# Power/control circuit isolation

#### smps97

I am designing a boost converter (of 50V, 500W) and I am unsure of isolation need.
I know that if there is AC voltage in a circuit (e.g. a rectifier/inverter) you have to isolate,
because voltage changes at various points, so there may be a current flow from the power to control circuit or reverse.
This situation produces noise and can even damage the control circuit if current is high.
I was wondering if we can avoid isolation in a high power boost converter (e.g. 100V, 5kW).
My question in fact is whether the need of isolation is because we want to avoid abnormal current loops or is also related with power level.

#### ron simpson

There is not enough information to know. I need to know much more about where the power goes and comes from. What is the load?

Example:
Input power is "+12V" and "Ground" 12V at 550 watts. Output is "50V" and "Ground2" at 500 watts.
If Ground and Ground2 can be the same thing. (connected together) then there is no isolation.
If Ground and Ground 2 must not be connected then you need isolation.

#### smps97

I have a solar cell array of 12V total and I want to power the TV antennas of the whole block of flats (about 7 TVs).
The voltage level needed is actually 36V (not 48V or 50V, I thought all telecoms work at this voltage level).
I am thinking of using and interleaved boost converter and go with the classic topology with the diodes
(not with synchronous rectification), to keep it simple.
So, I think there is no need of isolation, as you also say (with synchronous rectification I need isolation because of the high side MOSFETs).
Am I missing anything?

#### ron simpson

Boost almost never uses synchronous diodes. Synchronous diodes are only used at low voltage. (not 36V)
How much current? How much power per TV?

Here is an example, from data sheet, of 12V to 24V at 1A. Pretty simple. To get 36V try changing R1 to 1.5X. The max current will drop some.