Record efficiency?

Status
Not open for further replies.
Jump to Latest
#42
Yes, I am also troubled by you leaving this Forum.

I notice that you have gone elsewhere and you are already discussing the opportunity of parallel operation of power stages.

Please come back and explain how this should be done.

Then we can get Eva to build some for us.

DNA
 
#44
About precision of the efficiency measurement, it is really important if true RMS V/A meter in use (even for design with PFC), otherwise, with some type of multimeters, >100% we can get yet. 🙂 Funny, efficiency of the my regenerator circuit (input =230VAC@THD=3% to output =AC230VAC@THD=.00_%), obtained by the cheap multimeter = 98%+@500W, but it's two convertors -PFC+H-bridge! So, if that multimeter have quasy peak nature of measuring, the error can be up to 3%(input AC)+~1..2% (PFC current THD). I'll try my PCI oscilloscope as true RMS meter, i guess, that 10bit precision enough.
 
#46
How about the challenge of high'ish efficiency for a very low 0.15 ohm resistive load ribbon speaker?

I went to the Infineon site and noted that their Optimos3 IC process has a 30V, 100A N-MOSFET with 1.6 mohms of resistance. BSC016NO3LS G It does have 10,000pF of input capacitance, but the larger die area offers 1 C/W thermal resistance, while smaller MOSFETs with 3mohm resistance and 7,000pF input C have 1.3 C/W.

Can this team figure out how to drive it for good sound?

So, what size of output inductor would this need for 20 watts with a 430Khz TI xx5261 digital amp chip? All digital up to the speaker would be the goal, even with the non-feedback TI signal path.


http://www.infineon.com/upload/Document/BSC016N03LS_rev1.0.pdf


http://www.infineon.com/cgi-bin/ifx...age=/ep/program/typelist.jsp&pageTypeId=17099
 
#48
Drive it with TC442* serie of 9A mosfet driver...And use a hight saturation MPP output coil in ZVS topology ( output switch open in syncronisation with recovery of the output coil..), you will be able to achieve low switching lose for your load....And recovery will not be a probleme!

Fredos
 
#49
Diode recovery is not a problem for very low voltage MOSFETs, check diode reverse recovery charge in datasheets. Also, if gate capacitance becomes too large, resonant gate drive may be employed to reduce driver current and dissipation requirements considerably with a fixed switching delay as the only disadvantage. Brute force resistive high current gate drive is not always the best alternative.
 
#51
IVX said:
BTW, what kind/model of the ribbon speaker, has 150mOhm, and how it's usually driven, stepdown transformer?
Click to expand...


Hi IVX,

I have Apogee Full Range with a 84" x 2" ceramic magnet midrange ribbon and DIY with a 94" x 3" NdFeB magnet midrange ribbon, both with about 0.15 ohm resistive load. I have 1 ohm and 4 ohm step up transformers which show significant phase shift on a vector impedance bridge. I also used a non-inductive 1 ohm series resistor to get a load which my Krells can easily drive to far superior sound than with the transformers. I have designed a DIY 32 watt Class-A amp that can direct drive 0.15 ohms, but it burns 200 watts under full bias with 16 big Sanken output transistors. I also need a 14 opamp analog crossover to set up LR8 slopes, and their phase shifts and lack of room equalization bothers me.

SO

My next educational goal is for a full digital design from the CD/DVD player to the speakers. I was hoping a Class-D expert would spend some engineering to point me in the right technical directions, especially the output stage passive filters. Looks like some specialized inductors are required. The all digital path like the 432kHz TI xx5261 digital amp has some appeal to me, as it maintains a full digital path up to the speakers. This would allow me to build up a digital Xover and room equalization. I understand that there is no feedback loop with a TI xx5261, and that a robust power supply is required. If I can't find an all digital solution, I feel that I might as well just stay with my Class-A amps and analog crossovers. A hi/lo Class-A bias switch gives me some global warming savings.

The 1.6 m-ohm Infineon MOSFET I tagged looked like a pretty good starting point.
 
Status
Not open for further replies.
Top Bottom