F5 power amplifier

Jump to Latest
Sangram

Sangram

Moderator
Joined 2002
Paid Member
dbis said:
I'm using 10R 1W for R3 and R4 per attached F5 v3 BOM. I'm using the schematic from Nelson's F5 Turbo article. It's the first one in the article, "...official DIY F5 schematic". I hope those two are correct.
Click to expand...

Yes, 1W is the correct rating.

The supplies are 24V and the feedback resistors together dissipate around 6.5W peak, as the maximum output voltage cannot exceed 19.5V (4.5V is lost in the gate bias). In actual fact with music signals the dissipation will be significantly lower, but still not at the level a 1/4W unit can withstand.

A 1/2W unit is fine for lower listening levels, because music signals have a much higher peak-to-average ratio than sine waves and cause lower dissipation in those parts. Unless you're clipping the amp continuously or running sine wave tests, a 1/2W unit is adequate for 99% of users.
 
juma

juma

Member
Joined 2005
sangram said:
... maximum output voltage cannot exceed 19.5V (4.5V is lost in the gate bias). ...
Click to expand...
Not in the case of F5 since it uses output MOSFETs in Common Source topology (the output is taken from the Drains). In F5 you lose output swing on internal resistance of MOSFETs (negligible for IRFp240/9240) and MOSFETs' Source resistors (Id * 0R47 - not negligible).

Output swing is lost on Vgs ("gate bias") in Source Follower output stages (F4 for example).
 
Last edited:
A

AndrewT

R.I.P.
Joined 2004
AndrewT said:
music at an average of 20dB below clipping will reduce the dissipation in the feedback resistor down to 1% of the nearly clipping level.
Click to expand...

Nelson Pass said:
That would be the peak figure. With actual music, well within ratings.
Click to expand...
There is no doubt that replaying music is OK.

The Pmax doubts are being raised when test signals are being reproduced.

earlier I posted this
AndrewT said:
R3 & 4 are in series with the 50r NFB resistors.
The current through these is ~ Vpk/(50+10) = 30Vpk/60 = 0.5Apk.
The peak instantaneous dissipation will be 0.5Apk² * 10ohms = 2.5W
the average will be half of this if the output stays that high for long enough and the duty cycle is 50% so the average dissipation if testing with sinewave to 30Vpk (56W into 8r0) would be 0.625W
That is too much for a 600mW resistor.
But there is a recommendation (by a few) that the NFB resistor/s should not be exposed to dissipations exceeding 10% of their rating.
Ordinary VFA amplifiers can meet this requirement fairly easily. Not CFA and not the F5.
Click to expand...
CFA style amplifiers and the F5 behave differently from VFA.
Trying to apply a 10times factor to get a resistor Pmax would be somewhat silly for the F5 and variants.
It seems that we do need very good resistors in the feedback route to minimimise the thermally induced distortions. And need to take care with test procedure to ensure we don't damage these very good resistors.
Suzyj posted some results of testing comparisons for different Rupper in the feedback path. Illuminating to say the least. But these NFB resistors have been subject to this kind of scrutiny for decades. I became aware of it a long tme ago, maybe before I joined this Forum in 2004 and I am not an amplifier designer.
 
Last edited:
S

soldersmoker

Member
Joined 2005
I've got an oddball question: Will the F5 or F5T sound good with Sonus faber Chameleon B's? If anyone has direct experience with that combo, I'd appreciate learning what you think. The Chameleons are about 87db/2.83V/m at 4 ohms impedance. I don't listen to my music at loud volumes and my listening space is a typical American living room. I know it's probably not what most people would consider an ideal match. But . . .

I have a pair of Pass Amp Camp Amps that I like to use with a set of sensitive speakers (Large Advent Econowaves), but I really like how my Chameleons sound with a Yamaha A-S501. I have heard the F5 powering a pair of Altec 605s and that combination sounded sublime, and I'm very much aware of the radical differences between the 605s and my speakers. Mainly I'm wondering if the sweetness of more class A would raise the sound quality a notch or two.
 
B

brunogiovs

Member
Joined 2014
Hello Guys

Any opinions about non-inductive wirewound resistors for R9-12 and for R3-4?

I kind of decided to go with a 2w non-inductive wirewound for R3-4.
but for R9-12 I'm between a 5w non-inductive vs a regular 7w wirewound.

Anybody tried something different ? I noticed that on the pcb v3 there's holes for a T0-220 resistor, maybe somebody tried metal foil resistors?

Thanks in advance.
 
Last edited:
fgroen

fgroen

Member
Joined 2005
Hi guys,

Just got hooked into building this amp, already put the schematic into KiCad to get some PCB's manufactured.

Obviously, the hunt for parts is on. What got me wondering is whether there would be some more current alternatives out there that could do the job.

For the output MOSFETs (IRFP9240 and IRFP240) newer types with the addition "PBF" seem to have the exact same specs as the recommended ones. IRFP9240PBF and IRFP240PBF are readily available at Mouser.

Another story are in input devices. For the 2SK170BL and 2SJ74BL no obvious current alternative presents itself. As the transconductance is key here it means browsing through a lot of data sheets. Has anybody looked into this matter and now of some current devices that could replace the specified ones?
 
You must log in or register to reply here.
Top Bottom