This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi there,
This is my first post.
I have finished putting my Bride of Zen (1st design by NP)
together and I connected it to my (one channel only - not enough heatsinking yet :-( ) Zen. It works fine but now I have a few questions.
Before my BOZ was operational, my CD player was driving Zen
directly. Not all CDs sounded great.
Q1. Now (with BOZ) everything has changed BUT loud passages on some CDs clip somewhere (I don't know whether it is in BOZ or in Zen,
probably the latter). This surprises me since I haven't noticed this without the BOZ, especially that this happens when the volume pot is set for overall level lower than it would have been without the BOZ. I suspect that without the BOZ at loud passages the output of my CD player simply dropped. This is not the case with BOZ. Can anyone confirm this?
Q2. This is a minor one. There is an ambiguity in NP's on-line docs on BOZ. In some sources C102 is 47uF and in the others 100uF. This has to filter some noise. I use 47uF.
What changes if I use 100uF? (I want to know even if I probably would not hear that.)
Q3. I undersatnd that BOZ has more gain than the usual preamp. Can I possible destroy non-ZEN amp driving it
with the BOZ? Well, I tried this already with no ill effects.


Sounds like the BoZ clips.
Try to lower the bias current to about 30mA, adjust the pot so that you will get about 30V across the 1000 Ohm resistor at the drain of the MosFet.
Hope this helps! (I've heard other people saying that the BoZ sometimes clips.)

I have a little modified BoZ myself (increased gain, increased bias) running at a supply voltage of 100V.

The bias value as stated in the BOZ article at the Pass Labs site is incorrect and causes un-symetrical clipping of the Mosfet.

If you have access to an occiliscope you can bias the BOZ for symetrical clipping which on my BOZ is 28mv and which I have been told can be anywhere from 26mv to 30mv.

If you don't have an O-scope you can do the biasing by ear by playing a loud passage and adjusting the bias for the smoothest sound.

The credit for the information in this post goes to a fellow at the VT52 home page (who's name I unfortunatly forget) as he was the one who told me about the biasing problem with the BOZ.

Take Care.

Tony D.
Thanks guys,

I have played a little bit with different voltages
at 1kOhm resistor as Freddie suggested.
It helped, but not much.
Then I replaced the 1kOhm resistor with a 500Ohm one
(this reduces gain, as I understand).
This helped a lot.
But playing my CD player with the volume pot at min
position is not possible (I hardly play at 1/2 volume).
With the sources which have lower output level
(like my tuner) it works much better.
Also replacing the volume pot with a lower value helps.
However, NP writes that "higher values [of the pot] are OK",
so I am not sure about this modification.

I rechecked my last tweaking. Must have been something wrong there. Now with the reduced bias it works finally.
The sound is gorgeous.
(and it works with 10kOhm volume pot too)

PS. I had a member of a small orchestra at my place
2 days ago. She could not believe that you can produce
such sound yourself at home. She liked the sound of all
instruments and said it is as it should be.
We tried piano, chamber orchestras, symphonic, vocal
(male and female), violins, cello, brass,
classical, jazz, pop... you name it.
Hi all,
this is my first post.
I just built zen and boz, and found that boz did clip.
I managed to solve this problem by limiting the input signal. I put 22k resistors at inputs so my boz has voltage divider before mosfet gate (basically a fixed vol control).
Try to experiment with different values as resistor network at input acts as a RC network (with mosfet capacitance and cd player capacitance) and will affect the frequency response.

Another option is to move the volume control to the front instead of the end, this will solve the overload problem.

I do experience clipping in BOZ quite offen. Setting the bias to 30mA will solve asymmetric clipping but will not solve overload. Moving the volume will cure this.
You are right,
decreased bias solves problem with majority of CDs,
but there are still some that do clip. I will
attenuate the input signal or reduce the gain.
I did not use any "special" capacitors at the input
and output (of course they are not electrolits),
so I bypassed them with the smallest ones.
I don't know what to think of the result. The sound
is more crisp, but now the preamp lisps. So I think
I will remove them.


I did a lot of research on speakers after i built zen. speakers selection depends on your listening style and of course your budget. being a diyer, my budget is limited. what i find in speakers within my budget is: the more efficient a speaker, the more terrible it sounds (ex: new klipsch product lines). it is true that efficient speakers (>95db) will sound louder, but if you don't need a concert hall like sound in your home, you can still live with less efficient one.
after reviewing many speakers, i narrowed them down to dynaudio and kef (both bookshelf types). the final decision is kef rdm2 monitor. it has 6 ohms impedance and 91db efficiency. it sounds excellent especially for mid and high. bass is a little bit weak due to zen's low damping factor and low impedance, but I am not a big fan of thunder bass anyway.
I managed to tighten the bass by implementing pass' zen revisited. setting bias current = 3A improves damping factor.
if you really need a good and efficient diy speaker, check the stright8 (www.bottlehead.com). some people say it is a very good speaker for se amps, but it definitely is out of my budget.

I do not think the quality of the speaker is in any way related to efficiency, actually the Zen requires a relatively high efficient speaker. I constructed a pair of Ariel 6 ( 93db ) last year and it sounds great with my 8 Watt SOZ. I think high efficieny speaker may have an advantage since it requires less wattage to produce the same level of sound. Lower power output from the Amp means lower distortion.
As for the speakers.
I use 8Ohm speakers from Bose. I cannot tell you what
efficiency they have. They are ~10yrs old and I bought
them 2nd hand. They are of a traditional design, no
direction reflecion principle (or whatever they call it).
I do not think any of the problems I experienced was
due to the speakers. I do not play and do not intend
to play my music loud. Actually, Zen is quite loud for me.
And yes, I have enough bass.

So now the clipping in the BOZ is gone.
I adjusted the voltage at the drain of the mosfet
as you have suggested. It helped (in 99%).
Since I do not need that much gain I lowered
the valued of R104 down to 820Ohm (now it is ok in
the remaining 1%).

Then I turned my attention to ZEN. I had the original
circuit. I have upgraded to the Return of Zen.
(with R10 1kOhm, P1 25kOhm and R8 47kOhm and R11 6.2kOhm
instead of 3.3 as by NP). It playes better now, but
when I remove R11 I get the impression of more intimacy,
more dramatic sound. I know that feedback vs. no feedback
is subject to a huge debate. (Ok, I did not remove all the
I am listening to really huge distortions and enjoying them? Or do you think I really get some improvement?

The bad news is (in the low feedback case) that the Zen starts to clip when listened at higher positions of the BOZ's volume pot. To me it is not a big problem, because it is loud enough way before
it clips.

low feedback/distortions


I originally constructed a Zen amp using only N-channel mosfets. What I found, and only by using a distortion analyzer and a scope, was that unless there was about the same level of feedback applied as NP recommended, the frequency response was 5 dB down at 20 kHz. Feedback is necessary to get a decent response out to around 50 kHz. Of course, the more feedback, the less overall gain the amp has, but that is a definite trade-off with only a single stage.

The other problem with the Zen design is that the input source dramatically affects the frequency response. In fact, with some sources you could hear, and measure, the rise and fall in the upper harmonics: that is the input modulates the amp in a non-linear fashion. (This may not be a problem if the amp is matched with a NP mosfet preamp, as the single mosfet preamp effectively behaves as an input stage for the output mosfet in the Zen.) This non-linear response probably occurs because there is nothing to isolate the input from interacting with the feedback. I ended up adding another stage, and finally, after putzing around with various input stages, I scrapped the circuitry and built an Aleph style amplifier.

This isn't the thread to continue with the Aleph story, but I did modify the NP design and wound up with an amp that has 10 times less distortion, a freq. response to 200 kHz, and between 5 to 10 times greater S/N ratio (around -102 dB) than a stock Aleph. I've also simulated the design and optimized the IV control portions of the circuit so it can better deliver current into 4 and even 2 ohm loads, albeit with slightly more current and heat dissipation. I made a 4x7 inch circuit board that holds everything, including the power supply and up to 4 pairs of output mosfets: just plug in a transformer and the input and output jacks. If anyone is interested I can provide more details.
Re: low feedback/distortions

Would you be so kind to send me more details about you construction. I made some simulations of Hoffmanns ampamp like Alpeh 3)- the results were very bad - THD about 2%. So I am interested in your results ( NP declares THD less then 0.2%, for Alpeh 3). Are you better?? Koy
I read with interest all the comments on the BOZ preamp. Surprising none of you list noise (humming @ 50/60Hz and buzzing @ 100/120Hz) as a problem. See my posting in the other Bride of Zen category on this problem.

Putting the volume control at the input in my case resulted in circuit (oscillation?) at the extreme settings of the pot when coupled to the Leach Amp. Even when it's not oscillating, hum/buzz level is extremely high, measuring at least 50mV at the power amp output.

Nelson Pass suggested that I increase the gate resistance from 100 ohms to 1,000 ohms. This helped a bit, no more oscillation but still the high level hum/buzz problem........ Moving the pot to the output solved this problem.

Supposing the BOZ has a gain of 8 (18dB) and clips at an output level of 10V, the maximum input level must not exceed 1.25V; which any CD player will exceed. Again, Nelson's suggestion is to attenuate the input. I might try to lower the gain as stated in this forum.
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.