I'm working on a project including the OMAP4460 and TWL6030. Very similar to "Pandaboard". Now I experience the following:
I supply power (3,7V) to VBATT, I see all voltages ramping up on my scope. Just when the last voltage is fully ramped up, I see the power off sequence. NRESPWRON is never set by TWL. VIO1,8 is up for just 15,5ms. I see all voltages going down like shown in the datasheet. All voltages seem to reach full level, so I guess there is no short circuit.
Maybe I'm getting something wrong, but should the power on state last at least until the watchdog timer disables the device? Even if the OMAP is not responding? Since NRESPWRON is never released to HI level the OMAP will not start to boot.
Would a "short circuit" on a single rail result in a shut down? Or will it disable just a single rail.
I see VRTC present all the time. Also the 32khz clock is running. When I manually set the "PWRON" the whole sequence happens again. Power ramps up, power shuts down.
Thanks for any help on this!
We are also facing the same issue in our board. We are using TWL6030B1AACMR in our board.
Please let us know the complete part no. of the PMIC and have you have resolved it
Hi Hariseran, hi Ole,
we are having exactly the same problem with our board! The PMIC we are using is TWL6030B107CMR and the processor is OMAP4430 ES 2.2
I will be really appreciated, if you have solve this problem!
Hi Ole, Hans,
Yes we have resolved the issue. In our PMIC TWL6030B1AACMR primary watchdog timer was there. Watchdog timer expects all the PMIC outputs has to be available. We were not using VCORE3 SMPS and VMEM SMPS and connected the respective Feedback voltage to ground. So we connected some voltage to that feedback and it started working.
Hi Ole ,
Could you please tell me do you have any solution for this issue , you got yet ?
I am facing issue almost similar , But I couldn't get 1.8V ramp up too .
your point is correct I guess you also following the ref design as panda. they did same as Vcore3 and vmem are not used and Fb for those connected to GND .
and Datasheet of PMIC (TWL6030) also suggested same , if unused connect Fb pin to GND . So I guess may be some option to discard this situation instead of giving voltage to this balls ?
in my case I can't do this since I din't taken out this balls .. connected to GND in top layer itself under the BGA . :(
I wanted to try the modification Hariseran suggested. But It's the same here. I connected the FB balls to gnd, this connection is under the BGA now. I'm waiting for replacement chips which have the watchdog disabled. But I'm not sure that this is the solution.
I also contacted our TI FAE on this topic, he is checking this right now.
Sadly also the TESTEN pin is also inaccessible in my design, this pin can also disable the primary watchdog.
I hope the FAE will come up with a solution, If I have a solution I will post immediately.
Right now I have supplied my board voltages externally via a bench supply. So my board is working fine. But the PMIC still makes me grey hair.
P.S. I checked my supplier and my Chips should be P/N TWL6030B107CMR.
Hello Ole ,
I was facing some BGA Soldering issue , once i solved that now I am also in same stage as you face ,and I have the same part TWL6030B107CMR .
But as per the TWL6030 datasheet , this part is Primary watchdog disabled . how you concluded this is watch dog reset any particular observation for this ?
I was just guessing the watchdog is the problem. The "on time" of my PMIC is very reproducible every time I power cycle the board. This does not look like a "analog" problem to me. Even VRTC and 32k osc seem to work fine. Waking up the PMIC via "PWRON" pin works, but just for some milliseconds.
Hope TI FAE will come up with some solution or idea.
Hi Ole and Aneesh,
I and Hariseran work in same project. As per my observation there is short circuit protection inside PMIC which will trigger the primary watchdog to shut off the PMIC with in 10 ms(except for VRTC). I could see only feedback on each SMPS as sense line to detect short circuit and in our design we have grounded these FB lines (VMEM and VCORE3). I know this is contradicting, as datasheet recommends to ground the FB lines but other than this I don't see any explanation. We also used PMIC without watchdog in previous design and there was no problem even though feedback is grounded.
So I see my options are: New PCB with unused FB pins not connected GND, or a TWL6030 replacement with watchdog disabled. Do you have a list of available EPROM options / part numbers? I think the EPROM is not user programmable.
Hi Ole , please see the list of parts
Thanks for the list, I recently ordered TWL6030B107CMR as samples. Today I replaced my TWL with one of the samples. The problem still exists.
So sadly no news here.
Have you referred the schematics of Panda board with omap4460 or omap4430. There are some differences in the power connection of core1,2,3 for both these OMAP devices with the PMIC TWL6030.
Suhas told that the TWL6030B107CMR cannot be used for OMAP4430 as mentioned in the above table.
I was using the part "TWL6030B107CMR" by default , there we faced the issue . Now we changed the part with "TWL6030B1A4CMR" . PMIC is working properly with this Chip now .
I have analyzed the few things , and concluded that VMEM , Power supply FB connected to GND should be the cause . Observations to justify this are ,
i) in Panda Rev A they have the TWL6030B107CMR part and they using the VMEM power , Board is working with out issue .
ii) In panda ES , VMEM not used , FB is connected to GND . But they using the part " TWL6030B1A4CMR" .
iii) In Boot mode configuration for the PMIC ,
TWL6030B107CMR - VMEM will not disable only the voltage levels will change 1.2V or 1.35 V depends on the "Boot 1" pin state "low " or "high ".
TWL6030B1A4CMR - VMEM VMEM is not controlled by start-up sequence.
iv) If we select the Mode 1 as "high" . the same behavior we could produce on Panda ES Board .
I have tried all these from those I conclude VMEM Power , either we have to disable at power up (this option avilable only in TWL6030B1A4CMR or higher series .) else we have to use this power supply ( I mean should not connect the FB pin to GND as given in datasheet ).
Thank you for your detailed reply. I just ordered some TWL6030B1A4CMR from ti. I will post my results as soon as I get the devices.
Best Regards, Ole.
All content and materials on this site are provided "as is". TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with regard to these materials, including but not limited to all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement of any third party intellectual property right. TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with respect to these materials. No license, either express or implied, by estoppel or otherwise, is granted by TI. Use of the information on this site may require a license from a third party, or a license from TI.
TI is a global semiconductor design and manufacturing company. Innovate with 100,000+ analog ICs andembedded processors, along with software, tools and the industry’s largest sales/support staff.