Lateral CFA 120W - BSA

There is one interesting possibility to change a bit THD and harmonic shape at high frequencies, at 20kHz, by changing value of R20, the two pole compensation resistor.
Original value is 1k5 but could be lowered down to 220R and THD20k goes down, harmonics get more euphonic. Downside is lowering the Phase Margin but still stays very stable.
This could be done wile listening and decide what is better.
 
Last edited:
There is one interesting possibility to change a bit THD and harmonic shape at high frequencies, at 20kHz, by changing value of R20, the two pole compensation resistor.
Original value is 1k5 but could be lowered down to 220R and THD20k goes down, harmonics get more euphonic. Downside is lowering the Phase Margin but still stays very stable.
This could be done wile listening and decide what is better.

So it's true that certain compensation scheme can be used to tailor the distortion spectrum. I tend to search for the simplest compensation, which is critical for low bandwidth amplifiers, and an easy load for the driving stage.
 
This amp is simple to put in pure ClassA, 40W/8ohm and 80W/4ohm.
 

Attachments

  • 40W-LMOSFET-VAS-OITPC-ClassA-sch.jpg
    40W-LMOSFET-VAS-OITPC-ClassA-sch.jpg
    117.4 KB · Views: 825
  • 40W-ClassA-lateral-simple-FFT1k10_8.jpg
    40W-ClassA-lateral-simple-FFT1k10_8.jpg
    172 KB · Views: 805
  • 40W-ClassA-lateral-simple-FFT1k40_8.jpg
    40W-ClassA-lateral-simple-FFT1k40_8.jpg
    167.2 KB · Views: 761
Just curious (as in an engineering exercise for the 120 Watt BSA version) can a pair of ecx10's drive a 4 Ohm load ?
It seems based on the SOA curves probably not.

Would the output stage require an extra pair of mosfets ?

What about substituting the ecx10's for the ecw20's ?
In either case I guess the driver stage would need to be modified for the additional ecx10 or ecw20 mosfets ?
 
Last edited:
Just curious (as in an engineering exercise for the 120 Watt BSA version) can a pair of ecx10's drive a 4 Ohm load ?
It seems based on the SOA curves probably not.

Would the output stage require an extra pair of mosfets ?

What about substituting the ecx10's for the ecw20's ?
In either case I guess the driver stage would need to be modified for the additional ecx10 or ecw20 mosfets ?

Simulation shows it can drive 4 ohm at full power.
You can use one pair of ecw20 mosfets instead two pairs of ecx10, and to use 2 pairs I am not sure yet.

http://www.exicon.info/PDFs/ecx10n20.pdf
 

Attachments

  • 40W-ClassA-lateral-simple-85W_4-SOA.jpg
    40W-ClassA-lateral-simple-85W_4-SOA.jpg
    148.3 KB · Views: 647
  • 40W-ClassA-lateral-simple-20W_4-SOA.jpg
    40W-ClassA-lateral-simple-20W_4-SOA.jpg
    140.4 KB · Views: 632
Thanks for the information, if I read your charts correctly then 4 ohms at 50vdc will exceed a pair of ecx10 mosfets soa curve.
For the BSA 120w amplifier to operate into a 4 ohm load it is necessary to lower the dc rail voltage.

I guess to have the BSA 120w amp capable of operating unconditionally into either a 4/8 ohm load is to add additional ecx10's albeit with a different pcb layout or substitue with a pair of the ecw20 variant which would probably require component changes in the driver stage.
 
Last edited:
Thanks for the information, if I read your charts correctly then 4 ohms at 50vdc will exceed a pair of ecx10 mosfets soa curve.
For the BSA 120w amplifier to operate into a 4 ohm load it is necessary to lower the dc rail voltage.

I guess to have the BSA 120w amp capable of operating unconditionally into either a 4/8 ohm load is to add additional ecx10's albeit with a different pcb layout or substitue with a pair of the ecw20 variant which would probably require component changes in the driver stage.

Have you looked ClassA schematic, there is +-35V of power supply?
This amp is 80W on 4ohm.
 
Last edited:
I've now worked out that I was looking at this from the wrong angle and interpreted the mosfet soa curves from the datasheets from a dc voltage perspective and not the ac rms voltages and thusly the current and rms power into the load or do we use the peak ac voltage and current ?

If that is the case then reduce the dc rail voltage to give 100w into 4 ohms which is plenty of power for a typical room at home :)
 
Last edited:
You can take the pre-driver stage out, stability will improve, a little higher distortion, but more music

I suppose you meant without drivers, JLH did not use drivers in his 80W mosfet amp.

Here is this ClassA with no drivers. I increased the VAS current to cope with the mosfet drive, and here is the result.

Interestingly THD at low frequencies dropped but at high increased, specially at 4ohm load. Stability (PM and GM) did not changed, it is good as before, clipping a bit asymmetric.
Damir
 

Attachments

  • 40W-LMOSFET-VAS-OITPC-ClassA-supersimple-sch.jpg
    40W-LMOSFET-VAS-OITPC-ClassA-supersimple-sch.jpg
    113.5 KB · Views: 682
  • 40W-ClassA-lateral-super-simple-FFT1k10_8.jpg
    40W-ClassA-lateral-super-simple-FFT1k10_8.jpg
    175.5 KB · Views: 680