F5 power amplifier

Nelson pcb,
And when are these little Nelson Pass pcb's going to be available? I would like about 4 of them.

On another note. What is the preferred method of utilizing jfets in the input section when the voltage available is beyond the fjets range. Many members have their own ideas on this ie, resistor, cascode, series with mosfet etc. Which works best? Thanks. Tad
 
And when are these little Nelson Pass pcb's going to be available?

Master says soon. So somewhere this year?

What is the preferred method of utilizing jfets in the input section when the voltage available is beyond the fjets range

You can just cascode it - with whatever you want (bjt, mosfet, jfet). The resistor you mention is the load and will not ensure that the jfet sees sufficiently low voltage.

Have fun, Hannes
 
> I'm waiting for Nelson's pcbs. Sure I could do my own, but having at least a single Nelson Pass piece is a bit special

This place is called DIY Audio ??

> I'm still thinking of the Toshibas, it's just that I have more than enough Fairchilds already at hand...

Why not just build the Fairchild first ?
Then you can always decide to swap the Toshibas in later on.


Patrick
 
By the way, could somebody tell me what the limiting factor for the F5's power into 2R is?

Fig. 10 in the article shows clipping at 50W into 2R - at this power the F5 already left ClassA (5 A output current), but didn't not yet approach the rail (10 V output voltage).

What's happenin here?

Have fun, Hannes
 
Safe Operating Area limits of the MOSFETs. At least that's what I take it Papa means when writing: "The amplifier with 24 supply rails should be good for a 50 watt peak into 8 ohms, or about 2.5 amps. For a 4 ohm load we would want 5 amps, and for a 2 ohm load, we want 10 amps. Since we only have two output devices, we probably are best off stopping there."
 
Has anyone come up with replacements for te 2sk170 and 2sj108... and the one alternative given for each by Nelson?

My boards are 55mm x 130mm, virtualy fully semetrical, with the outputs about 100mm appart.

The 2sj108 and its alternative specificaly seem to be hard to find...
 

Attachments

  • pass_f5.gif
    pass_f5.gif
    31.2 KB · Views: 2,221
I would perhaps move the power supply connections to the bottom left and right corners to aviod any voltage difference between the common nodes of the MOSFET source resistor and the JFET drain resistors. Also makes it easy to double up to a balanced on one PCB. And for that matter perhaps use a soldered jumper to connect the 10R (for which I use 3x 33R 0.25W in parallel) JFET source resistors to Gnd, which allows you to convert to an X-Balanced later. All options then open.

Otherwise quite well done, I thought.
;)


Patrick
 
Hi thanks for the reply...

I thought about cornering the two power points, but decided to stick with the stared layout giving all components accesing those nodes, connection at a single (reference) point, in the same way you might be used to seeing ground stars. It has the benefit of not makeing high and low current paths shareing the same tracks... (Quieter operation)

I do not see myself going balanced soon, but I will try to find the balanced schematic in this long thread, to get a grasp of what you intended wit the 10R's.

Also, I still need to sim the boards to check dissipation etc, so it will likely still change a little...