Full-bridge vs Half-bridge Fault Tolerance

[EE4me]

TL;DR Summary

Why is it said that full-bridges are fault tolerant compared to half-bridge

Why is it said that full-bridges are fault tolerant compared to half-bridge?

 

[Baluncore]

Why is it said that full-bridges are fault tolerant compared to half-bridge?

Where is that said ?
Is it said about rectifier bridges, amplifier H bridges, or Wheatstone bridges.

 

[EE4me]

Where is that said ?

One example is this paper..."Open-Circuit Fault Diagnosis and Fault-Tolerant Strategies for Full-Bridge DC-DC Converters".

Another is this one..."Implementation of a Fault-Tolerant AC/DC Converter for Permanent Magnet Synchronous Motor Drive Systems" which can be read here https://internal-journal.frontiersin.org/articles/10.3389/felec.2021.670077/full

I know what I've heard that's one of the benefits of going with a full-bridge topology as well but I've just never understood why. I've read a few papers in the past trying to understand it but I've never understood why.

 

[EE4me]

Is it said about rectifier bridges, amplifier H bridges, or Wheatstone bridges.

I think this part was added after I replied. I'm not sure if those are fault tolerant as well because that's actually part of my question. I guess it is not clear though so I apologize. What is fault tolerance and how is a full bridge when a half bridge is not?

I would assume there are topologies that offer fault tolerance just based on the nature of it's operation (like adding a zener for voltage clamping) but my guess is the full bridge would be more of a "fault tolerance if" kind of thing that comes down to the controls implemented since the two bridges could be operated synchronously making it equivalent to a h bridge with a higher current rating. There would have to be some control scheme offering the protection. I am unsure of what type of protection it can offer and how though.

 

[berkeman]

They may just mean that a full-bridge DC input circuit is polarity insensitive (it makes the correct DC output voltage no matter which way the +/- input is wired). A simple half-bridge DC input circuit blocks the input for the wrong polarity of input connection.

We use full-bridge inputs for our low voltage AC/DC power inputs on our modules.

 

[EE4me]

They may just mean that a full-bridge DC input circuit is polarity insensitive (it makes the correct DC output voltage no matter which way the +/- input is wired). A simple half-bridge DC input circuit blocks the input for the wrong polarity of input connection.

We use full-bridge inputs for our low voltage AC/DC power inputs on our modules.

I think it has more to do with it offering protection for the load. In the event of a fault, you can circulate the high current through the bridge to save the load. Thats about as far as my understanding goes, which is why I need to learn more. But it is possible as you mentioned because you can re else polarity.

 

[Baluncore]

Two half bridges make a full bridge. It seems sensible to throw out half bridges as they fail. By gating of the control signals the circuit is restructured and the power is reduced, which results in a limp mode.

 

[EE4me]

Two half bridges make a full bridge. It seems sensible to throw out half bridges as they fail. By gating of the control signals the circuit is restructured and the power is reduced, which results in a limp mode.

I'm not referring to just redundancy. Here is a quote directly from the paper "
A Comparison of the Battery Fault Tolerance of
Modular Multilevel Converters with Half-Bridge
and Full-Bridge Submodules"

"Since the half-bridge SM has fewer components
and lower power losses, it is more commonly used in MMC applications without batteries [3]. Nevertheless, the full-bridge SM provides some advantages, such as blocking the dc-fault current of the converter, while the half-bridge SM cannot.
Hence, the full-bridge SM is preferred in some applications."

 

[EE4me]

For clarification, SM is submodule and MMC is modular multilevel converter.

 

[Averagesupernova]

In the event of a fault, you can circulate the high current through the bridge to save the load.

That sounds like a lot of mumbo jumbo that makes no sense.

 

[EE4me]

That sounds like a lot of mumbo jumbo that makes no sense.

Maybe you can help me with a simpler more elegant
way to explain how it protects from the dc fault current of the converter? I posted the full paragraph and cited the paper I got it from, but ill post part of that except below again. This is referring full bridge submodules in a MMC.

"the full-bridge SM provides some advantages, such as blocking the dc-fault current of the converter, while the half-bridge SM cannot."

 

[Averagesupernova]

...how it protects from the dc fault current of the converter?

This also sounds like mumbo jumbo to me. Here's the thing: Until you can define what "fault current from the convertor" actually is, your questions are somewhat pointless.

 

[DaveE]

I agree, the question is nearly meaningless without specific examples. Real EE questions about SMPS pretty much require a simplified schematic. "Half-bridge" has several different contexts in my experience. Also, as others have said, what sort of fault?

I suspect their comment is somehow a result of one end of the winding (xfmr, motor, etc) being switched, and thus able to be disconnected, while the other end isn't controlled (like split capacitors, for example). So there may be a path for ground fault currents through the unswitched end of the winding. But, then this is also mumbo-jumbo, since I don't know what circuits I'm discussing.

 

[EE4me]

This also sounds like mumbo jumbo to me. Here's the thing: Until you can define what "fault current from the convertor" actually is, your questions are somewhat pointless.

Thats the way that its worded in all of the papers that I've read regarding the MMCs that mention the benefit of the full bridges fault tolerance. So it seems like the community reading it understand what's meant and we do not. Thanks for trying to help. Now I don't feel like I'm missing something obvious since you all don't get it either. Thanks for the responses.

 

[Averagesupernova]

Thats the way that its worded in all of the papers that I've read regarding the MMCs that mention the benefit of the full bridges fault tolerance. So it seems like the community reading it understand what's meant and we do not. Thanks for trying to help. Now I don't feel like I'm missing something obvious since you all don't get it either. Thanks for the responses.

Don't leave us hanging. If you find out more info, please share. The word 'bridge' in the context you've set here could mean several things. My mind goes to the bridge rectifiers in a power supply, or an H-bridge of mosfets that drive transformer or motor, etc.
-
Concerning the phrase 'fault current', that's pretty broad. Fault current refers to current that flows that shouldn't be flowing due to a short between a conductor and something else conductive.

 

[EE4me]

Don't leave us hanging. If you find out more info, please share. The word 'bridge' in the context you've set here could mean several things. My mind goes to the bridge rectifiers in a power supply, or an H-bridge of mosfets that drive transformer or motor, etc.
-
Concerning the phrase 'fault current', that's pretty broad. Fault current refers to current that flows that shouldn't be flowing due to a short between a conductor and something else conductive.

Sorry, I didn't realize there was confusion on what full bridge was referring to. It is referring to totem pole MOSFETS...or any other active semiconductor switches I suppose.

My understanding of what's meant by fault current is the same.

By fault tolerance, I THINK they're referring to protecting the load and not the converter. I also think it has to do with being able to invert the voltage.

I'll be sure to post an update when I figure more out.

 

[Baluncore]

Forget the "fault currents", they are irrelevant.

Power converters are typically designed as three phase converters. Each phase can be switched by either one half bridge as a single ended converter, or by two half bridges, one at each end, described as a full bridge converter.

A half bridge converter drives an LC series load against ground. If the half bridge fails, the LC load cannot be driven, so that phase of the converter is lost.

In a full bridge converter, one half bridge drives an LC series load against another half bridge. If either half bridge fails, it can be disabled, and the LC load connected to ground or power, while the other half bridge can continue to drive the LC load, but with only half the voltage swing on that phase.

Which explains why it seems that “full-bridges are fault tolerant compared to a half-bridge”, whatever that might mean.