Nothing in particular.
But a hi-end manufacturer uses it everywhere, in combination with the 36P15.
So must be that they are idiots ?
Patrick
But a hi-end manufacturer uses it everywhere, in combination with the 36P15.
So must be that they are idiots ?
Patrick
An odd comment.
I was not aware it was popular, and I am the last to assume other manufacturers are idiots.
Perhaps it was cost? Or Ciss? Or gm?
HD
I was not aware it was popular, and I am the last to assume other manufacturers are idiots.
Perhaps it was cost? Or Ciss? Or gm?
HD
"The FQA36P15 and FQA28N15 seem to have very close transfer characteristics, while the FQA46N15 has very close gate characteristics with the FQA36P15."
Thanks egra.
F5 power amplifier
Woud you happen to know of any measurements to back up that statement ?
Patrick
F5 power amplifier
Woud you happen to know of any measurements to back up that statement ?
Patrick
Not yet. I quoted this statement from opc. However Tvicol uses FQA46N15 and FQA36P15 pair in his Quad amplifier successfully, so I decided I will use these too, the Cg is high for both FETs, but almost identical (measured). In both side I'll use a 120 ohm gate resistor.
HexFet Amp Based on Philips AH578 and LMK
HexFet Amp Based on Philips AH578 and LMK
I haven't tried FQA fets yet. This will be the first amp with these when the PCBs finally arrive. What kind of experience do you have with FQAs?
Nelson Pass used FQAs in his original F5, of which I have also built one proto.
I also used FQPs in the F5 headphone amp.
Otherwise I only use Toshiba's.
But I am open to take a look at Fairchild (Onsemi) these days.
As well as laterals.
Patrick
I also used FQPs in the F5 headphone amp.
Otherwise I only use Toshiba's.
But I am open to take a look at Fairchild (Onsemi) these days.
As well as laterals.
Patrick
i have FQA40N25 and FQA36P15 for five years in Class AB amps with excellent results, and in fact 480W mosfets as well from IXYS.
I have had very few failures; those sheet home to shorting and other misadventures. The gates are tough to drive, so I run low impedance, high current drivers. I do think that larger mosfets sound different to bipolars; generally a little less resolving but it's arguable.
I have not tried laterals yet but they are on the list.
HD
I have had very few failures; those sheet home to shorting and other misadventures. The gates are tough to drive, so I run low impedance, high current drivers. I do think that larger mosfets sound different to bipolars; generally a little less resolving but it's arguable.
I have not tried laterals yet but they are on the list.
HD
Hugh, you wrote this somewhere: "Nigel, Most of the time there will be no oscillation without snubbers. However, in a day, a week, a year, it will appear and destroy the device......" According to the simulation, the GD snubber significantly reduces the phase margin. Is it worth using snubber for a 36v supply voltage and with a single pair FETs? Thank you.
I really like LATFETs. Stable and reliable devices. They have a bit more extensive and warmer hights than HEXFETs. Very easy to listen to them.
The high-end manufacturer that uses has nothing to say, I've seen quite a few adventurous high-end circuits, so I couldn't find the words.
Just read the service manual with technical expertise from Ecler for Mosfet P + N and only N-Channel, Mosfet power amplifiers, then you will see how faulty some circuits from HIGH END manufacturers are.
Thanks for your comments.🙂
That was a duplicated 2.2u cap - i will remove it and replace the other one with 220n.
I will also check and if necessary spread the gate protection diodes so a hex spanner can fit between them. The height of those diodes should be lower than the Fet bolts anyway but a few more mm clearance won't harm. That remaining 2u2 cap also needs to be moved to provide more clearance but again, the height should be lower than the Fet bolts.
PS: Sprint Layout tips: The zone or rectangle tools are easier to use to create copper areas instead of a network of tracks. Alt-clicking pads and then selecting through hole is the recommended way to generate those pads instead of placing through hole pads over top of existing ones. The latter creates DRC errors.😀
You're right Patrick. In the meantime I remembered that there are some FQA28N15 in my stock so I measured them too. The difference between Cgs is significant, so the gate resistors must be changed to 100 ohms on the P side and 200 Ohms on the N side. Many consider this N fet to be the most ideal one for the FQA36P15 because of their similar transfer characteristics.
Attachments
