Questions about a CCS-X-BOSOZ

Ok, I've got my Aleph-X working and now I need a better pre. So I looked around and compiled a few designs into my own design, which is really pretty standard. But there are few things about the general design of the circuit that I would like to understand better.

I have attached a schematic which I will refer to in this post:

First question: How do I calculate the gain of a SuSy line stage? I would like to get at least 10 db of gain, and have it adjustable up to about 20 db (just like the original bosoz). R21 and 22 are the feedback resistors and I just copied the value from some other designs I looked at.

Question 2 relates to the above question: are the values of P5 and R15 also the correct values to get the required gain adjustment? Since the values in my schematic are lifted straight form the BZLS article, I'm guessing something will have to be changed since the feedback resistors will reduce the gain. Once again I don't know how to go about calculating the gain.

Question 3: Are the input and output pots the correct values and configured the correct way for optimal input/output impedence and performance. I will most likely use ladder type attenuaters for the both input and output.

Q 4: I am using the dual CCS approach because I would like to have adjustable gain. Can someone tell me if this design is appropriate for this circuit? Also, are the values of R3 and 4 correct in order to get about 40mA through each leg. This is exactly the same current source I used for my aleph-x. It's a 9.1V zener bypassed with a .1uF film cap to reduce noise and a J505 1mA current regulator as a supplemental stabilizer to keep the current through the zener constant. Anyone used one like it?

Hope you all can help, I believe my current pre is inadequate to get full performance of my new Aleph-X, and I would like to change that as quick as possible!


  • dualccsxbosoz.gif
    9.2 KB · Views: 942

answer 1: By looking at the feedback resistors R18,19,21,22. These define the input output ratio.

answer 2: normaly you should leave p5 and r15 out to have enough gain for the feedback to work.

answer 3: yes although you can also use 20k input pots for a nicer load on your cd/phono stage. I wouldn´t use the input pots but a fixed pair of resistors, set to get the right attenuation (in my X-BOSOZ I only use pots on the output).

answer4: I think one ccs would do (see answer 2) but the resistor value seems correct for 40mA (9-3,5)/150 = 37mA.
Why do you want the adjustable gain?

jh6you, can you show me the calculations involving P5/R15? I understand the concept of how they work but not the math.
Also, doesn't turning P1/2 effectively vary the feedback ratio? That doesn't sound like a good thing do do, or does it matter that much? And doesn't increasing R21/22 decrease the "X" factor? I guess I am actually trying to find out how much feedback I need to get good susy performance. Does changing R21/22 to 220K still offer good susy performance?

Wow, there are a lot of questions in that paragraph. . .

Anyways, moving on. . .

Wuffwaff, I would like adjustable gain because I will be using both +4dBu (1.23V nominal) and -10dBV (.316V nominal) sources. I just did the math and the level difference between these standards is 11.8 dB. So when using my mixer or soundcard outs (this is NOT a soundblaster) I will need 11.8 dB less gain than when I will be using my CD/DVD/Phono sources to get the same output level to the power amp.
Because of this I have also rethought my gain requirements. If 10dB of gain is enough for the lower level sources (cd/dvd/etc) I won't need any gain for the higher level sources. So I'm thinking of making this pre switchable or adjustable between unity and 10dB gain.
This brings up the dual ccs. I would like to keep the feedback ratio the same when changing gain settings in order to keep the amount of susy content the same, so I that leaves the resistance between the diff pair as the gain adjustment. I also seem to remember reading that several people liked the sound better with some resistance there, but that might be my imagination. Can anyone confirm or deny?
Another reason I wanted so much adjustment was so that I could hear the differences between different gain levels, feedback ratios, input/ output levels, and find out what sounds best.
Wow, there are a lot of questions in that paragraph. . .

I feel myself farting in front of the elephant. Whenever I talk about this type of technical matter, I sweat fearing the elephant looking down at me. You know who he is. . .

When I thought about the gain, it was after disconnecting P1 and 2.

In the PassDIY article, Nelson Pass says that the amp gain, considering only R1, R2, P5 and R15, is formulated by (R1+R2)/(P5+R15). I interpret this as an open loop gain (A). Meanwhile, he also says that the gate of the gain MOSFET could be assumed as a virtual ground. If so, the feedback arrangement looks like the op-amp feedback connected to the inverting amplifier (B). If you have the values of A and B, the math is your piece of cake, IMHO.

Of course, if the P1 and P2 are involved in, the calculation might merge into bloody hell and the gain values might be different.

I have no idea about the X factor you are mentioning.
By the way, Nelson Pass says there is still very little of X-factor even without having the feedback resistors, R21 and 22. My BOSOZ has this very little of X-factor. The sound is awesome. . .

The place where the greatest effect on X-ness will occur is P5 and R15. The higher their resistance, the greater the attentuation of the susy cross-communication current. Increases in R21,22 do increase the X-ness, but in proportion to the increase to the output signal level with decreasing feebdack, so relatively speaking, no significant gains there in terms of X-ness.

Overall, it a game of balancing tradeoff's. To max the susy, you have to give up gain adjustment using P5, and instead regulate the gain by increasing the feedback level through R21,22. That being said, with the schematic values you have posted, you should still get a significant susy effect, and still have your adjustable gain.

The middle ground would be to eliminate R15, thereby increasing the susy, decrease R21,22 to bring the gain back to the same level. Leaves you with some gain adjustment, but gives you the max X-ness for any given gain level.

My two cents, for what they're worth.


P.S. I too am astounded with the sound of my BOSOZ, except mine is thoroughly X-ed. Whatever you decide for this circuit, you will not be dissappointed.
If you want to have the adjustable between unity and 10dB gain, I would recommend to try it with the total value of P5+R15 variable inbetween 1k-124. Good luck.

I buy lot of the cheapest resistors having different sizes and continue re-soldering them until I meet amount and quality of the sound to my satisfaction. . . At the end, I replace them with a little bit better grade resistors. . . This works much better than my math. . . And I find p2p as convenient for this practice. . . :)

Best Regards
OK, so after hearing from you all and doing some research, I have changed my mind and have taken away the dual ccs and replaced it with a single ccs. Beyond that, I will do gain adjustment one of two ways: either a switch that selects from two different feedback resistor values, or a pot which will vary the same resistors. Now, I just have to decide on exactly how much gain I need. I will be running balanced into my 120W aleph-x, but using both balanced and unbalanced sources. And I don't have a thing for VERY LOUD MUSIC. So lots of gain is not needed. What do you all think?
Hmmm... Are the source resistors still necessary, as the CCS is presenting a high impedance to the sources already? When I was building my X'ed BOSOZ, Nelson advised me to eliminate the source resistors altogether.

On the gain subject, I did the quick calc, and with a preamp gain of 9.8dB, a 120W Aleph-X will clip on maximum signal peaks, assuming a CD player conforming to normal output levels. I built my preamp with 13dB, have 89dB/Wm speakers, and I find that only on the rarest occasions do I turn the volume up past halfway.

Cheers, Terry
Among others. . . I have thought about the transeconductance. I usually use unmatched Q1 and Q2. They might have different characters in their transconductances. Therefore, I have thought that having source resistors would be of help. . . one step before the same help by the X-feedback. . .

Nelson Pass, is my thinking an idle thought?


Nelson Pass

The one and only
Paid Member
2001-03-29 12:38 am
jh6you said:
Among others. . . I have thought about the transeconductance. I usually use unmatched Q1 and Q2. They might have different characters in their transconductances. Therefore, I have thought that having source resistors would be of help. . . one step before the same help by the X-feedback. . .

Source resistors are generally recommended to equalize and
linearize the performance without loop feedback.
Well y'all, after even more research, I have decided not to build an active pre and instead I am going to build an aleph-ono/xono and a passive pre/signal selector/attenuator since it seems that most modern components put out enough voltage to run without an active pre.
Thanks all for your input on this doomed project, but I'm sure I will hear from you all when I get stumped with the ono.
Once I have tried one passive pre. . . Uh. . . not successful. . . My BOSOZ has the same gain as a passive pre, but with the BOSOZ shows a much better difference. . . I am still wondering why. . .

One of my colleges said to me in my living room, sharing a glass of red wine with us: ¡°I¡¯m using CD player directly to my power amp, you know Electrocompaniet. . . ?, but I think I need an active pre amp.¡±

Good luck and Regards