MX535 and MC34708 Startup

[G1z]

I have a custom board with multiple iMX535's and MC34708 PMICs. Each processor is independent, there is no connection between them. (The design is very similar to the QSB.) One of the processor starts up just fine and enters serial boot mode as it is strapped to do. The other processor stays in reset. I think the MC34708 PMIC is the USER OFF state, RESETB is low, but RESETBMCU is high. All of the output rails are at the correct voltages. I've tried forcing different turn on events as described in section 7.5.3.4 of the datasheet, but RESETB stays low. There is nothing else driving the RESETB signal.

What condition(s) on the inputs to the MC34708 will cause it to assert RESETB low?

Is there is a simpler way to force a turn on event and put the MC34708 in ON mode?

[YixingKong]

Hi Geff,

DId Jorge answer your question? If yes, please click Correct Answer button, so that we can close the DI.

Thanks,

Yixing

[JorgeRama_rezRi · Accepted solution]

Hi Geff,

What condition(s) on the inputs to the MC34708 will cause it to assert RESETB low?

>> The RESETB pin is driven according to the operating mode of the MC34708. Based in the condition you specified (RESETB low, RESETBMCU high), your device is in User Off mode as you said. The strange fact here is that User Off mode is only entered by SW, the processor has to write a couple of bits in the MC34708 so this mode is achieved, plus, by default, regulators will be off in User Off mode, the only way to keep them on is by writing their respective SWxUOMODE bits by SW, so if your regulators are on, the MC34708 is most likely in on mode. Per your description, this issue is happening at first power up, right? If no SW is being executed, I'd look for connection issues.

Is there is a simpler way to force a turn on event and put the MC34708 in ON mode?

>> The simplest ones are battery attach and power button press. Does your system have the capability to generate these?

Questions:

- Please note that this pin is an open drain output and requires an external pull-up resistor. How are you connecting it?

- Do you see the CLK32K clock? This would be another indicator that the system is on.

- The i.MX53 should be able to start since the regulators are on and the RESETB signal is not related to it. Are you trying to run any SW? Is the only issue that the RESETB signal is low? Or are you experiencing problems trying to run your system?

Best regards.

Jorge.

[G1z]

Thanks Jorge for the info, very helpful.

Turns out a peripheral device also connected to the RESETB signal contains a large internal pulldown resistor. The voltage divider created by this pulldown and the external pullup for the MC3708 open-drain output would not always result in a high enough voltage level on the reset line.

The condition was made more confusing since some boards and processors would work fine while others would not. Tolerances and variations in the boards caused some to work and others to be on the edge!