BOZ mods - using with "standard" amp

[Tieftoener]

Ok,

I decided to get my feet wet and try out the BOZ preamp. For the record, THANK YOU Mr. Pass for sharing your designs with us!

Today, prior to beginning my PCB layout, I decided to calculate the gain of the BOZ...

"The gain of the circuit is the ratio of the output loading divided by the apparent resistance formed by 100ohm R108 and the inverse of the transconductance of the MOSFET." - per Nelson Pass

So, after some quick calculations (I'm using a 20k pot on the output, rather than the 5k spec'ed - just cuz I happened to have a couple nice Panasonics left over in my stash)... I came to the conclusion that the numerator of the gain ratio would be about 925ohms for me. There is a 33k input impedance on my Parasound amp... with the typical denomonator of the gain ratio being about 110ohms (per Nelson Pass in BOZ document - R108 + inverse of transconductance)... this results in an approximate gain ratio of 8.5 V/V.

What this means is that a full scale output of a CD player (let's say 1.5V) would put 12.7V into my Parasound amp. NOT GOOD!! (...the Parasound has an input sensitivity of 1.2V for max output). Unfortunately, it never really dawned on me to calculate all this before I ordered the parts (dumb, I know - but I was motivated to get something to replace my loaned Adcom GFP-555). I was inspired after re-reading the intro and history of the design of the BOZ to be mated with the Zen amp. So...........

I came to three solutions for my problem. If I make R108 1kohms, rather than 100ohms, I'd have an approximate gain of 0.92 V/V. That suits the purpose much better, I feel. However... this would greatly reduce the current biasing of the FET to 4mA, rather than 40mA. And a large part of the reason the BOZ sounds so good (from what I understand) is that the bias current is able to be set so high. So... that basically defeats the major benefits of the circuit.

To get around this (and bring the bias up to 40mA again), the bias voltage could be increased at the gate, but that would require about 44V at the gate... and the Vgs breakdown is about 20V on the MOSFETs. So, that clearly wouldn't work at all.

My second solution would be to add a large series resistance to the output in order to create a voltage divider with the input impedance of the Parasound (33kohms). To get the roughly 8.5V/V gain ratio of the default schematic down to about 1V/V, I would have to add about 275kohm in series with the trimmer output of the volume pot.

My third solution would be to reduce R104, currently 1kohm. If I made this 100ohms, that would also reduce the gain greatly, close to the desired level of about 1V/V. Unfortunately 40mA through the 100ohm drain resistor is only going to reduce the bias voltage of 60V to about 56V. That means there will be 56V riding at the output. Fortunately, the DC blocking caps I purchased are 100V rated so this would still work i suppose. And the IRF610 is rated for a Vds of 200V, so the FET should be OK too. I just don't know if having 56V at the output is such a good idea... Am I missing something else that would be changed as a result of making this modification????

I surmise that the second solution would sound better than the first, due to the 40mA bias. Granted I will be hearing more "resistor" ... but I think the cons of the restance are greatly outweighed by the pros of the greatly increased biasing.

All that, just to ask... "What do you think????" Opinions please!?

Many thanks!

To avoid any confusion. Right now, I just want to get the BOZ working with my Parasound amp. Buildig a matching Zen amp right now isn't an option I'm considering. Eventually I WILL build an amp... but not at this time. I'm leaning more towards the Aleph designs anyway...

[stappvargen]

Why don't you use your 20KOhm pots at the input of the BOZ, instead of R107, as a voltage divider. These you can fiddle with till you get the gain you want. Then buy two new pots, preferably with no more than 5KOhm, to be placed at the output. If you ever get to building an amp that requires the full gain, just remove the pots at the input.

Do you really need a pre amp with your Parasound? Perhaps you could go passive.

/Niclas

[Tieftoener]

Thanks for the reply...

The pot at the input is a fine idea... I just thought it would have lower noise if I kept a full signal from the CD player (or other source) into the FET. It's certainly worth considering I guess...

Then buy two new pots, preferably with no more than 5KOhm, to be placed at the output.

Why is 5kohm limit important? I see that it could reduce gain quite a bit if I got say a 1kohm pot actually. Other than effecting the gain, is this important? Perhaps drawing more current from the FET improves sound? Curious, I just don't know...

Do you really need a pre amp with your Parasound? Perhaps you could go passive.

No, I don't really need a preamp with my Parasound. I have been debating doing up just a simple passive design with a mutliposition input selector switch and just a pot. But, I have had little exposure to pure class A - and I just wanted to build something conceived by Mr. Pass, 'cuz he's the man

I just wanted to get something working - simple, one imput for now. Then I wanted to build another version with multiple inputs to be used with my future Aleph project .

Any more suggestions!!?!?!?!?!?!?!?

Thanks!

[stappvargen]

Why is 5kohm limit important? I see that it could reduce gain quite a bit if I got say a 1kohm pot actually. Other than effecting the gain, is this important? Perhaps drawing more current from the FET improves sound? Curious, I just don't know...

Maybe the 5KOhm pot is chosen because of the Zens low input impedance (4.75KOhm), which it is intended to drive. According to Mr.Pass's article on the Zen:

The input impedance is 4.75 KOhms (on the Zen), and the gain is about 8.5dB. This means that the amplifier must be driven by an active source capable of delivering 3.5 volts at 700 mircoamps.

If you have a high output impedance on your pre-amp it will be highly influenced by a low input impedance of the source it must drive. Maybe it doesn't matter in your case since your input impedance is 33KOhm. There are problably other factors to consider, such as low and high frequencey cutoff and distortion, but I leave that to the gurus.

Sincerely yours,
Niclas