What is better for DRSSTC: full-bridge or double half bridge?
BSVi, Tue Feb 23 2010, 07:27AMI'm going to redesign my DRSSTC (
and I'm in doubt whitch is better for DRSSTC - full-bridge or half-bridge with two paralleled IGBTs.Half-bridge with paralleled devices gives double curent, and full-bridge halfes rise time.
In Terry's pdf "Steve's vs. Terry's coils..." he says that rise time is critical.
So what is better?
Re: What is better for DRSSTC: full-bridge or double half bridge?
Luca, Tue Feb 23 2010, 08:46AM
I don't think that paralleling IGBTs is a good idea... Conduction mechanisms in IGBTs are a mix of MOS and bipolar mechanism (actually they are bipolar devices), and it is well known that it is not a good idea to parallel bipolar devices (BJT), unless you put resistors to equally divide the current in the devices.
I'm not totally shure but I think that this applies also to IGBT.
Regards,
Luca
Luca, Tue Feb 23 2010, 08:46AM
I don't think that paralleling IGBTs is a good idea... Conduction mechanisms in IGBTs are a mix of MOS and bipolar mechanism (actually they are bipolar devices), and it is well known that it is not a good idea to parallel bipolar devices (BJT), unless you put resistors to equally divide the current in the devices.
I'm not totally shure but I think that this applies also to IGBT.
Regards,
Luca
Re: What is better for DRSSTC: full-bridge or double half bridge?
mmt, Tue Feb 23 2010, 10:12AM
i have never see a paralled igbt's..
mmt, Tue Feb 23 2010, 10:12AM
i have never see a paralled igbt's..
Re: What is better for DRSSTC: full-bridge or double half bridge?
Steve Conner, Tue Feb 23 2010, 10:22AM
Full bridge is better!
Steve Conner, Tue Feb 23 2010, 10:22AM
Full bridge is better!
Re: What is better for DRSSTC: full-bridge or double half bridge?
BSVi, Tue Feb 23 2010, 10:22AM
There are several IGBTs especially designed for parallelling. They have slightly positive temperature coefficients and well matched.
BSVi, Tue Feb 23 2010, 10:22AM
There are several IGBTs especially designed for parallelling. They have slightly positive temperature coefficients and well matched.
Re: What is better for DRSSTC: full-bridge or double half bridge?
Luca, Tue Feb 23 2010, 11:45AM
I see...
In any case, I would choose full-bridge because losses on resistive paths (windings, cables, ESR of capacitors...) increase with current^2...
Regards,
Luca
Luca, Tue Feb 23 2010, 11:45AM
BSVi wrote ...
There are several IGBTs especially designed for parallelling. They have slightly positive temperature coefficients and well matched.
There are several IGBTs especially designed for parallelling. They have slightly positive temperature coefficients and well matched.
I see...
In any case, I would choose full-bridge because losses on resistive paths (windings, cables, ESR of capacitors...) increase with current^2...
Regards,
Luca
Re: What is better for DRSSTC: full-bridge or double half bridge?
BSVi, Tue Feb 23 2010, 03:42PM
>Full bridge is better!
Would you please explain why?
BSVi, Tue Feb 23 2010, 03:42PM
>Full bridge is better!
Would you please explain why?
Re: What is better for DRSSTC: full-bridge or double half bridge?
GeordieBoy, Tue Feb 23 2010, 03:52PM
> Would you please explain why?
...because you develope twice the peak-to-peak voltage across the load with a full-bridge than you do with a a half-bridge.
Losses can also be spread evenly between 4 switches in a full-bridge which reduces temperatures and improves reliability.
-Richie,
GeordieBoy, Tue Feb 23 2010, 03:52PM
> Would you please explain why?
...because you develope twice the peak-to-peak voltage across the load with a full-bridge than you do with a a half-bridge.
Losses can also be spread evenly between 4 switches in a full-bridge which reduces temperatures and improves reliability.
-Richie,
Re: What is better for DRSSTC: full-bridge or double half bridge?
Sulaiman, Tue Feb 23 2010, 04:48PM
Full bridge requires less capacitance across the dc bus for the same ripple/hummmm.
The caps are not cheap.
Sulaiman, Tue Feb 23 2010, 04:48PM
Full bridge requires less capacitance across the dc bus for the same ripple/hummmm.
The caps are not cheap.
Re: What is better for DRSSTC: full-bridge or double half bridge?
BSVi, Tue Feb 23 2010, 06:04PM
>..because you develope twice the peak-to-peak voltage
Yeap, but you can double voltage with double current halfbridge too - all you have to do is increase "on" time on your interrupter.
Anouther way to double voltage is to decrease capacitance and increase inductance.
As from my point of view all difference beetwen dullbrigde and double halfbridge is ringup slope (which is doubled in FB design) and maximum current (which is doubled in HB design).
BSVi, Tue Feb 23 2010, 06:04PM
>..because you develope twice the peak-to-peak voltage
Yeap, but you can double voltage with double current halfbridge too - all you have to do is increase "on" time on your interrupter.
Anouther way to double voltage is to decrease capacitance and increase inductance.
As from my point of view all difference beetwen dullbrigde and double halfbridge is ringup slope (which is doubled in FB design) and maximum current (which is doubled in HB design).
Re: What is better for DRSSTC: full-bridge or double half bridge?
Steve Conner, Tue Feb 23 2010, 06:36PM
That's not really true.
Doubling the on-time doesn't double the voltage, because the spark loading limits the ringup.
If you took a halfbridge DRSSTC, and replaced the halfbridge with a fullbridge without changing anything else, then because of the extra drive voltage, the primary current would increase and the sparks would get bigger.
If you paralleled extra devices in the halfbridge, the spark length wouldn't change, and you'd have to decrease the number of primary turns and increase the tank capacitance to get the benefit of the extra devices.
Steve Conner, Tue Feb 23 2010, 06:36PM
That's not really true.
Doubling the on-time doesn't double the voltage, because the spark loading limits the ringup.
If you took a halfbridge DRSSTC, and replaced the halfbridge with a fullbridge without changing anything else, then because of the extra drive voltage, the primary current would increase and the sparks would get bigger.
If you paralleled extra devices in the halfbridge, the spark length wouldn't change, and you'd have to decrease the number of primary turns and increase the tank capacitance to get the benefit of the extra devices.
Re: What is better for DRSSTC: full-bridge or double half bridge?
101111, Wed Feb 24 2010, 09:25AM
For those who have never seen.. What has once been seen can never be unseen! ( )... (not it is not anything nasty, quite the opposite)
101111, Wed Feb 24 2010, 09:25AM
For those who have never seen.. What has once been seen can never be unseen! (
)... (not it is not anything nasty, quite the opposite)Re: What is better for DRSSTC: full-bridge or double half bridge?
BSVi, Wed Feb 24 2010, 03:01PM
>and you'd have to decrease the number of primary turns
Ok, I expressed my question incorrectly.
Lets imagine that we have two DRSSTC's with optimal primary/capacitor bank combination. First DRSSTC is driven by full brigde, and second one by halfbridge with paralleled devices.
Which of them is better?
BSVi, Wed Feb 24 2010, 03:01PM
>and you'd have to decrease the number of primary turns
Ok, I expressed my question incorrectly.
Lets imagine that we have two DRSSTC's with optimal primary/capacitor bank combination. First DRSSTC is driven by full brigde, and second one by halfbridge with paralleled devices.
Which of them is better?
Re: What is better for DRSSTC: full-bridge or double half bridge?
Steve Conner, Wed Feb 24 2010, 05:23PM
Neither is better, they both perform the same because you reoptimized the primary coil.
In practice, the fullbridge is slightly better for the following reasons:
You don't have to worry about current sharing of parallel IGBTs. In theory they will share current, as was said, but it's VERY easy to destroy the sharing with a bad layout that adds too much inductance in the emitter connections.
A higher drive voltage makes it somewhat more convenient to optimize the primary coil and capacitor bank.
And if you tried to build a very large DRSSTC with a halfbridge, the optimal primary might end up less than one turn, which is impossible.
Steve Conner, Wed Feb 24 2010, 05:23PM
Neither is better, they both perform the same because you reoptimized the primary coil.
In practice, the fullbridge is slightly better for the following reasons:
You don't have to worry about current sharing of parallel IGBTs. In theory they will share current, as was said, but it's VERY easy to destroy the sharing with a bad layout that adds too much inductance in the emitter connections.
A higher drive voltage makes it somewhat more convenient to optimize the primary coil and capacitor bank.
And if you tried to build a very large DRSSTC with a halfbridge, the optimal primary might end up less than one turn, which is impossible.
Re: What is better for DRSSTC: full-bridge or double half bridge?
BSVi, Wed Feb 24 2010, 07:25PM
Steve McConner, thank you! That's what I wanted to hear :)
BSVi, Wed Feb 24 2010, 07:25PM
Steve McConner, thank you! That's what I wanted to hear :)
Print this page