I've created a PCB that has two buck/boost converters, the second one being connected to the output of first one. The enable pin of the second buck converter is connected to the PG pin of the first buck converter, so whenever the first converter has a stable output, the second one will be turned on.
To convert 12-16 V down to 5 V, I'm using the TI TPS55165, in a design generated by WeBench. To convert the 5 V to 3.3 V, I'm using the TPS63050, with the design generated by WeBench.
Now, whenever I try and power the PCB's, either the 5V circuitry comes on, the board draws quite some current (500 mA) and the 3V3 is around 5 V, or the 5 V power manager shuts down because of over current protection inside of the chip.
Today I've downloaded PSpice for TI to have a go at simulating the different designs and each design works individually, but when they are in series, the overall design stops working.
It does seem to replicate my real world scenario by some degrees, since the 3V3 output shoots up to around 6 V, before it stops working.
I just checked the simulation again and I'm not really expecting the 3V3 buck/boost controller to do anything really, since the 5V's PG is not yet active. (or, could it be that the EN pin has a different polarity than I thought)
This is the simulation schematic
This is the schematic as produced
Debugging notes:
- Pulling down 5V PG, thus disabling the second state conversion, results in a clean 5V output, with regular current draw.
- One board sprung to life, with just a few mV's over on the 3V3 line (Around 3.5 V). It's strange because the example schematic (3V3 output) of this chip doesn't use a resistor divider on the feedback line, but WeBench suggested to use one. According to the simulation, the output voltage with the divider should be around 3.325 V
- Some of the boards draw a significant amount of current when supplying 5 V manually. The 3V3 circuitry tries to come alive, which is not working.
- On another board, when manually powering the 3V3 everything seems fine, but the 5 V circuitry does not work.
Here are some traces:
- Yellow: 3V3 line
- Pink: 5V PG
- Blue: 5V output
So what is this converter doing, it's not listening to it's enable input. I've checked again in the datasheet and EN is tied to Vin in the example diagram, so EN polarity is active high with 5V input allowance.
Full vertical zoom, when taking these screenshots, the circuitry was loaded with 6 Ω.
5 V (0.83 A):
3V3 (0.55 A):
The 3V3 circuitry seems to supply 3V3 only when loaded, when unloaded it supplies around 3.5-4 V.
