How high efficiency is possible with class D. I saw a sales brochure claiming 97% efficiency in a range of 100W - 500W Public Address amplifiers. I have designed both class D amps and switching power supplies, and my first reaction was that this was a load of hogwash.
Is it at all possible to achieve this kind of efficiency in an amplifier?
I mean, I would have to struggle hard to achieve this kind of efficiency in any kind of power supply!
Is it at all possible to achieve this kind of efficiency in an amplifier?
I mean, I would have to struggle hard to achieve this kind of efficiency in any kind of power supply!
i've at moment 97% H-bridge 400VDC 100KHz 300W, but after replacing the poor IRF mosfets to new ST, i'll wait >98% and at 600-700w. So even 99% for low frequence switching and for half bridge isn't impossible.😎
I reach it, but you should use APT or IXYS fast and low RDSon mosfet, with low lose rectifier, etc....Impossible to reach with ''commercial'' mosfet, you have to move to industrial parts.
Fredos
www.d-amp.com
Fredos
www.d-amp.com
Fredos can reach at least 100%. 😀
In a complete amplifier (power connector->filter->rectifier->PFC->DC/AC->trafo->rectifier->choke,capacitor->bridge->filter, plus fan, modulator, signal processor circuitry, etc...) I can't believe such efficiency. In the power modul, it's OK.
What kind of amplifier is this? (400V 0,8A?)
Be careful! 3% loss or 1% loss is a huge difference! Don't underestimate the power of switching loss! 🙂 (Unless you use ZVS)
Pafi said:
There's a thread "Free Energy Devices" that shows a device said to generate more energy given it. 😀
fredos, why only APT or IXYZ? Infineon presented month ago 500v 130mOhm 48nC, not so bad thing. BTW, how to obtain samples from Infineon?
290nC...Total gate charge...Nothing impressive, specialy the 120ns turn-off...Far from APT or IXYS!
Fredos
Fredos
IVX / BWRX: " ... BTW, how to obtain samples from Infineon? ..."
Infineon buys stuff, they never sell stuff. Check out their web site = filled with "carreer opportunities" and new supplier information ... meaning they are in expansion mode. Their customers appear to be big companies with long term projects ... or race tracks and car shows. (As Forbes.com would say: "short the stock.")
😕
Infineon buys stuff, they never sell stuff. Check out their web site = filled with "carreer opportunities" and new supplier information ... meaning they are in expansion mode. Their customers appear to be big companies with long term projects ... or race tracks and car shows. (As Forbes.com would say: "short the stock.")
😕
record efficiency?
Class-D is a new breed of cat. Just as flouresent bulbs are more efficient than incandecent bulbs, "switching" pulse width modulation amps are more efficient than "brute force" or current dump amps (type A or AB). The over all thermal efficiencies tells the true nature of Class-D, being as they deliver most of the power to the load instead of consuming a significant amount of power as heat. The number being bandied about for "efficiency" of Class-D are generally erronious. The numbers to look for for comparison should be thermal efficiencies. Modern motor speed control using pulse width modulation is significantly more thermally efficient than simple variable resistance control (same, same w/ Class-D) ... that's why the EPA & the EU want to see more of 'em = less heat generated = more environmentally correct = less power consumed for the job.
Put your hand on Class-A or Class-A/B amp heat sinks after several hours of continuous operation
... then try the same with a Class-D amp of the same power, relatively same heat sinks == 😎
A well designed pulse width modulation amp (Class-D) should have thermal efficiencies greater than 85% or more ... over all thermal efficiency ... from power plug to speaker terminals.
(Also that 6.3 volt filiment in a vacuum tube is called a heater .. for good reason ... )
The true tests should be the specs like THD ... and the listening experience, not the efficiencies. Unless you want sodium vapor lighting in your living room, there is still no substitute for an incandecent bulb to read by IMOP.
Class-D is a new breed of cat. Just as flouresent bulbs are more efficient than incandecent bulbs, "switching" pulse width modulation amps are more efficient than "brute force" or current dump amps (type A or AB). The over all thermal efficiencies tells the true nature of Class-D, being as they deliver most of the power to the load instead of consuming a significant amount of power as heat. The number being bandied about for "efficiency" of Class-D are generally erronious. The numbers to look for for comparison should be thermal efficiencies. Modern motor speed control using pulse width modulation is significantly more thermally efficient than simple variable resistance control (same, same w/ Class-D) ... that's why the EPA & the EU want to see more of 'em = less heat generated = more environmentally correct = less power consumed for the job.
Put your hand on Class-A or Class-A/B amp heat sinks after several hours of continuous operation
... then try the same with a Class-D amp of the same power, relatively same heat sinks == 😎A well designed pulse width modulation amp (Class-D) should have thermal efficiencies greater than 85% or more ... over all thermal efficiency ... from power plug to speaker terminals.
(Also that 6.3 volt filiment in a vacuum tube is called a heater .. for good reason ... )
The true tests should be the specs like THD ... and the listening experience, not the efficiencies. Unless you want sodium vapor lighting in your living room, there is still no substitute for an incandecent bulb to read by IMOP.
I've got a 95% efficient amplifier handy.
It's an H-bridge used to drive a ginormous motor.
Hey, you never said it had to be able to produce frequencies above 20hz!
It's an H-bridge used to drive a ginormous motor.
Hey, you never said it had to be able to produce frequencies above 20hz!
Easy - a commercial mosfet is one used in commercial products while an industrial mosfet is one used in industrial products 😉
Seriously, he's mainly talking about devices used for different power levels. "commercial" devices being the ones used in "lower" power applications and "industrial" devices being the ones used in "higher" power applications.
Seriously, he's mainly talking about devices used for different power levels. "commercial" devices being the ones used in "lower" power applications and "industrial" devices being the ones used in "higher" power applications.
What I call a commercial mosfet is something like Fairchild, IRF, MOSPEC, Infineon, low power series of ST. A fews $ mosfet in other word. What I call Industrial mosfet is APT, IXYS, big ST mosfet, a mosfet that you can push to 150% of it capacity without fear of faillure! But maybe 10 time the price of an commercial mosfet!
Fredos
Fredos
There are really interesting Infineon parts readily available through Farnell in Europe, like a 600V 34A MOSFET with ultrafast body diode, 0.1 ohm typ. Rds-on and 163nC typ. total gate charge. Not to mention that this is the one I'm using in the last version of my +/-200V +/-35A LF class-D amplifier (with great success BTW).
On the other hand, after browsing several dozen datasheets, I was unable to find any device worth even trying from IXYS or APT. I can almost guarantee that at least all the MOSFET stuff from those manufacturers available through Farnell is useless and extremely expensive. Am I missing something or what?
On the other hand, after browsing several dozen datasheets, I was unable to find any device worth even trying from IXYS or APT. I can almost guarantee that at least all the MOSFET stuff from those manufacturers available through Farnell is useless and extremely expensive. Am I missing something or what?
fredos, topic is "Record efficiency", but not "Record output power", however, 23A at 400VDC H-bridge give several KW, obviously. I guess, tales like this -"really cool can be APT/IXYS ONLY", actually just a popular myth. Commercial IPP50R140C3 from Infineon, outperforms any of the comparable APT "COOLMOS", IXYS will offer pretty good mosfet soon, but at moment many interesting specs is "tbd" (see http://ixdev.ixys.com/DataSheet/f7f3bb14-c5ae-434e-a188-022b16bbb606.pdf). Maybe "commercial"/"industrial" just a cost ratio?
fredos said:
How do you manage to push MOSFET dissipation (or even drain current) 150% above rated while still keeping 99% efficiency? I may be missing something again...
I don't mind at all whether my switching MOSFETs or IGBTs are rated at 75W like low cost TO-220 parts or at 500W dissipation like ultra-expensive IXYS and APT stuff. They are rarely subject to more than 25W in my circuits.
- Status
- Not open for further replies.