Can't reliably start 4P2 regulator for a 5V only setup (AN4199).

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Can't reliably start 4P2 regulator for a 5V only setup (AN4199).

2,849 Views
rabryn
Contributor III

I have an i.MX283 design with which I'm facing a problem reliably starting the 4.2V regulator on a new unit.  My hardware follows figure 9 of AN4199 with the optional RTC backup precisely.  I've modified mainline u-boot (2013.04) to configure the i.MX283 per the 5V only instructions outlined in the application note.  I have a 60mF super capacitor backing up the on chip RTC when power is removed.  I use Freescale's MfgTool to program brand new boards.  If I charge the super capacitor externally before powering up the board everything works well, however, if the super capacitor is completely drained (as it is on a brand new board) the 4.2V regulator will not start and the system will continually reboot preventing the MfgTool from loading the programming application.  Attached is my mainline u-boot patch for 5V only operation.


Is anyone successfully using AN4199 in the 5V only setup with a super capacitor backing up the RTC?  This product is actively being built and shipped, too much time is being spent charging super capacitors by the production team.  Any help would be greatly appreciated.

Original Attachment has been moved to: 0001_5V-only.patch.zip

Labels (2)
0 Kudos
Reply
9 Replies

2,411 Views
Yuri
NXP Employee
NXP Employee

You wrote about  5V-only configuration, in such case, according to section 4.1.2.2 (VDD5V source only: software configuration) of AN4199,

"The battery charger is disabled". I think for the case, when a super cap is applied it makes sense to use the default Linux BSP configuration.
"Be sure that neither NO_DCDC_BATT_SOURCE nor NO_VDD5V_SOURCE is defined in power_prep.c on the bootlet codes."

0 Kudos
Reply

2,411 Views
ryanschroeder
NXP Employee
NXP Employee

Hi Yuri, the question still remains. Does the circuit work with a discharged super cap?

0 Kudos
Reply

2,411 Views
Yuri
NXP Employee
NXP Employee

> Does the circuit work with a discharged super cap

In hardware sense - yes.

0 Kudos
Reply

2,411 Views
Yuri
NXP Employee
NXP Employee

   Please pay attention,  DCDC_BAT and  BATTERY pins of the i.MX28

should be connected (with each other), and - according to the i.MX28 Datasheet(s) -

their voltage should be in range  3.3 (3.1) - 4.2  V (in any case). This means battery
should not be "dead'.

0 Kudos
Reply

2,411 Views
ryanschroeder
NXP Employee
NXP Employee

Hi Yuri, it does not seem like Figure 9 in the app notes shows that the DCDC_BAT and BATTERY pins are connected per your suggestion. Do you know why that is? Does the Freescale app note need to be changed or am I reading the connections wrong?

0 Kudos
Reply

2,411 Views
Yuri
NXP Employee
NXP Employee

Really the main requirement regarding the DCDC_BAT and BATTERY is that they should be closed each other to avoid leakages.

0 Kudos
Reply

2,411 Views
rabryn
Contributor III

Thanks for the response.  Regarding AN4199, figure 9, how do you propose to bring up a board with C201 installed but not charged (as in a NEW capacitor)?

0 Kudos
Reply

2,411 Views
Yuri
NXP Employee
NXP Employee

VDD4P2 should provide the needed voltage offset.

0 Kudos
Reply

2,411 Views
bradsundstrom
Contributor II

Hi Yuri,

Do you know if circuit shown in figure 9 from AN4199 has been confirmed to work with a discharged supercap? Can you provide any guidance on what would need to be modified in uboot with regards to the PMU settings to allow the processor to boot and run with a completely discharged supercap?

0 Kudos
Reply