Some manufacturers hang the transformer from a strap across the width of the chassis.
This strap can be strong, as a beam, to support the weight of the transformer, even during abuse.
But the strap can have very low torsional stiffness, allowing the transformer to dangle and vibrate without transferring all the vibration to the Chassis.
I only did this for one amplifier using a very tall Chassis. Smoothing caps in the bottom. Toroid and rectifier just above. Then the Amplifier and other hot parts in the upper half.
Air entering the bottom of this ClassA amp was cool and kept both the capacitors and transformer relatively cooler compared to a wide low Chassis.
The radiation from the inside surfaces of the tall sinks would be considerable and I thought this "tall" arrangement would be better for reliability.
This strap can be strong, as a beam, to support the weight of the transformer, even during abuse.
But the strap can have very low torsional stiffness, allowing the transformer to dangle and vibrate without transferring all the vibration to the Chassis.
I only did this for one amplifier using a very tall Chassis. Smoothing caps in the bottom. Toroid and rectifier just above. Then the Amplifier and other hot parts in the upper half.
Air entering the bottom of this ClassA amp was cool and kept both the capacitors and transformer relatively cooler compared to a wide low Chassis.
The radiation from the inside surfaces of the tall sinks would be considerable and I thought this "tall" arrangement would be better for reliability.
Thanks for the help Guys, I shall have my sinks side by side 3 on each side this will be a very Big F5 turbo. 11.8" high 
No point stacking them tall if its going to waste heat-sinking abilities
Yes revision 3 of the Diyaudio boards look really good so I think that's the way forward.
Do you think 2x 300Va transformers would be ok ?
No point stacking them tall if its going to waste heat-sinking abilities
Yes revision 3 of the Diyaudio boards look really good so I think that's the way forward.
Do you think 2x 300Va transformers would be ok ?
Not my idea.
Copied from some manufacturer.
Better if the transformer were hanging from some springs with a damage limiting hook arrangement. Grommets do not work if they are the wrong frequency.
But do allow for the equipment being dropped on rare occasions.
potting compound is not runny enough.
Even removing the outer insulation, the potting compound will not penetrate all the air spaces in the outer windings. The potting compound has no hope of penetrating the next layer of insulation.
You would be wasting your time using potting compound to impregnate the toroid.
However, JohnH is reporting good results with encapsulation, i.e. using potting compound around the toroid while inside a retaining vessel.
That helps keep the noise inside the transformer.
Is it really about current?
How many watts do you really need to have on tap? My speakers are around 95dB efficient from bass to treble, and dip no lower than 8 ohms.
Amplifier Power Required Calculator
I'd think that the diodes are irrelevant.
How many watts do you really need to have on tap? My speakers are around 95dB efficient from bass to treble, and dip no lower than 8 ohms.
Amplifier Power Required Calculator
I'd think that the diodes are irrelevant.
I'm not using diodes, and my speakers are only 85dB. The impedance does not dip below 6 ohms and even a standard F5 did not run out of steam when driving them.
However, they did run out of voltage swing, so I am building a balanced version for myself (with a few changes).
You may want to think about a standard unit but using lower feedback if you want something different.
I am not sure that removing the gate resistor biases the MOSFET to zero Vgs, but an intermediary value based on the source resistance and power supply voltage - the so-called self bias. I think the results of removing all drive is like coasting downhill in neutral. You need voltage control over the gate, so shorting it to the source is a better option.
However, they did run out of voltage swing, so I am building a balanced version for myself (with a few changes).
You may want to think about a standard unit but using lower feedback if you want something different.
I am not sure that removing the gate resistor biases the MOSFET to zero Vgs, but an intermediary value based on the source resistance and power supply voltage - the so-called self bias. I think the results of removing all drive is like coasting downhill in neutral. You need voltage control over the gate, so shorting it to the source is a better option.
Yes shorting G and S gives max Idss as 1/-1 uA for N and P channel respectively.
But I also need to disconnect those pesky diodes. And there is no way of doing that without removing them. It's not such a simple task once it is all wired up.
I'm going to play with P3 more tomorrow -- see how little overall distortion I can get.
Does it matter much which frequency you choose? The harmonics are very close to the grass at 2 watts and 1 KHz. They are above at 2KHz or 4KHz -- the noise floor vanishes.
Sure, buy enough to build two channels of 4 pair F5T and you'll have two extra front end boards. Use these with a P2P output section of your choosing. I may end up there.
Another question:
Why does no one produce a Front end board that does not use P channel jfet? Can some sort of "quasi-complementary" architecture be implemented on the front end board?
Last edited:
already made , in several iterations (few as official PL products)
and ,one of them (not official PL), it'll drive F4
Hi
Once you take Vgs down to zero the MOSFET stops conducting and the diode in the source tail also does not conduct.
A few microamps here and there shouldn't matter
For 95dB speakers, a standard F5 sounds like the right choice, much better than the F5 Turbo. Come to think of it, since you need only about 10 watts (I think) There are lots of nice Class A candidates, even single-ended ones that will do what you want.
Assuming the speakers are 95dB/w at 150Hz and not 1Khz.