Common-Ground vs Non-Common-Ground DC–DC Converters

Thread Starter

Perapat2545

Joined Oct 9, 2025
1
I’ve seen that most conventional DC–DC converter topologies (like buck, boost, and buck-boost converters) share a common ground between the source and the load.

I want to know that what are the advantages of having a common ground between source and load? Additionally, in what applications would a non-common-ground converter be used instead?

I'd appreciate any insights into the advantages, disadvantages, and real-world applications of each type of configuration.
 

BobTPH

Joined Jun 5, 2013
11,463
A common ground means the two sides are not isolated. This means the two sides must be at the same potential. If they must be at a different potential, an isolated (no common ground) converter must be used.

With a common ground, signals can be connected directly between the two sides.

Without, signals going between them must go through some form of an isolator, usually an opto-coupler for DC or a transformer for AC.
 

michael8

Joined Jan 11, 2015
472
Also the failure modes can be different. For example if the pass transistor shorts in a non-isolated buck
converter the input voltage passes through to the output unreduced. An isolated converter would
have a transformer in the path and without AC from the transistor switching the output would be zero.
 

panic mode

Joined Oct 10, 2011
4,864
different reasons exist but isolation can prevent certain types of problems that would plague non-isolated types. and if needed one can always turn isolated converter to a non-isolated simply by providing a connection. and the reverse is not true - non-isolated converters cannot be used to make isolated one.

in this example inputs are on the left, outputs are on the right and red line indicates that B and D are internally connected... somehow (may not be direct short... even if that red connection is 10MOhm it is still not isolated).
1760019692182.png

isolation allows you to connect two circuits even if they are non-isolated but do not share common equipotential. this is also safe approach when interfacing to unknown equipment (when you are not sure if it is isolated or not). for example many old appliances (TV sets, Radio sets etc) had no isolation so one side of the circuit was practically connected to mains. and if the mains plug could be connected the other way, then metal chassis of the device you are using (or troubleshooting) would be live! even when life is not at risk, lack of isolation can open possibilities to ground loops and that means dramatically higher noise levels, rendering most functions useless.

consider this case... there are three circuits each with their own 5V supply (E1..E3) but they are not isolated from each other. in this case interfacing circuits powered by E1 and E3 is easy since they share same reference point. but if you want to interface circuits that are powered by E2 and E3, you must use isolation... if not E1 may be shorted and then sparks will fly.

you may think 'so what, i could account for E1 voltage'. but what if E1 could be changing? maybe it is adjustable PSU. you would be much better off to use isolation between any connections between E2 and E3 so that your circuit does not depend at all on state of E1 - even if E1 is switched off or unplugged from circuit.

1760020705673.png

Suppose you do have an isolation transformer or isolated DC-DC convertor but your application does not require isolation... then you have flexibility do some interesting things. For example you can combine input and output voltages to get other values. in AC this is called an AutoTransformer. I am not aware of term for DC but same rules apply.
 
Top