in other words - diodes are there to make Klonk!! here and there across/for poor resistors , decreasing their Hogging! (or - Behave!! ) role for nasty mosfet's nature .....
luckily - they're on both sides , so that sounds more like Klonk!!knolK , so less objectionable .............
we can call them - Unhogging! or Unbehave!! parts ......
conclusion - diodes in F5T are same as Jaccolina on DiyA
Attachments
I feel better now that i know that i was not the only person who didnt completely understand what was happening. Apperently everyone here is an engineer in diguise and i am one man proletariat.
When one side is switched off or not conducting anymore
Less than mosfets treshold voltages on gate.
I think
So pick your rail voltages find out what bias your sink let you have
Find out your normal listening level
Pano tread may help to find out your normall volume levels.
set up the source resistors at there about to get out of class A when you are having party levels. otr pick best compromise
I normaly have 2 or 3 places where volume pot alwaise land.
Serious listening level when I Just there to enjoy my music is normaly lower than the level I use while doing dishes ironing dusting and cooking which is lower than when I am having party
Pick best compromise that suits you.
Main thing that finaly sunk in is. one do not have to use 0.50000 Homs
Or my personal opinion and just that. 0.5 homes is prety good value for 62 V rails
And another thing
The sorce resistor is also there to stop one of the more exuberant Mosfets from current hogghin so if you match your sets prety well that can be smaller.
May even let you get all the currants one need before output flatens out due to clipping
I think so
but on What rail voltages? and are your sink able to take it?
There are a number of question to answer before one chanages Papa specs
and keep things reasonably safe.
F5 was prety much right on the ball.
F5 turbos is Papa answer to our greedy egos.
I fell that Papa wanted us to go walk about and find things out by ourselves.
First of all is how far one feel is safe to push thermal load on mosfets.
second can you get heat out of mosfets properly or is your mica and goop working ?
Can you find out how hot the mosfets realy get by mesuring themperature on drain pin (this is in direct contact with the metal tab on thee back)
One may have cool sinks but the mosfet is boiling away if this is not done properly.
Third can your sink take it?
fourth do you realy need 2 or what output do you need out of the amplifier and speakers?
I'm looking at building a balanced F5T with 24V rails. I'm targeting 50W Class A into 4ohms and wanting to leave some headroom (hence the little higher rail voltages)
I would parallel a pair of N and P channel FETs on each half of the balanced amp (8 total FETs per channel). I would use (4) Conrad MF25-151 heatsinks per channel. With two FETs per sink the 65W dissipation should give roughly a 26 degree rise (0.29C/W x 1.4 x 65W). Seems like a nice safe temp range to me.
I would parallel a pair of N and P channel FETs on each half of the balanced amp (8 total FETs per channel). I would use (4) Conrad MF25-151 heatsinks per channel. With two FETs per sink the 65W dissipation should give roughly a 26 degree rise (0.29C/W x 1.4 x 65W). Seems like a nice safe temp range to me.
Since you are basically bridging two amps together, you don't need as high of rails. Double the power into double the load is what I always remember as the rule of thumb.
Are you saying there is no real benefit to using the diodes with 24V rails?
if you want 50W of ClassA into 4r0 from the balanced amplifier, then go back to the most basic rule of balanced and bridged amps.
A bridged/balanced amp gives double the power into double the impedance.
Therefore, aim to bias each half of the amplifier to do 25W into 2r0.
That requires a bias current of 3.6Adc in each half of the amplifier.
The PSU will see a demand of 7.2Adc on each supply rail.
On 24Vdc supply rails, everything is going to get very hot.
Have a rethink.
A bridged/balanced amp gives double the power into double the impedance.
Therefore, aim to bias each half of the amplifier to do 25W into 2r0.
That requires a bias current of 3.6Adc in each half of the amplifier.
The PSU will see a demand of 7.2Adc on each supply rail.
On 24Vdc supply rails, everything is going to get very hot.
Have a rethink.
Thanks for that analysis Andrew.
Correct me if I am wrong, but a bias current of 3.6A per half of a balanced amplifier would give you 50W of Class A peak? If you wanted 50W RMS, wouldn't you have to up the bias current to 5Adc per half? I know with 24V rails this are going to get toasty. I've thought about lowering them down, but I like the idea of some Class AB headroom.
Correct me if I am wrong, but a bias current of 3.6A per half of a balanced amplifier would give you 50W of Class A peak? If you wanted 50W RMS, wouldn't you have to up the bias current to 5Adc per half? I know with 24V rails this are going to get toasty. I've thought about lowering them down, but I like the idea of some Class AB headroom.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.