F5 power amplifier

[jackinnj]

Originally posted by pro
jackinnj

try 1u or higher, it do the job.

I've seen 47uF on a design note from nat semi, and 200pF on another design note.

Care and feeding of MOSFET gates --

[AndrewT]

Originally posted by AndrewT
Tini,
The bias is set. It remains unaltered, if the temperature compensation is accurate, irrespective of the load impedance.

Originally posted by tinitus


So, you are saying that into 4ohm, idle bias will be only half (0.6A)
A stereo amp will consume only 30watt pr channel, instead of 60watt pr channel, at idle bias

This situation would mean that one could use smaller heatsinks, for that particular use
And if staying within the classA limit, smaller trafos as well

I guess you are right
I dont question that,

the answers are given to you in various forms/words and still you won't read.

[npapp]

Originally posted by AndrewT


and yes if one chose 100dB/W/m speakers to go with an F5 then you will have higher SPLs available then if one, somewhat stupidly, selected 85dB/W/m speakers to go with the F5s

stupidly like a fox: my mains are 91db and what I am using now... but when I was testing this amp early on I used 85.5 db/w/m speakers. I was able to drive the amp directly from my DAC. One straight wire with no knobs. The combo filled my large room beautifully, but the wife missed the knob


Andrew, thanks for the thoughts on how much output I may actually need for the transients.

Garcia and Grisman are pick'n through my F5

[bobodioulasso]

Originally posted by jackinnj


I've seen 47uF on a design note from nat semi, and 200pF on another design note.

Care and feeding of MOSFET gates --

My (bjt) ES200 Audiolabor (Wachowiak) has 1000uf between bases.

[Bill Fuss]

Originally posted by tinitus


So, you are saying that into 4ohm, idle bias will be only half (0.6A)
A stereo amp will consume only 30watt pr channel, instead of 60watt pr channel, at idle bias

This situation would mean that one could use smaller heatsinks, for that particular use
And if staying within the classA limit, smaller trafos as well

I guess you are right
I dont question that, but sounds strange to me

We can break this down to make it simpler to understand.
Let's look at only half of the amp. Your rail voltage is set at around 24V so if you have a 4v drop across the fet you have 20v available at the output. The current available is set by the bias current of 1.3A, and that is set by the source resistor. You now have two limiting factors in your circuit, your maximum output voltage is fixed, as is your maximum current. To understand this you must forget about the rest of the amp and just consider this as an electrical circuit and not a part of an audio amp.
The optimum load for max power is easily found, E(20v)/ I(1.3a) = 15.4 ohms, easy enough. Max power is 26Watts
If we use a 20 ohm load then we have 20v/20ohms = 1A which means there is a reserve of current still available but we are at the limit of available output voltage. Max power is 20 Watts
If we use a 10ohm load then we have 20v/10ohms= 2A, but we only have 1.3A available so we are now current limited. Max power is about 17 Watts
This is a basic electrical circuit, please do not try to read anything else into it.

Best, Bill

[tinitus]

Well, in that case

Originally posted by Nelson Pass


25 watts Class A into 8 ohms implies a 1.25 A bias current.

I understand that a load into half the impedance means only half the voltage is needed
But is that also equal to whats available
Im really trying to understand your point, but dont
This is really not my strong side, and has nothing to do with "not reading"
But I have always thought that current would rise in low impedance load

[Bill Fuss]

What does that have to do with what I posted? You have to learn to walk before you learn to run.

[tinitus]

Its obvious we cannot expect 50watt classA into 4ohm
Perfectly clear

Though to me it will stay 25watt classA, no matter what you say about that
Bias is fixed, supply voltage is fixed
To me its simple logic
And I will build with the suggested trafo rating requested, "more than 10A peak"
Banging my head with better knowing has never worked on me

Obviously I will never understand your calculations
Im even more happy now that all information is easily available, in clear text

[Nelson Pass]

50 watts into 4 ohm means a 2.5 amp bias current, which is 50
watts dissipation per Mosfet. If you have adequate hardware
you can do it.

[bobodioulasso]

Hi Tinitus

Staying in class A: two times the bias current in the load is the limit.

12.5w into 16 ohms / bias 0.625A
25w into 8 ohms / bias 1.25A
50w into 4 ohms / bias 2.5A

At 1.25 bias and 4 ohms load we leave class A at 12.5w
when load current reaches 2times the bias current

P peak = 2Prms.

The only formula is P=RxI^2

Is that clear for you?