Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

MOSFET (Si/SiC) Forum Discussions

PowerStarter
Level 2
First like received 25 sign-ins 10 replies posted
Level 2

Hello guys,

I'm currently comparing the Full-Bridge LLC (FB-LLC) and the Half-Bridge LLC (HB-LLC) and I'm wondering until when the usage of the HB-LLC makes sense.

HB-LLC has multiple advantages over FB-LLC. Especially when splitting the resonance capacitance Cr. This would apply more stress to the capacitance, but it also stabilizes the input dc-link voltage more. Additionally it would prevent incomplete body diode reverse recovery, which can damage the power FET due to large current caused by the high reverse recovery charge.
Therefore it seems that the HB-LLC is the more reliable and more forgiving topology.

The FB-LLC has the main advantage that it uses the whole input voltage Vin, when the HB-LLC only uses half of it Vin/2.  This reduces the current stress on the components in the FB-LLC.  Due to the lower current and the two more switches the heat dissipation would be less and more distributed.
With a combination of phase-shifting and frequency variation the input voltage range of the FB-LLC can be increased.
One issue in a full-bridge LLC relates to holding the input voltage bus constant enough not to effect the resonant waveforms. [1]

The formulas also indicate that a smaller Cr and Lm is needed in the FB-LLC, while the Lr gets bigger. This would decrease the resonance circuit, but the bridge would get bigger and more expensive due to the higher component count.

For a high power LLC converter (1.5kW) the appropriate topology would be the FB-LLC.
But when I run the numbers for the HB-LLC topology than the current of the equivalent load on the primary winding side is only 8.33A. The magnetizing current is 3.44A and the resonant current is 9.01A. These current stress values seem as manageable.

Am I missing something or is the HB-LLC here still a good choice?

Best regards

[1] https://www.researchgate.net/post/What-are-the-challenges-in-full-bridge-LLC-resonant-converter

0 Likes
1 Solution
SPCH
Moderator
Moderator 50 sign-ins 5 solutions authored 25 sign-ins
Moderator

Hi,

Thanks for posting at Infineon Community.

Q: HB-LLC has multiple advantages over FB-LLC. Especially when splitting the resonance capacitance Cr. This would apply more stress to the capacitance, but it also stabilizes the input dc-link voltage more. Additionally, it would prevent incomplete body diode reverse recovery, which can damage the power FET due to the large current caused by the high reverse recovery charge. Therefore it seems that the HB-LLC is the more reliable and forgiving topology.

A: For the Power, you are working (1.5kW) definitely HB-LLC is a good choice.
Regarding Split capacitor your understanding is correct.

Q: With a combination of phase-shifting and frequency variation, the input voltage range of the FB-LLC can be increased. One issue in a full-bridge LLC relates to holding the input voltage bus constant enough not to affect the resonant waveforms.

A: In fact, a slight variation in Input DC BUS/LINK Voltage can provide optimal benefits from FB-LLC such as,
1. Operating FB-LLC in the vicinity/around Resonant frequency
2. Improves efficiency
3. Improves power density etc..

Q: The formulas also indicate that a smaller Cr and Lm is needed in the FB-LLC, while the Lr gets bigger. This would decrease the resonance circuit, but the bridge would get bigger and more expensive due to the higher component count.

A: For the power rating you are working, we recommend Split capacitor HB-LCC. If you go with FB-LLC as you stated component count will be more.

Q: For a high-power LLC converter (1.5kW) the appropriate topology would be the FB-LLC. But when I run the numbers for the HB-LLC topology than the current of the equivalent load on the primary winding side is only 8.33A. The magnetizing current is 3.44A and the resonant current is 9.01A. These current stress values seem as manageable.

A: For 1.5kW you can prefer HB-LLC.

Q: I missing something or is the HB-LLC here still a good choice?

A: HB-LLC is a good choice you can go ahead.

 

Regards,

SPCH

View solution in original post

3 Replies
SPCH
Moderator
Moderator 50 sign-ins 5 solutions authored 25 sign-ins
Moderator

Hi,

Thanks for posting at Infineon Community.

Q: HB-LLC has multiple advantages over FB-LLC. Especially when splitting the resonance capacitance Cr. This would apply more stress to the capacitance, but it also stabilizes the input dc-link voltage more. Additionally, it would prevent incomplete body diode reverse recovery, which can damage the power FET due to the large current caused by the high reverse recovery charge. Therefore it seems that the HB-LLC is the more reliable and forgiving topology.

A: For the Power, you are working (1.5kW) definitely HB-LLC is a good choice.
Regarding Split capacitor your understanding is correct.

Q: With a combination of phase-shifting and frequency variation, the input voltage range of the FB-LLC can be increased. One issue in a full-bridge LLC relates to holding the input voltage bus constant enough not to affect the resonant waveforms.

A: In fact, a slight variation in Input DC BUS/LINK Voltage can provide optimal benefits from FB-LLC such as,
1. Operating FB-LLC in the vicinity/around Resonant frequency
2. Improves efficiency
3. Improves power density etc..

Q: The formulas also indicate that a smaller Cr and Lm is needed in the FB-LLC, while the Lr gets bigger. This would decrease the resonance circuit, but the bridge would get bigger and more expensive due to the higher component count.

A: For the power rating you are working, we recommend Split capacitor HB-LCC. If you go with FB-LLC as you stated component count will be more.

Q: For a high-power LLC converter (1.5kW) the appropriate topology would be the FB-LLC. But when I run the numbers for the HB-LLC topology than the current of the equivalent load on the primary winding side is only 8.33A. The magnetizing current is 3.44A and the resonant current is 9.01A. These current stress values seem as manageable.

A: For 1.5kW you can prefer HB-LLC.

Q: I missing something or is the HB-LLC here still a good choice?

A: HB-LLC is a good choice you can go ahead.

 

Regards,

SPCH

PowerStarter
Level 2
First like received 25 sign-ins 10 replies posted
Level 2

Thank you SPCH for the quick response.

I've taken a look at the EVAL_3K3W_LLC_HB_CFD7, I couldn't find the two CoolMOS IPW60R018CFD7 (T3, T4) in the schematic. Are they hidden behind the U1 block?
Do you know if it is possible to run the EVAL_3K3W_LLC_HB_CFD7 only with low DC-voltage, or is it just getting too hot too quickly?

Regarding the splitted capacitor: It's simple extending the capacitor of the existing dc-link voltage right? So in order to keep the previous determined dc-link voltage ripple, the dc-link capacitor has to be become smaller.

When using e.g. GaN HEMTs (they don't have an intrinsic body diode), so there shouldn't be the problem with incomplete body diode reverse recovery.

Best regards

0 Likes
SPCH
Moderator
Moderator 50 sign-ins 5 solutions authored 25 sign-ins
Moderator

Hi,

Q: I've taken a look at the EVAL_3K3W_LLC_HB_CFD7, I couldn't find the two CoolMOS IPW60R018CFD7 (T3, T4) in the schematic. Are they hidden behind the U1 block?

A: The FETs are connected Under Ref U1. Please refer to the below Fig.

SPCH_0-1655315113918.png


Q: Do you know if it is possible to run the EVAL_3K3W_LLC_HB_CFD7 only with low DC voltage, or is it just getting too hot too quickly?

A: The transformer magnetizing component is a function of operating frequency and applied DC-Link voltage, improper combination can lead to FET hard-switching so, we recommend you operate the EVAL board as per the boundaries mentioned in an application note.

Q: Regarding the split capacitor: It's simple extending the capacitor of the existing dc-link voltage right? So in order to keep the previous determined dc-link voltage ripple, the dc-link capacitor has to become smaller?

A: Split Capacitor config. will help to reduce RMS currents demand from DC-Link capacitor as compared to single capacitor half-bridge. Accordingly, you can size the DC-Link Capacitor (For a stand-alone HB-LLC).

Q: When using e.g. GaN HEMTs (they don't have an intrinsic body diode), so there shouldn't be a problem with incomplete body diode reverse recovery.

A: Since Qrr for GAN HEMTs is 10-fold lower/Negligible, the Recovery effect may not cause device failure.

Thanks,

SPCH