Hi, A little confused about the statement that the diodes must not be driven into reverse bias? If they are not driven into reverse bias, then they will have a forward voltage drop of lets say 0.7V. So if the bottom diode at 9:17 is not to be in reverse bias, then the + terminal should be negative. I am missing something here?
I say thatbecause, if the + voltage on A is large enough to drive the two series diodes into reverse bias to the mutual grounf (- terminal on B), then the regulator would basically short out. This is another way of saying your reverse breakdown voltage needs to be large enough so that the output voltage is dropped across your load and not across your diodes.
Placing 2 regulators in parallel is a bad idea. There is part-to-part tolerance. For example: 1 3.3V LDO, could be actually 3.28V, and the other could be 3.31V. If you placed these in parallel you're just burning power as one would drive the other (is syncable), or 1 would be used significantly more than the other. This is never done unless you have a very (very) unique circumstance, and for building an actual product, is a bad idea. Series is used all of the time, for example: Power regulators which have +/- rails for op-amps. Nonetheless, good overview and conversation pieces for us EE's. 👌
Thanks, I did mention it's best if they are designed for it, I agree that doing it with something like 2 LDOs is not a good idea especially if they have no way to track and sync outputs.
@@Zachariah-Peterson What if you put some diodes to isolate the individual outputs and the main output? I know the voltage drop would reduce the total output power and it's not recommended to do that for an actual product. But my usecase is that basically I want to test the power handling capability of a multi channel SMU (120 channels) by connecting all of them in parallel to an electronic load. If I connect them in parallel would that potentially damage the instrument? If I put diodes in between each output to prevent back powering every source channel would it have any unintended consequences?
What would be the problems if you were to parallel two buck converters that did not have any sync capabilities? Would this be an absolute no? Or is it still possible with diodes on the outputs?
It depends, be careful about that because you now have two components trying to drive each other other during switching, and the switching waveforms can be out of sync and could even be at different frequencies. The problem is that any time the potential difference looking back into one of the drives gets too high you could drive the gate into reverse or even drive one of the pins too far into reverse resulting in failure of the device, but that's generally if they don't have any built-in overload protection or the ripple voltage is very large. At minimum, you will have the FETs or the internal switcher just burning up power and creating excess heat due to reverse driving. The syncing basically aligns the ripple waveform in phase and frequency, so they will oppose each other and reach very low potential difference.
8:55 I am a bit confused about purpose of 2 diodes and what kind of "isolation" they provide. Can somebody explain?
Maybe , in case one of the regulators shuts down you don't drive current in it
Hi Zach, Make a video on the Power Sequencer.
Thank you 👍👍👍👍
Hi,
A little confused about the statement that the diodes must not be driven into reverse bias? If they are not driven into reverse bias, then they will have a forward voltage drop of lets say 0.7V. So if the bottom diode at 9:17 is not to be in reverse bias, then the + terminal should be negative. I am missing something here?
I say thatbecause, if the + voltage on A is large enough to drive the two series diodes into reverse bias to the mutual grounf (- terminal on B), then the regulator would basically short out. This is another way of saying your reverse breakdown voltage needs to be large enough so that the output voltage is dropped across your load and not across your diodes.
Placing 2 regulators in parallel is a bad idea. There is part-to-part tolerance. For example: 1 3.3V LDO, could be actually 3.28V, and the other could be 3.31V. If you placed these in parallel you're just burning power as one would drive the other (is syncable), or 1 would be used significantly more than the other. This is never done unless you have a very (very) unique circumstance, and for building an actual product, is a bad idea. Series is used all of the time, for example: Power regulators which have +/- rails for op-amps. Nonetheless, good overview and conversation pieces for us EE's. 👌
he said in the video @17:14 that you need to make sure that they are design for it.
LT304x LT309x
Thanks, I did mention it's best if they are designed for it, I agree that doing it with something like 2 LDOs is not a good idea especially if they have no way to track and sync outputs.
@@Zachariah-Peterson yeah I know thanks for replying!
@@Zachariah-Peterson What if you put some diodes to isolate the individual outputs and the main output? I know the voltage drop would reduce the total output power and it's not recommended to do that for an actual product. But my usecase is that basically I want to test the power handling capability of a multi channel SMU (120 channels) by connecting all of them in parallel to an electronic load. If I connect them in parallel would that potentially damage the instrument? If I put diodes in between each output to prevent back powering every source channel would it have any unintended consequences?
What would be the problems if you were to parallel two buck converters that did not have any sync capabilities? Would this be an absolute no? Or is it still possible with diodes on the outputs?
It depends, be careful about that because you now have two components trying to drive each other other during switching, and the switching waveforms can be out of sync and could even be at different frequencies. The problem is that any time the potential difference looking back into one of the drives gets too high you could drive the gate into reverse or even drive one of the pins too far into reverse resulting in failure of the device, but that's generally if they don't have any built-in overload protection or the ripple voltage is very large. At minimum, you will have the FETs or the internal switcher just burning up power and creating excess heat due to reverse driving. The syncing basically aligns the ripple waveform in phase and frequency, so they will oppose each other and reach very low potential difference.
Hello sir. Can you please provide a reference schematic for this combination of power regulators