General > General Technical Chat
DC-DC converter with floating ground
electrodacus:
--- Quote from: Zero999 on December 28, 2023, 04:00:26 pm ---
No, it's not possible to do this. A voltage regulator which goes inline with the input like that, must have a common i.e. 0V connection. There is no way to design a regulator which goes series with a power supply like that, be linear, or switched mode and regulates the output voltage. All voltages are relative, so common connection is required to monitor the output potential difference.
--- End quote ---
When you say not possible. Are you referring to not possible to have V2 > V1 continually or not possible to even have V2 for any amount of time above V1?
In the diagram you quoted it is possible to have V2 up to 3x V1 (30V) for short periods of time before V2 decreases below V1 in order to recharge the capacitor back to 20V.
Andy Chee:
--- Quote from: electrodacus on December 28, 2023, 04:09:50 pm ---When you say not possible.
--- End quote ---
I believe he's talking about the voltage feedback to the switching controller. At minimum, the feedback network MUST be referenced to the output voltage ground in order to measure the output voltage.
OTOH, if this regulator is just a thought experiment, then you don't really need feedback and just leave it open loop. However, without feedback you can't really call it a "regulator" can you? ;)
electrodacus:
--- Quote from: Andy Chee on December 28, 2023, 04:28:39 pm ---I believe he's talking about the voltage feedback to the switching controller. At minimum, the feedback network MUST be referenced to the output voltage ground in order to measure the output voltage.
OTOH, if this regulator is just a thought experiment, then you don't really need feedback and just leave it open loop. However, without feedback you can't really call it a "regulator" can you? ;)
--- End quote ---
You can have an internal microcontroller running say at 3.3V from a separate capacitor where you have a voltage reference or just use the 3.3V supply to fairly accurately charge the capacitor to 20V.
I do not want for this to get complicated the point I'm trying to make is that is possible to have V2 higher than V1 but only intermittent and V2 needs to drop below V1 else the capacitor can not be charged.
Zero999:
--- Quote from: electrodacus on December 28, 2023, 04:09:50 pm ---
--- Quote from: Zero999 on December 28, 2023, 04:00:26 pm ---
No, it's not possible to do this. A voltage regulator which goes inline with the input like that, must have a common i.e. 0V connection. There is no way to design a regulator which goes series with a power supply like that, be linear, or switched mode and regulates the output voltage. All voltages are relative, so common connection is required to monitor the output potential difference.
--- End quote ---
When you say not possible. Are you referring to not possible to have V2 > V1 continually or not possible to even have V2 for any amount of time above V1?
In the diagram you quoted it is possible to have V2 up to 3x V1 (30V) for short periods of time before V2 decreases below V1 in order to recharge the capacitor back to 20V.
--- End quote ---
I mean it's impossible to regulate the output voltage, without a common connection to the input.
You schematic is also drawn contrary to the convention of having the input on the left and output on the right.
Here it is flipped the right way round, with the common/0V connection added.
--- Quote from: Andy Chee on December 28, 2023, 04:28:39 pm ---
--- Quote from: electrodacus on December 28, 2023, 04:09:50 pm ---When you say not possible.
--- End quote ---
I believe he's talking about the voltage feedback to the switching controller. At minimum, the feedback network MUST be referenced to the output voltage ground in order to measure the output voltage.
OTOH, if this regulator is just a thought experiment, then you don't really need feedback and just leave it open loop. However, without feedback you can't really call it a "regulator" can you? ;)
--- End quote ---
No, it still won't work because of KCL.
electrodacus:
--- Quote from: Zero999 on December 28, 2023, 05:47:41 pm ---
I mean it's impossible to regulate the output voltage, without a common connection to the input.
You schematic is also drawn contrary to the convention of having the input on the left and output on the right.
Here it is flipped the right way round, with the common/0V connection added.
--- End quote ---
I do not think there is any convention regarding where the input needs to be but if that makes it easier to visualize I have no problem with that.
I do not want that connection added. That connection missing is the entire point of the exercise.
With the floating 0V negative connection a circuit inside the green box can still get the V2 = 3*V1 as peak voltage.
So I can have a V1 = 10V constant the internal circuit will be able to calculate that it has 10V at the input or 10.1V using a micro controller with cleaver algorithm and current and voltage measurements.
But I do not want to get the circuit and explanation to complex. I just want for everyone to agree that V2 can be three times V1 for short periods of time and V2 needs to decrease below V1 in order to be able to recharge that internal capacitor back to 20V and so increase the V2 again.
--- Quote from: Zero999 on December 28, 2023, 05:47:41 pm ---No, it still won't work because of KCL.
--- End quote ---
Kirchhoff current law will not prevent V2 to be higher than V1 temporarily with no other connection on that green box other than Battery+ and Load+ (the 100Ohm resistor).
The box can contain capacitors that are energy storage devices.
Those capacitors can be charged from battery trough the 100Ohm load resistor so V2 < V1 then after those are charged they can be connected in series with the battery in order to have V2 >> V1 for some limited amount of time as capacitor will of course discharge.
You can have another capacitor internally that supplies a low power micro controller and be able to accurately calculate the battery voltage and even the resistor value as you can have an internal voltage reference say a 1.2V to be able to compare against.
So you can accurately fluctuate V2 to be between say 5V and 30V
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version