Jfet BOZ

I wired up a PtP version with 30V, a 6ohm source resistor and a 1.2k drain resistor.
I connected a 600ohm headphone, one channel and I'm currently building the other channel to be able to listen properly.
FET is the BL version and bias is 7.3mA.
The gain is more than adequate for normal listening levels.

/Hugo
 
Nelson Pass said:
Since the question has come up and not answered, the
situation with the gain, bias, and resistor values is as
follows.

The device is self biased, and we want the highest value of
bias possible for lowest distortion which makes us want to
use a low value Source resistor to ground.

At the same time, it would be nice to not have too much gain
by virtue of that Source resistor value. The apparent resistance
of the JFET is about 50 ohms, so the gain into a 10 K ohm load is
about 1000 / 60 which is about 24 dB.

We can get lower gain, but the distortion will increase - it's
all a trade-off. I didn't make any effort to exactly optimize these
values - they were just thumbnail calculations.

:cool:


thanks for information
... I was wondering.
But you know these JFET and MOS better than me, Nelson
... for the time being. (what is your personal interest, you will know about)


I have had only some practical experience of
when the JFET gain was in order of 0.5 x gain of one Bipolar TO-92.

In this case you tell the gain is like 16.67, where a bipolar would have ~90.00
which makes as low a coefficient as: ~ 0.2 :att'n:

I understand this is dependent on several things, and most of which actual JFET we would use.
Very much Like when we would use different TUBES, in same sort of setup.
-----------------------------------------


x16, +24dB, is quite alright gain for a preamplifier.
It will be more allround amp
,
especially when signal comes from Sources
with lower output voltage than CD-players.
Say IPOD, MP3 and whatever they use nowadays.


Sorry, if I confused anyone here.
I just want to be supportive.
Now, we know a bit more, in this special case.


Thanks
lineup
:)
 
lovechild said:


Ahem ... 1000/60 would be more like 16.6667
... but I got no clue about calculating audio circuits, so forgive me if my comment is stupid :confused:

cheers, LC

You are 100% correct.

Now in these circuits, the actual values are never 100%
the '1000' bit maybe be +/- 5%
the '60' bit, may be +/- 10 %

so we should need to use, (in 99 cases of 100 in audio circuits),
~ xxx
Just about anything we can imagine or can not think of
could effect the circuit a bit, more or less

so mathematically correct would be to say:
~16
... or say:
Voltage gain from input to output will be like 16, give or take some 15%

On paper we could get some figures.
In reality we can measure the Actual Case

For example one transistor marked BC550C
can have a HFE Gain between 420-800
This is what the specification from the manufacturer tells!

is like +/-30 %

Same with one transistor JFET like 2SK170.
It is wide spread tolerances we talk about.

We can make a good approximation on paper, in schematic.
If we, like Nelson Pass, has got many years experience of Audio Reality we can make a fair guess. But still is a guess.


Regards
lineup
:)
 
Hi,

I just want to thank Nelson once again for this precious gift. This preamp is just great. No,I didn't build it yet... Even though I'd like to do it this weekend, there are some priorities...(Secret revealed: I am getting married on 1st of July!!!). So I will consider this preamp as a wedding gift from Nelson! ;)

Best Regards

Vix

p.s. Steen, where are you? Secretly building it? :clown: :D
 
Zen Mod said:



congrats!!

hehe ,soon you'll be :bawling: like we all (ok- majority of us)....

kiddin' off course .........


Thanks! Well, since we agreed that the amps and speakers are going to stay, I hope it won't be that bad :D

(She didn't have anything against a set of BIBs either, but an OB with B200 and 15" woofer will have to be negotiated; New Jfet BoZ from Papa is considered to be an important ingredient to achieving better sound quality-increasing the chances that She will allow even an OB :clown: )

p.s. There are no grills on the speakers yet. I guess I've got enough time to solve a potential future problem.... (kidsfingersondustcaps)

Regards,

Vix
 
Done.
:)
Oh I forgot drilling.
:bawling:
Manu.
 

Attachments

  • bozjfet.jpg
    bozjfet.jpg
    99.6 KB · Views: 5,594
lovechild said:

I knew there's a hidden factor (or at least hidden to me), wherever it's coming from. :xeye:

thx!

20 x log(100/6) = ??? (dB gain)
where dB is the level, the volume we actually are supposed to hear

( deci - Bell, Bell is probably the same man who invented telephone , which was the first time we could HEAR, in very long distances.
before this was 'Morse' another Man, who made some way to communicate via electrical impulses, Short and Long. = Telegrafi.
But this did not invlove actial sounds)

correct formula to get the dB, in this case:
1. Where Rs = 10 Ohm (source resistor)
2. Where Rjfet ~50 Ohm ( an imaginary resistor, representing the JFET characteristic.
3. Where Rd ~1000 Ohm (actually 991 Ohm, like )

( the paralleled value of 2k2//2k2//load, where load is selected to be 10k, which is some 'worst case' in put impedance of a 'thought up' power amplifier )

Now the combined load resistance as seen at junction of DRAIN = output
is paralleled values of: 2k2 resistor 2k2 resistor and this 10k Ohm.
Formula for getting this drain load into one value:

Rd =

1
------------------------------- ~ 991 Ohm
1 1 1
---- + ----- + ------
2200 2200 10000

If you go back to my first post, I think I mention this value. 900 Ohm something.


Now voltage gain is roughly Rd / (Rs + Rjfet)

We get 991/ ( 10 + 50 ) = 991/60 (which is almost 100/6)
Gives ~ 16.51 times higher voltage at output, compared to input
if the power amplifier side has 10 kOhm input impedance.

GainV = 16.51
now, people can not hear Voltage. Like with morse telegrafi this just electrical impulses.
They have to be translated into Air pressure Changes, for example in a Loudspeaker.
And how strong a Sound is ( SPL = sound pressure level )
is measured in deciBel ...
I am beginning to think this has nothing to do with Alexander Graham Bell. There must have been a french guy called Bel, or something :D whpo was heavily into the business of exploring sounds.



Now the relationship between VoltageGain magnitude and sound increase level
is this formula, as been posted by friend Zen Mod:

Gain in dB = 20 x { Log10( GainV ) }

GainV = 16.51
gives Log10 ( 16.51 ) ~ 1.217
x 20 = 24.35 dB

The amplifier will have a gain of 24 dB
.. or more correct -24.35 dB
... negative sign to tell that the signal at Drain, will be opposite Phase of input.

We call this type of amplifier Inverting Amplifier.

If we use 2 transistors in same configuration as JFET BOZ here,
and let it pass both in series, we get the signal back to normal.

The combined gain of using 2 modules after each other would:
- - (24.35+24.35) = +48.70 dB

As two - make one +

for Decibel you simply add the gain.
And expressed in VoltageGain, you multiply.




Therefor 16.51 x 16.51 = +48.70 dB
A gainV of 272.58 !

Put +0.005 Volt (5mV) at input and get +1.36 Volt at output!
This combined 2 modules amp would be suitable as a Microphone .. at least regarding GainV.




lineup :) do not claim this article to be fit for a Book on the subject. Far from :D
But you get the picture, if you can dis-regard the use of terminology, in some lines.
... it is not that easy all the time, you know,
when youre bound in writing/thinking fast .. in a FOREIGN Language:
like British english