Powerup problems with custom board

[sh9]

I am currently trying to bring up a custom board with the OSD3358-512M-BSM that is powered through VIN_AC only. I can”t get it to start, and I wonder why.

I haven’t wired up VIN_USB and VIN_BAT, because I don’t need it. A regulated 5V source feeds VIN_AC, and its risetime is way within limits (a clean ramp with 2 ms duration). Unfortunately, I don’t have access to unused signals, because they’re not wired out from underneath the package.

What I could measure is that the SYS_RTC_1P8 rail (i.e. the LDO1 output of the PMIC) doesn’t ramp up. This, presumably, prevents further steps in the powerup sequence from occuring. It doesn’t seem as if the SYS_RTC_1P8 was shorted to anything, AFAICT, and there is no load on it except RTC_KALDO_ENN. Yet it doesn’t move.

I left PMIC_BAT_SENSE floating instead of connecting it to VIN_BAT, because I assumed that to be of no importance when not using the battery functions at all. Is it conceivable that this could be the source of my problem? Or what else could be amiss?

Thanks for any hint that could help.

[Neeraj Dantu]

Hey sh,

Given your description of the behavior, there could be more than one reasons for the failure. As you mentioned, please make sure that the power rails are not shorted. SYS_VOUT powers the SYS_RTC_1P8V LDO. So, checking SYS_VOUT is a good first step. Apart from that, please go through the schematic checklist(https://octavosystems.com/app_notes/osd335x-schematic-checklist/) for any other potential issues with the design.

Other reasons for the part not booting up could be bad reflow profile, part orientation or not pre-baking the parts.

How many boards have you verified to have this fault condition?

Best,

Neeraj

[sh9]

Sorry for the delay in replying, I had a couple of days off due to holidays.

I went through the schematic design checklist document, and these are the deviations from it:

• PMIC connections are as in Figure 1 (RTC subsystem enabled), even though the RTC is not intended to be used. However, CAP_VDD_RTC and RTC_KALDO_ENN are connected as in Figure 2.
• No clamping circuit is used.
• The BAT pins are connected with each other, but are otherwise floating. BAT_SENSE is also unconnected (i.e. not connected to BAT).
• WARMRSTN is undriven. It is not used to reset external devices. It merely connects to the JTAG connector.

Regarding the clamping circuit, I have my reservations which I open another issue for.

Some crucial signals have, unfortunately, not been brought out from underneath the package to test points, so my opportunities to measure their voltage are quite limited. I basically can manage to probe balls on the outermost row.

I am not sure how my “mixing” of Figure 1 and Figure 2 connections impacts powerup. I find the PMIC data sheet difficult to understand in this respect.

Another suspicion of mine is that not connecting BAT_SENSE to BAT might disturb SYS_VOUT switching, but unfortunately I have no access to SYS_VOUT, because it hasn’t been brought out to a test point.

It doesn’t seem as if anything was shorted signals or badly orientated parts. I don’t know if the soldering profile or part handling was correct, I would have to investigate that if necessary. I have two boards, both show the same behavior.

[Neeraj Dantu]

sh,

In theory, the mixing of figures 1 and 2 in your circuit seems reasonable. Not connecting BAT_SENSE to BAT might cause power up issues, but the PMIC should at least try to bring up the power rails.

Another signal you can check is the PMIC_NRESET which if grounded could be causing the PMIC to be in RESET.

SYS_VOUT is the power supply input for the LDOs and DCDC converters in the PMIC(https://octavosystems.com/app_notes/osd335x-sm-power-application-note/). So, a short on this rail could also be causing the power rails to not come up. It is unfortunate that you do not have access to this rail.

One more thing to verify is the footprint for the OSD335x-SM being used. Please make sure all the pin names are correct and they are in the correct positions.

As mentioned earlier, pre-bake and reflow profile are additional avenues to investigate.

Please let us know if that helps or if you have an update.

Best,

Neeraj

[Author]

Posts