Unfortunately, values given will be theoretical as there is no load values given. However, as each FET will conduct 13A continuously or 80A switching. Take your pick using Ohms Law.
Have a look at the spec sheet; http://www.vishay.com/docs/91237/91237.pdf
Have a look at the spec sheet; http://www.vishay.com/docs/91237/91237.pdf
Well, yes, you are right, it should be possible to get higher power than 1000W - 1500W. I measured the ON Power, and it is PDon = 15A x 0.5V = 7,5W. I didn't measure the switching losses but I am sure is not too much. Anyway I should measure the case temperature, but I am sure is not hotter than 55°C. I'd like to hear real experiences about how much power have people got.
yes 4000 Watt should be easy in controlled lab test
so what happens? , the lights go out. LOL
measured Vds and Id waveform would be useful ( in case of device failures )
what loads and output voltage ( check source and load capabilities 1st)
active loads? tested branch breakers? etc
Last edited:
Oughhhh....I´m far from 4kW!! With a conventional hard switch PWM Full Bridge??? Ok, I can measure Vds and Id with a home made current probe. Actually works fine, so I can trust of it.
Output voltage: 130V.
Output Current: Max I got: 10A
Active loads??? What is that? I just put nichrom wire home made resistances for load.
I made a case temperature measurement last night of the mosfets cases. They are 60° more or less.
What about the output rectifier? Two diodes or four diodes??
Output voltage: 130V.
Output Current: Max I got: 10A
Active loads??? What is that? I just put nichrom wire home made resistances for load.
I made a case temperature measurement last night of the mosfets cases. They are 60° more or less.
What about the output rectifier? Two diodes or four diodes??
I'll made measurements of case temperature on 1000w of power, but I guess that is the problem, mosfets are overheating...we'll see..
Infinia: I was thinking about yours 4000w....suppose your smps has a efficiency of 80%. So you have 800W dissipating on your smps. Imagine that in the better case, the 50% of 800W is dissipated on the mosfets, so every mosfet dissipates 100W (conduction, switching and drive losses). How can mantain a low temperature??? Is that power continuous??
Infinia: I was thinking about yours 4000w....suppose your smps has a efficiency of 80%. So you have 800W dissipating on your smps. Imagine that in the better case, the 50% of 800W is dissipated on the mosfets, so every mosfet dissipates 100W (conduction, switching and drive losses). How can mantain a low temperature??? Is that power continuous??
older CPUs dissipated 100 Watts, not a problem for MOSFETs either
active loads or electronic loads
I don't think 4KW is a realistic goal for a conventional hard switching SMPS maybe 1500 is practical limits. then you spend all your time chasing EMI.
I think the biggest one I worked was around 5V/160 Amperes
AN1114 excerpt
In the full-bridge converter, four switches have been
used, thereby increasing the amount of switching
device loss. For applications requiring output power of
more than 1000 watts, the loss in the switching device
becomes impractical to handle in a full-bridge
converter.
The conduction loss of a MOSFET can be reduced by
using a good MOSFET, and switching losses can be
reduced by using either a ZVS (zero voltage switching
during turn ON transition), a ZCS (zero current
switching during turn OFF transition), or both
techniques. Shaping the input current sinusoidal to
achieve ZCS, increases the peak and the RMS current
through the MOSFET in the high power application,
thereby increasing the conduction losses. At high input
voltage, the ZVS technique is preferred for the
MOSFET.
active loads or electronic loads
I don't think 4KW is a realistic goal for a conventional hard switching SMPS maybe 1500 is practical limits. then you spend all your time chasing EMI.
I think the biggest one I worked was around 5V/160 Amperes
AN1114 excerpt
In the full-bridge converter, four switches have been
used, thereby increasing the amount of switching
device loss. For applications requiring output power of
more than 1000 watts, the loss in the switching device
becomes impractical to handle in a full-bridge
converter.
The conduction loss of a MOSFET can be reduced by
using a good MOSFET, and switching losses can be
reduced by using either a ZVS (zero voltage switching
during turn ON transition), a ZCS (zero current
switching during turn OFF transition), or both
techniques. Shaping the input current sinusoidal to
achieve ZCS, increases the peak and the RMS current
through the MOSFET in the high power application,
thereby increasing the conduction losses. At high input
voltage, the ZVS technique is preferred for the
MOSFET.
Last edited:
Infinia. Thanks for your honesty. That is useful for me. I am agree with the realistic limit. Anyway I have to work with the temperature of the mosfets, I guess there is the problem.
A couple of questions:
- 2-Diode or 4-Diode for output rectifier?
- RC Snubbers on Drain to Source mosfets?
- External fast diodes on Drain to Source?
- Snubbers on output rectifier diodes?
A couple of questions:
- 2-Diode or 4-Diode for output rectifier?
- RC Snubbers on Drain to Source mosfets?
- External fast diodes on Drain to Source?
- Snubbers on output rectifier diodes?
one snubber across the primary > maybe 5-10 watt non inductive
snubber across every secondary diode 2-5 watt
IDK maybe you need try soft recovery rectifiers (biggest EMI source since Vo is 130V )
edit> switching and snubber losses can be reduced by lowering switching frequency maybe make some improvements on transformer windings leakage too. ( interleave pri and sec shoot for 5 or more layers )
snubber across every secondary diode 2-5 watt
IDK maybe you need try soft recovery rectifiers (biggest EMI source since Vo is 130V )
edit> switching and snubber losses can be reduced by lowering switching frequency maybe make some improvements on transformer windings leakage too. ( interleave pri and sec shoot for 5 or more layers )
Last edited:
analyzing secondary rectifier current in full bridge is most interesting!
app note - energy contained in "snap off" region http://www.irf.com/technical-info/appnotes/an-989.pdf
app note - energy contained in "snap off" region http://www.irf.com/technical-info/appnotes/an-989.pdf
Well, I actually had to put a couple of snubber on every diode (MUR3040). Otherwise, I had a terrific current spike on the secondary, causing an ugly noise on transformer. But, that snubbers creates a current spikes on current mosfets 
I am having more troubles on secondary side than primary actually.
I'll read that AN soon. Thks!
I am having more troubles on secondary side than primary actually.
I'll read that AN soon. Thks!
Well, I can get the RHRP1540, and I was comparing with the MU3040 (that I have now in my smps), and the RHRP1540 has soft recovery characteristics and is a ultra fast, and MUR3040 is not. I will try it.
I actually didn't know about that, so I will try. Thks guys!
I will make measurements on temperature on the cases of mosfets too, I guess the swichting losses rises the temperatures and limits the top of power.
I actually didn't know about that, so I will try. Thks guys!
I will make measurements on temperature on the cases of mosfets too, I guess the swichting losses rises the temperatures and limits the top of power.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.