Phase Shifted Full bridge with Voltage Mode Control - Does it really needs DC blocking Capacitor at Primary?

[Pulasthi_Perera]

I have designed and tested a Phase shifted full bridge with Voltage mode control and i designed it without the DC blocking capacitors at the primary. But there are mentions in literature that when PSFB is operating in Voltage mode control it needs a DC blocking capacitor to avoid possible transformer saturation due to flux imbalance so far i have operated the board up to 400W with up to 5A in the primary current but i didn't experienced any transformer saturations. Can Someone explain or give your opinions on this matter?

 

[cupoftea]

Sometimes if there is relatively high resistance in the primary fets or the winding, then the flux walking doesnt so much happen.
Though if you scope your primary current, through eg a low value resistor, then you will be able to see if there are signs of impending saturation. (ie, it has flux walked a bit, but not yet catastrophically).
Also, if you have done an "integrated transformer", ie, with inbuilt "extra leakage inductance", then it will be more immune to saturation, as the leakage inductor is effectively air-cored and so wont saturate.

 

[Pulasthi_Perera]

Sometimes if there is relatively high resistance in the primary fets or the winding, then the flux walking doesnt so much happen.
Though if you scope your primary current, through eg a low value resistor, then you will be able to see if there are signs of impending saturation. (ie, it has flux walked a bit, but not yet catastrophically).
Also, if you have done an "integrated transformer", ie, with inbuilt "extra leakage inductance", then it will be more immune to saturation, as the leakage inductor is effectively air-cored and so wont saturate.

Yes the transformer is have extra leakage inductance to achieve ZVS not a shim inductor. That might be the reason that I don't see the flux walking.FET resistance is 60mOhms.It is high cause voltage rating is at 650V.May be the both effects are combined and make the flux walking vanished.Thank you for your informative reply

 

[Easy peasy]

" Do I really need DC blocking Capacitor in the Primary? "

If your control can guarantee zero volt seconds average on the transformer primary - then you don't

Properly done peak current mode on the pri side can give this sort of control if done well

If you don't have this - then a blocking cap may well be a good idea - we can't see how good or bad your control circuit is, we can't see how you have made your transformer - we can't see how symmetric your circuit is . . .

 

[Pulasthi_Perera]

" Do I really need DC blocking Capacitor in the Primary? "

If your control can guarantee zero volt seconds average on the transformer primary - then you don't

Properly done peak current mode on the pri side can give this sort of control if done well

If you don't have this - then a blocking cap may well be a good idea - we can't see how good or bad your control circuit is, we can't see how you have made your transformer - we can't see how symmetric your circuit is . . .

Well i didn't Implement a PCMC because it is bit complex control algorithm compared to Simple inner and outer loop current control algorithm. So far i didn't had issues with transformer saturation but i fears it might occur if i ran it for a very long time such as 2hours or so.

 

[cupoftea]

some people kapton gap the cores, so that if it does go toward saturation, then it doesnt so suddenly saturate.
There is a way of "Looking" at the flux excursions by using a "search coil" inside the transformer. Just a very small extra coil wound in there, and have a very small load on it, ...that will show the flux waveform, albeit with the nuisance of having this extra test coil in your transformer.

 

[mtwieg]

There's no inherent law that says a voltage mode circuit with no blocking cap will saturate. But proving it won't requires analyzing every aspect of the circuit which contributes, and this should be done for all operating conditions (line, load, temperature, component tolerance, etc). It turns out to be an extremely involved exercise, as opposed to adding a capacitor or a current control loop.

Well i didn't Implement a PCMC because it is bit complex control algorithm compared to Simple inner and outer loop current control algorithm.

Peak current mode control is a simple innter/outer control scheme. At least I'm not aware of a simpler one.

So far i didn't had issues with transformer saturation but i fears it might occur if i ran it for a very long time such as 2hours or so.

Testing with a single prototype is absolutely not sufficient.

 

[Easy peasy]

You really need to test at full power, vary Vin, and then at max temp - also if you step the load from 20% to 100 % a few times - this will show any tendency to go to saturation - don't forget to put a quick blow fuse in the DC supply to the H bridge !

 

[Pulasthi_Perera]

There's no inherent law that says a voltage mode circuit with no blocking cap will saturate.

Yes exactly but I think it will be better if i add a DC blocking anyway it seems.

You really need to test at full power

Yes my rated power is 2.5kW but i still operated up to 600W.I might run in to saturation if i went full power.

Has anyone ran in to saturation while operating without a DC blocking capacitor?

 

[Easy peasy]

Has anyone ran in to saturation while operating without a DC blocking capacitor?

Ah - yes - anyone is a very wide range of punters - we have seen lots of " interesting " designs . . .