Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Full Bridge SMPS ….why so much better than "identical" 2 Transistor Forward?

Status
Not open for further replies.
T

treez

Guest
Full Bridge SMPS ….why so much better than "identical" 2 Transistor Forward?

Hello,
The attached shows two LTspice simulations (schematics also included). One is a Full Bridge SMPS, the other is a Two Transistor Forward converter. They both are in DCM, currentmode, and at 100khz = F(sw) , pout=100W.
The transformer turns ratio, on-time and output inductor are the same for each converter. The primary and secondary currents are the same also. The transformer leakage inductances are the same, so are the RC snubbers across the secondary diodes.
Also, vout, vin and pout are the same for each converter.

However, the Full Bridge SMPS suffers far less overvoltage ringing than the Two Transistor Forward converter on its secondary diodes. (600V versus 1000V) Also, the total dissipation in the secondary-side snubber resistors is far less with the Full Bridge SMPS than the Two Transistor Forward. (0.6W total versus 1.8W total ).

Why is this?
 

Attachments

  • Schematic _Full Bridge SMPS _100W.pdf
    24.3 KB · Views: 94
  • Schematic _2 Transistor Forward _100W.pdf
    20.7 KB · Views: 81
  • 2 Transistor Forward _100W.txt
    10.6 KB · Views: 61
  • Full Bridge SMPS _100W.txt
    14.8 KB · Views: 66

Re: Full Bridge SMPS ….why so much better than "identical" 2 Transistor Forward?

For a 2 sw forward the Vout of the Tx must be twice the desired Vout as the o/p consists of effectively of a number of pulses with say 45% duty cycle to get Vout ave = desired, you need Vout Tx to be >2x Vout desired.

With a full bridge the Tx Vout is only a little higher than the desired Vout as you get pulses with ~90% duty cycle (very little dead time in the gaps compared to >50% dead time in the gaps for the forward)

Thus FB diodes need not be rated so high compared to the straight forward, and can be faster recovery...
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Re: Full Bridge SMPS ….why so much better than "identical" 2 Transistor Forward?

Thankyou, yes I agree with everything you say…..and what you are saying is obviously the right general comment about these two converters.
However, the question here is,
Why is there more overall loss in the secondary snubbers of this particular forward than the Full Bridge……they are “essentially” equivalent in every possible respect, so why is the forward dissipating more (overall more) in its secondary snubbers?

I mean the forward has less snubbers , but still dissipates more in its secondary snubbers overall.

(The full bridge smps here has been “pranged” to have exactly the same pri and sec current waveforms as the two transistor forward.)
 

Re: Full Bridge SMPS ….why so much better than "identical" 2 Transistor Forward?

I covered that, for the same o/p volts the FWD converter has twice the turns on the sec winding and therefore higher stresses on the o/p diodes, Power dissipated proportional to V^2 in the snubbers, so much less in the FB.
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Re: Full Bridge SMPS ….why so much better than "identical" 2 Transistor Forward?

Thanks, but the particular above forward and full bridge converters have exactly the same transformer. I appreciate you are talking "general case", whereas I am speaking of these 2 particular cases.
Its unfortunate that the simulations take so long to run, because if you run both simulations, you can see exactly what I mean. In both the forward and full bridge of the top post, the off-state diode voltage calculates out exactly the same, (however, due to the worse ringing with the above 2 tran forward, it sees far higher overvoltage on the diodes then the full bridge of the top post)
 

Re: Full Bridge SMPS ….why so much better than "identical" 2 Transistor Forward?

For the same o/p volts and a fully utilised Tx you cannot have the same Tx, the F.B. needs to have close to the desired Vout on the sec (just above) and the FWD needs to have just over 2x the desired Vout on its sec,

This is the particular case for comparison. Unless you are running twice the peak flux in the same Tx to get the desired Vout in the FWD configuration - that will affect losses also

When the FWD flies back (fets off) there is a lot more stress on the o/p diodes and snubbers than for the same Vout F.B., similarly when the fets come on there is 2x the driving voltage applied to the o/p stage (diodes, snub's) giving rise to more losses...
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Re: Full Bridge SMPS ….why so much better than "identical" 2 Transistor Forward?

Yes, thanks, but you are still talking of the general case, in this case, both are operating in DCM, and the off-state voltage to the secondary diodes is the same in the above (top post) case, for full bridge and forward. I deliberately "pranged" the full bridge of the top post so that the diodes receive the same off-state voltage as the Forward. However, there is still far more ringing with the forward and its the overvoltage ringing that's higher with the forward. The question is why is there more overvoltage ringing with the forward than the full bridge.

The transformer in both of the simulations of the top post are the same for the full bridge and the forward. (whether they "should best be" or not is not relevant to this post)

(if its of any use, in a different scenario we are using a 100w full bridge in dcm , but with a duty of 0.7, and vout=400v, vin = 390v for a led light supply....we are comparing and seeing if their any advantage in changing to a two transistor forward, but as the top post shows, forwards appear to suffer far more ringing and snubber loss, even when the full bridge is "kneecapped" to be the "same" as the forward in terms of on-time and transformer.)
 

Re: Full Bridge SMPS ….why so much better than "identical" 2 Transistor Forward?

over volt ringing on the o/p diode? this will always be due to the Tx sec L and the diode capacitance for DCM. For a slow diode there may be some energy stored in the FWD Tx (esp if it has a gap) sec L which will need snubbing.

In a F.B. you have four o/p diodes and four snubbers, and four fets and four snubbers, thus more bits to soak up energy at turn off...

Thus it makes sense that if you reduce the topology to essentially one o/p diode it will require at least 2x the snubbing to get similar turn off waveforms...

You don't get much saving in a FWD Tx either, as it has to have the same volt seconds as a F.B. to get the same average o/p volts (so same core size), and, lets say you operate at 50% PWM, the rms current is 70.7% of the full current, operate at 45% rms ~ 78% of full current - then you have to have a largish o/p choke to limit the current ripple to reasonable values (upping the rms even more for smaller chokes, in fact for DCM the rms I in the Tx could well be more (i.e. 2x) than for a F.B. in CCM with low ripple), with a F.B. you can run up to 98% PWM (as seen at the Tx out) and have quite a small o/p choke...
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top