but the buck converters had a hard time outputting the right voltage, so I went with the top design. It seems to work fine, so I'm just wondering if the first example is safe to use?
slipstick:
Either can work but the second version is better...why add extra load on output of the 5V converter?
Steve
When connecting them in the way of the second picture, they become unresponsive when adjusting their output. They either jump high or low, and don't change when near the center V, which is where my desired 5V is. Also, when I adjust one, the other changes. Should this not be the case?
If you mean it is difficult to adjust the voltage but it stays consistent once you take your screwdriver off the potentiometer, that is a problem with the quality of the potentiometer.
The main concern when cascading buck converters is their efficiency - If you feed a converter with an efficiency of 80% into another with an efficiency of 80%, your overall efficiency is 64%. And they are generally more efficient the higher the input voltage (and the higher the output voltage).
If they interact, you may need more capacitance on the power input. What 12 V supply is it anyway?
Usually efficiency is greater when input and output voltages are close to each other, so daisy-chaining
is likely to mean each converter is being more efficient than with independant converters, but that won't translate to overall efficiency unless the lower voltages are using much less power that the higher voltages.
So perhaps my take is use a direct conversion for anything using significant power, daisy chain the lower
power converters as is convenient.
Most complex electronics devices tend to have a cascade of supply converters, an advantage is you get repeatable and well defined power-up sequencing.
And you can cascade a buck converter with a linear regulator to optimize efficiency and get low-noise
supply as well - so for instance with 12V in, buck convert to 5V, linear regulate the 5V down to a clean
3.3V analog rail. Instead of dropping 8.7V in a linear regulator, you only drop 1.7V, so less waste,
less heat to get rid of, and no switching noise on the 3.3V rail... The 5V supply may be wanted anyway.
Paul__B:
If you mean it is difficult to adjust the voltage but it stays consistent once you take your screwdriver off the potentiometer, that is a problem with the quality of the potentiometer.
The main concern when cascading buck converters is their efficiency - If you feed a converter with an efficiency of 80% into another with an efficiency of 80%, your overall efficiency is 64%. And they are generally more efficient the higher the input voltage (and the higher the output voltage).
If they interact, you may need more capacitance on the power input. What 12 V supply is it anyway?
I mean they work great when adjusting each buck converter separately/ positioned in pic. 1. I am easily able to regulate voltage with upmost precision using the screw. The problem arises when trying to adjust them when they are positioned in the second picture. The voltage doesn't change unless you quickly torque the adjustment screw, but then it is either too high or too low. When it gets towards the center, ~4 - 7V, it just jumps either high or low. Very weird. The 12V supply is a car battery.
MarkT:
Usually efficiency is greater when input and output voltages are close to each other, so daisy-chaining
is likely to mean each converter is being more efficient than with independant converters, but that won't translate to overall efficiency unless the lower voltages are using much less power that the higher voltages.
So perhaps my take is use a direct conversion for anything using significant power, daisy chain the lower
power converters as is convenient.
Most complex electronics devices tend to have a cascade of supply converters, an advantage is you get repeatable and well defined power-up sequencing.
And you can cascade a buck converter with a linear regulator to optimize efficiency and get low-noise
supply as well - so for instance with 12V in, buck convert to 5V, linear regulate the 5V down to a clean
3.3V analog rail. Instead of dropping 8.7V in a linear regulator, you only drop 1.7V, so less waste,
less heat to get rid of, and no switching noise on the 3.3V rail... The 5V supply may be wanted anyway.
Many of my projects are 12V/5V/3.3V mixes (too many sensors that still demand 5V...)
Typical power chain: 12V - buck converter - 5V - linear converter (AMS1117-3.3 usually) - 3.3V. Makes for a very clean 3.3V supply to the microcontroller.