Hi William.
The xsoz isn´t like most poweramps. It´s not trivial to keep it right on track, the highend rolloff is the big challenge to keep alive, it has to do better than -3db at 100kHz. The sound is degraded already when its highend roll off is -3db at 80kHz, I have tried a lot of times.
I have attaced a graph showing an 7 db attenuation done with a resistor at 18 Ohm across the + and - terminal at the xbsoz output.
This is quite a normal listening level with my speakers in my room. And it is obvious why this dosen´t sound good, we are down -3db at 65kHz at the lowpass at the output of the xsoz.
Regards
The xsoz isn´t like most poweramps. It´s not trivial to keep it right on track, the highend rolloff is the big challenge to keep alive, it has to do better than -3db at 100kHz. The sound is degraded already when its highend roll off is -3db at 80kHz, I have tried a lot of times.
I have attaced a graph showing an 7 db attenuation done with a resistor at 18 Ohm across the + and - terminal at the xbsoz output.
This is quite a normal listening level with my speakers in my room. And it is obvious why this dosen´t sound good, we are down -3db at 65kHz at the lowpass at the output of the xsoz.
Regards
Attachments
Hi again William.
I have worked a little more with your idea with an attenuator at the output of the xbsoz, and I think you are right after all, it might be done, and it´s easy to try in real life since I can try it outside the box. But the values of the resistors in series with the signal has to be as low as 56 Ohm, and for 6db attenuation the resistor across + and - has a value of 220 Ohm. At this attenuation I can keep the lowend roll off at 100kHz. I haven´t ben looking at the open and closed loop gain in the xsoz, but I believe they should keep the ratio. I will find out later.
Thanks for the suggestion William.
Regards.
I have worked a little more with your idea with an attenuator at the output of the xbsoz, and I think you are right after all, it might be done, and it´s easy to try in real life since I can try it outside the box. But the values of the resistors in series with the signal has to be as low as 56 Ohm, and for 6db attenuation the resistor across + and - has a value of 220 Ohm. At this attenuation I can keep the lowend roll off at 100kHz. I haven´t ben looking at the open and closed loop gain in the xsoz, but I believe they should keep the ratio. I will find out later.
Thanks for the suggestion William.
Regards.
Attachments
more observations
after drilling some holes in the top of the preamp and power supply I can leave them on permanently without worrying too much about overheating, fire etc.
Things seem to have burned in by now and the sound is very good indeed. Did some more listening with balanced and unbalanced connections. The difference is smaller than the difference between my balanced and unbalanced cables so for the moment I´m back to an unbalanced connection.
Finished the knobs for volume and power supply and only have to attach some parts of the fronts so I hope to be ready tomorrow.
william
after drilling some holes in the top of the preamp and power supply I can leave them on permanently without worrying too much about overheating, fire etc.
Things seem to have burned in by now and the sound is very good indeed. Did some more listening with balanced and unbalanced connections. The difference is smaller than the difference between my balanced and unbalanced cables so for the moment I´m back to an unbalanced connection.
Finished the knobs for volume and power supply and only have to attach some parts of the fronts so I hope to be ready tomorrow.
william
William ,
Do you drive the XBOSOZ unbal at the input ? Is that what you mean ?
Did you keep resistors or did you use a CCS ?
I guess you are using the balanced input of the XBOSOZ otherwise as Nelson states , you don't really get the X benefit ( a single pole output carries "more " distortion than using the 2 halves , where , according to the Supersymmetric concept , the distortion and noise is made identical on the 2 halves , thus vanishes at the output
The THD+N of a single pole output may carry a lot of degradations ! Who knows ? Ear will tell ...
Do you drive the XBOSOZ unbal at the input ? Is that what you mean ?
Did you keep resistors or did you use a CCS ?
I guess you are using the balanced input of the XBOSOZ otherwise as Nelson states , you don't really get the X benefit ( a single pole output carries "more " distortion than using the 2 halves , where , according to the Supersymmetric concept , the distortion and noise is made identical on the 2 halves , thus vanishes at the output
The THD+N of a single pole output may carry a lot of degradations ! Who knows ? Ear will tell ...
nar,
I drive it balanced (ONO) and unbalanced (CD). The differences I´m talking about are the unbalanced versus the balanced output. Both into my Aleph5
I´m using resistors (4 x 680 ohms) no current source (I will try this later but was put off by Henriks comments on the sound of these)
I´m not shure if 10x or 100x less distortion is so important if the "standard" distortion is low enough. I will eventually sell my unbalanced Kimber at EBAY and get the balanced Version in time for my Aleph-X.
William
I drive it balanced (ONO) and unbalanced (CD). The differences I´m talking about are the unbalanced versus the balanced output. Both into my Aleph5
I´m using resistors (4 x 680 ohms) no current source (I will try this later but was put off by Henriks comments on the sound of these)
I´m not shure if 10x or 100x less distortion is so important if the "standard" distortion is low enough. I will eventually sell my unbalanced Kimber at EBAY and get the balanced Version in time for my Aleph-X.
William
on the back there is a 230V input on the right with switch and fuse. On the left are the two XLR-connections to the preamp. One 6- and one 7-pole (so I don´t feed +-60V to my ONO). The 6 pole feeds power to the X-BOSOZ, the 7-pole to the ONO. Plus, minus and ground are seperate for both channels.
Attachments
and this is the preamp without the top. On the right is one veroboard holding both x-bosoz channels, on the left are the two ono boards. All the ground connections are connected to the case at one point ( 4 grounds from the power supply, 2 from the x-bosoz, 2 from the ono, two from the inputs and two from the outputs plus grounding for the tonearm).
At the back you can see the Elma 24pos switch used as a balanced output attenuator.
At the back you can see the Elma 24pos switch used as a balanced output attenuator.
Attachments
Hi William
Very VERY nice job.
Housewarming makes sence in a new way, doesn´t it!
Which part of your preamp reaches 43 degree C?
I can´t see your output fets on the photos, are they mounted under the veroboard at the bottom of the housing?
Once again brilliant.
Cant wait to read about your tests on current sinks, some like them some just don´t. I have used them to reduce the negative rail in the outputstage of my CDP, they served a purpose and that was just ok.
Regards
Very VERY nice job.
Housewarming makes sence in a new way, doesn´t it!
Which part of your preamp reaches 43 degree C?
I can´t see your output fets on the photos, are they mounted under the veroboard at the bottom of the housing?
Once again brilliant.
Cant wait to read about your tests on current sinks, some like them some just don´t. I have used them to reduce the negative rail in the outputstage of my CDP, they served a purpose and that was just ok.
Regards
Moe, Henrik,
thanks for the compliments.
Henrik,
yes the fets are below the board mounted on the bottom of the case. Temperature is the same all over the case (after a few hours). Total dissipation in the preamp should be a bit over 40 watts (not measured it yet) so a high temperature was to be expected.
I think you´ll have to wait a while before I can tell something about the current sinks. I will leave it like it is for the next few weeks to get a bit of rest back in the system (changed quite a lot in the last few months including (and most important) changing to balanced ac power) and listen to some more records and cd´s. The difference between my OPA627 based preamp and the new one is getting bigger with every new record I play
attached is a picture of the x-bosoz where you can see the holes for the hex-key to fasten the fets. (picture was taken before I doubled up the output caps)
william
thanks for the compliments.
Henrik,
yes the fets are below the board mounted on the bottom of the case. Temperature is the same all over the case (after a few hours). Total dissipation in the preamp should be a bit over 40 watts (not measured it yet) so a high temperature was to be expected.
I think you´ll have to wait a while before I can tell something about the current sinks. I will leave it like it is for the next few weeks to get a bit of rest back in the system (changed quite a lot in the last few months including (and most important) changing to balanced ac power) and listen to some more records and cd´s. The difference between my OPA627 based preamp and the new one is getting bigger with every new record I play
attached is a picture of the x-bosoz where you can see the holes for the hex-key to fasten the fets. (picture was taken before I doubled up the output caps)
william
Attachments
Too much gain
Hi, I just modified my BOSOZ into X - , - and use 20k attenuator at input.
Problem is that it has so much gain, that I get to much noise, and not low enough volume.
even at first step, I'm blasted away...
Is there a easy way to reduce gain, without just attenuate further, on in or output?
Arne K
Hi, I just modified my BOSOZ into X - , - and use 20k attenuator at input.
Problem is that it has so much gain, that I get to much noise, and not low enough volume.
even at first step, I'm blasted away...
Is there a easy way to reduce gain, without just attenuate further, on in or output?
Arne K
Hi Arne K
Something is wrong, if you use my scematic the gain in bsoz and xbsoz is quite the same.
It seems as if you run the thing in open loop and not closed.
When you short the sources the gain rises, but is again lowered to normal by the feedback and the inputresistor.
Be shure you have the 10K inputresistors in place and that the loops are connected properly. If you run the xbsoz unbalanced in, you short -in by the 10K to ground.
Regards
Something is wrong, if you use my scematic the gain in bsoz and xbsoz is quite the same.
It seems as if you run the thing in open loop and not closed.
When you short the sources the gain rises, but is again lowered to normal by the feedback and the inputresistor.
Be shure you have the 10K inputresistors in place and that the loops are connected properly. If you run the xbsoz unbalanced in, you short -in by the 10K to ground.
Regards
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.