x soz

Jump to Latest
#344
Ralf

You are quite right, in my tests I have these C3 and C4 even they are not in my scematic, my fault.

For C101 and C102 I used some blackgate at 47uF/50v, and for C201 and C202 I used Panasonic HFS 220uF/25V.

In my final version I will go for Blackgates, C-type 50V for C101/102 and BG-Nx-type 6.3V for C201/202.

But I think that Panasonic FC or similar type will do fine.
 
#345
XSOZV2 R3 and Frequency Response

Henrik

I finished Revision 3 of XSOZV2 with the following alterations:
  • MOSFETs to IRFP140
  • R2 to 430
  • R5 to 4.5k
  • R4 to 5k-pot (to be adjusted as about 3.3k)
Results are:
  • Lower Fc = 2.813Hz (-3dB)
  • Upper Fc = 51.94kHz (-3dB)
  • At 20kHz, about -0.65dB
I could not lift the upp Fc further up. I send all of you SOS.
The new circuit and frequency response curve are attached.
 

Attachments

  • xsozv2_r3.jpg
    xsozv2_r3.jpg
    42.5 KB · Views: 972
#349
Henrik

From the frequency formula f = 1/(2x3.14xRxC), I could see
three possibilities to lift the upp fc up.

1. Minimizing the resistance of the signal source
2. Minimising the Miller capacitance by selecting proper MOSFET
3. Minimizing the voltage gain

Accordingly, your suggestion of R8 down to 100 is a good idea.
If there is no problem of parasitic oscillation, I would even remove R8.
Then, the upper fc could rise up somewhat.

:dunno:
 
#350
Keep up the god spirit!

I suggest You do all tree, but You need about 20db gain, 15db as minimum.
I think You need the gateresistors at min. 100 Ohm.

I have almost finishd my test CCS´s in my XBSOZ, may be I will come back tonight with som feedback on this.
 
#352
I have just listend to my XBSOZ with and without CCS at the tail.

I have only singleended signal sources, so the XBSOZ have to convert into balanced mode.

The sound of the CCS is more soft, round and less dynamic, sweeter, but the directnes of all the precented instruments and voices is signifcantly degraded, and because of that the feeling of tactility and space in the soundstage suffer from this too.

This is my first impressions, I need to listen to this difference some more, but I think I will end up with resistors only and no CCS, not of "religious" reasons but purely sonically.
 

Attachments

#353
XSOZV2 R4 and Frequency Response

In order to get the widened frequency response, I finally
adopted the reduction of gain by introducing negative
feedback loop. I will examine sounds of both Revision 3 and this
Revision 4.

Results of Revision 4:

* Lower Fc = 2 Hz (-3dB)
* Upper Fc = 100 kHz (-3dB)
* I expect 16W/ch with 8-ohm load.

<Revision 4>
 

Attachments

  • xsozv2_r4.jpg
    xsozv2_r4.jpg
    49.3 KB · Views: 1,059
#357
Henrik
It is cheerless to hear that the constant current source
was helpless in sound effect. I would however recommend
listening further for certain period before you make a
conclusion. I avoid one day A/B test. I often find mine
after long hour A/B test, minimum one week.
 
#360
Henrik
You will manage well.

I have my understanding of the constant current source (CCS):
  • For the balanced-in and balanced-out, the CCS will maximise CMRR.
  • For the unbalanced-in and balanced-out, the CCS will afford symmetrical output swing.
I hope my understanding would come true with sonic effects.

By the way, I might think about the splitting CCS, port and starboard,
for X-effect. Otherwise, I presume there is no X-effect.
With one CCS in the middle, how the left and right channel can exchange
the signals...?
 
You must log in or register to reply here.
Top Bottom