50w Single-Ended BAF2015 Schade Enabled

Hi, I'm wanting to build a pair of these Mosfet amps.

I have quite a few bits in my spares box that can be used to get something up and running at reasonable cost.

One question to start with, there have been a few suggestions as to the 4.7v bias supply.

The 4.7v zener idea seems the most simple, has anyone tried it yet, and if so where did you take the feed from ?


Philip.
 
Subscribed :)

Interesting thread this is!

Read it twice now, and the BAF2015 also.

Where do you guy's buy the Jensen transformer here in Europe btw?
Also i don't exactly understand, if there is any preferable mosfets to use, or is it only matter of power? - I mean, i have IRFP240/250 in stock here home!

Anyway i would prefer to build a lower -powered version, if possible.

Rgds; Jesper.
 
Thank's again...

Soo, i have to make two supply's... one with e.g 30-50vdc, and a stable one with 4,7vdc (Been reading a lot of this in Schade threads)...

The part with the 4,7vdc is a bit confusing to read, as there is no really clear solution for this (Not that i didn't read about it).

Have anyone found a good solution for this?

Rgds; Jesper.
 

Attachments

  • 50wSESchade-red.JPG
    50wSESchade-red.JPG
    71.1 KB · Views: 685
Member
Joined 2000
Paid Member
The part with the 4,7vdc is a bit confusing to read, as there is no really clear solution for this (Not that i didn't read about it).

Make a variable voltage supply and connect it to bias section. You can then later remove the variable resistor and measure the resistance and plug in the closest fixed value. I would personally keep the variable resistor in case other value change over time and you can set it back to 4.7v or whatever voltage value you want. Don't listen to me though, I have no idea what I'm doing. :p

Adjustable Voltage Regulator – Electronic Circuits and Diagram-Electronics Projects and Design

Physical example:
https://www.parts-express.com/posit...c-in-dc-out-based-on-lm317t-regulato--320-614
 
Nice! :)

I am trying hard to really understand how the circuit work's.

Is there anywhere a description of the circuit? (I have seen the video with Nelson, but English is not my native) even if i understand most, it's difficult to understand when looking at schematic.

Normally i understand quite how circuits work, before doing them, but this must be to simple for my strange brain sry... :rolleyes:

Rgds; Jesper
 

Attachments

  • 50wSESchade.JPG
    50wSESchade.JPG
    59.3 KB · Views: 761
Member
Joined 2000
Paid Member
The video is the best for information on this circuit.
In someway, the ACA (amp camp amp has some characteristics of this amp. The SE 50w has feedback that give the mosfet triode characteristics, so it's special that way.

There are other SE Schade feedback amp examples here but none this simple. If you go to firstwatt.com and passdiy.com most of this is broken down in the articles sections.

Breaking down into 3 areas, input section, output section and current source.

Input: uses a 1 to 1 transformer, no gain. Provides good isolation for noise on input.

Output: uses very heavy duty mosfet to handle the power and is set with Schade feedback to get the triode curves. It's also single ended where one mosfet handles 100% of the cycle.

Current Source: uses another heavy duty mosfet and clever iso-coupler to set the constant current source (ccs)
So no matter what happens with the voltage, the current delivery is constant.

A 50w build will be large so mono blocks are recommended.

I'm sure I'm missing something, so someone else please correct and fill in the gaps.
 
Last edited:
Breaking down into 3 areas, input section, output section and current source.

Input: uses a 1 to 1 transformer, no gain. Provides good isolation for noise on input.

Output: uses very heavy duty mosfet to handle the power and is set with Schade feedback to get the triode curves. It's also single ended where one mosfet handles 100% of the cycle.

Current Source: uses another heavy duty mosfet and clever iso-coupler to set the constant current source (ccs)
So no matter what happens with the voltage, the current delivery is constant.

Isn't the input transformer part of the Schade feedback arrangement?
 
If my simulation is right it looks like the bias adjustment will be pretty touchy with a single 5k pot. Might have to use both a coarse and a fine adjustment. I used this circuit with the baf2015 circuit and in simulation the 1000u cap that is connected to ground took care of all the ripple. And wow! Won't it take about a whole minute to reach the peak bias voltage because of the 10k ohm and 1000u values.
 

Attachments

  • DCbias.png
    DCbias.png
    12.2 KB · Views: 671