carefully look at the battery behaviour itself. Not only the temperature range of the battery tends to influence this (some references mention issues when operating it at very low temperatures) but also the actual charge of the battery, which may be overvoltaging the device. To get a complete picture of the problem, I would try to use a data logger (a multimeter, a DAQ, etc.) to measure the voltage applied to the device over temperature and time to see if there are any power sags or overvoltages.

The temperature of the battery can influence its lifespan (not relevant for this problem) but also its maximum voltage, which may be just enough to overvoltage the device.

Also, although the nominal voltage of a LiIon battery can be 3.6 or 3.7V, depending on factors such as charge and construction it can reach higher voltages that may also put it above the device specifications. 

In both cases, the inclusion of a series schottky diode would reduce the battery voltage by 0.2~0.3V, which may be just enough to make the device work properly. 

Hope this helps,

Rafael

Hi,

Thanks for the additional details. From the description I can't necessarily say what is the exact problem, but your last post highlighted to me the brown out reset source (sorry I missed that from your original post). In general a brownout is indeed an undervoltage (a voltage drop below 1.8V on the device), therefore it would be interesting to take the voltage behaviour of the battery over time to see if there is a drop over temperature.

Your last post mentions turning the VDDS_LOSS_EN bit off, but to ignore the brown out reset detection mechanism the bit VDDS_LOSS_EN_OVR must be zero as well. I would do this as a test only, since I personally tend to refrain from permanently silencing such "alarms" as this might be hiding a deeper issue with the system with the battery or another system-level peripheral.

Regarding the inability to recover from reset, where exactly did you put the diode that prevented the effect from happening? Was it in series with VDDS, VDDR or with the battery itself? Perhaps a schematic diagram could help. Also, keep in mind there might be other factors on the system being introduced that may be preventing the device from coming up, such as external devices or the crystal, for example. Our app note SWRA495 contains a list of recommended crystals and other considerations.

At any rate, I will consult with colleagues to see if there are any additional aspects that I might be missing.

Best regards,
Rafael

We use CC2640 with 3.6V lithium battery. In VDDR power circuit
installed inductor 6.8 uH. Conducted several tests with different
capacitors in the VDDR power circuit at -5 degrees and below.
1. Capacitor 10uF. CC2640 stops working.
2. Capacitor 22uF. CC2640 restarts at a random time with
cause "Brown out detect on VDDR". After a successful OAD, the CC2640 does not
starts. Starts with a new program only at temperatures above
10 degrees.
3. Capacitor 44uF. CC2640 is stable. After a successful OAD
CC2640 starts up with "Brown out detect on VDDR" cause.
4. Capacitor 66uF. CC2640 is stable. After a successful OAD
The CC2640 starts up with the reason "Software reset via SYSRESET register".
5. Capacitor 44 uF, and in the power circuit 3.6 V is additionally installed
Schottky diode. CC2640 is stable. After successful OAD CC2640
starts with the reason "Software reset via SYSRESET register".
6. Capacitor 44 uF, and in the power circuit 3.6 V is additionally installed
resistor 18 Ohms. CC2640 is stable. After successful OAD CC2640
starts with the reason "Software reset via SYSRESET register".
7. Capacitor 44 uF, and in the power circuit 3.6 V is additionally installed
inductor 220uH. CC2640 is stable. After successful OAD CC2640
starts with the reason "Brown out detect on VDDR".

———————

please comment why is this happening? all nutrition advice says 10uF!!!

can i get your recommendation?
I ask to involve the above-mentioned colleagues in the discussion

10 uF: The CC2640 runs on a 3.2V battery, and does not work when
voltage 3.67 V.

Hi,

Thanks for the additional information. A few questions:

I assumed this is happening in your own hardware board, right? If so, does this happen in one of our development boards such as a Launchpad? The reason is that I am trying to isolate the device or the surrounding circuitry.

Were you able to monitor and log the voltages at VDDR and VDDS during the event? That would give additional insights on what may be happening.

The scenarios with the capacitors indicate that higher values (thus higher time constants to hold the voltage in case of a power fault) are providing the desired SYSRESET condition after an OAD operation. Given the OAD will write the device's internal Flash (a higher power operation when compared to a simple execution), it is possible the supply system is faltering. Again, a voltage monitoring would be useful to get additional visibility.

We have an app note about battery usage and characterization with CC devices - although it mostly covers coin cells, its discussion about peak current and internal resistance might give additional ideas.

I will keep the conversation with other folks and report back any additional findings and ideas.

Best regards,

Rafael

there are no peripherals on Launchpad that we use in our board, especially since there is another mcu there.
can we set 66uF instead of the recommended 10?

As for power monitoring, why do it if we connect a 3.2V battery and everything works well?

Hi,

About an hour ago our team discussed the design you sent offline and a colleague will provide his analysis and suggestions to this.

With the design in hands, we do not recommend to change the reservoir capacitor of the VDDR line (the output of the DCDC_SW pin) as it will certainly bring instability to the system.

The battery app note below (sorry, I though I had sent it on my prior post) has a method to calculate the supply reservoir capacitance when using batteries, which can reach the values you mentioned in your post but for the VDDS line.

https://www.ti.com/lit/wp/swra349/swra349.pdf

What I suspect the Schottky diode may be doing in this case is preventing the VDDR capacitor from discharging back through the DCDC_SW pin, which maintains the core voltage longer if the battery falters.

You should hear from my colleague with additional details in a very short time.

Regards,

Rafael

desouza said:

What I suspect the Schottky diode may be doing in this case is preventing the VDDR capacitor from discharging back through the DCDC_SW pin, which maintains the core voltage longer if the battery falters.

We do not use the Schottky diode in these tests, because we change the type of batteries
We reached 3.2v by connecting two AA 2x1.5v batteries
3.6v we use LiSOCl2 from !!!different manufacturers!!!, as well as Li-Ion. Batteries were placed both in the freezer itself and outside.

So that there were no questions about battery power, today we connected an external power supply with regulation .. And changed the voltage from 3v to 3.6v. the effect was the same (3v works fine, 3.6v restarts).

Hi,

You should have received a design review analysis. Please let me know if you are missing this. 

Regards,

Rafael

I did not receive anything

desouza said:

You should have received a design review analysis. Please let me know if you are missing this

Hello. they didn't send me anything. is it possible to speed up the process?
two weeks I stand still

Hi,

I apologize for the delay. I think the reviewer did not send the files. I just sent them myself to the e-mail account attached to the design review form. Please let me know if you received it.

Best regards,

Rafael

I bet the problems are caused by some kind of design issue. Either incorrect component values or poor layout. If you can share snippets of your schematic and layout showing all power connections I may be able to help. 

hello write your email

No, post pictures here if you want me to help. 

why only here?

hello, I got your analysis, I'm studying. As will be the result I will write. Thank you

Hi,

Thanks for letting me know. You can reply directly on the e-mail if you prefer.

Our reference designs are tested through the temperature range of the device, thus we are confident on its stability at the temperature boundaries you are using. 

Regards,

Rafael

Hello. Thank you. i answered you

my issue is still unresolved

maybe with a small consumption, we have 130 uA .. dc-dc is not stable. and should we try LDO?

Is there any recommendation as to which converter is best to use when?

The DCDC is stable in all conditions. It automatically switch to LDO if necessary. My offer still stands by the way.

I'm all ears

The L1 inductor value is wrong, have you tried using 10 uH? (With 10 uF cap?).

If that does not solve it, can you share the layout? Not the whole thing, power circuitry is sufficient.

Fredrik,

We have been working with the customer on design reviews (second so far), where this and a few other details were already raised. This is being worked internally. I will post updates.

Best regards,

Rafael

fredrikgk said:

The L1 inductor value is wrong, have you tried using 10 uH? (With 10 uF cap?).

If that does not solve it, can you share the layout? Not the whole thing, power circuitry is

yes, but it didn't help. new scheme

We have been working with the customer on design reviews (second so far), where this and a few other details were already raised. This is being worked internally. I will post updates.

I understand, but it is clear that you have not solved his problem. Which is interesting after two rounds of reviews. 

Why do you not post review feedback here in the thread so other people can understand where the problems are and thus hopefully avoid them themselves? It is possible to do that without disclosing details about the customer´s design. 

Alex Stilpans said:

yes, but it didn't help. new scheme

Ok, in that case I can only contribute if I can see the layout. 

How many boards have you tested? Do you see the same problem on all of them?

300 boards, 80% of them with a problem

Hello. ordered a board according to all the recommendations. now the board reboots at + temperature, and - it freezes and starts working only at +

electronic components used only those that you said