6k7g/6j7g
brent
uses a 5u4g rectifier and a pair of 6k7g pentodes.The manufacturer mentioned
I could use 6j7g in place of the 6k7g. I did switch the tubes and the
differance in sound was signifigantly for the better.Seemed to tune the amp
right in--focus it in.I am not knowledgeable about electronics and specs but
maybe someone who is could tell me the qualities of a 6j7g that would make
it better than a 6k7g?
Thanks
Patrick Turner
brent wrote:
The 6J7 is a "sharp cut off" small signal pentode.
Its transfer curve is much straighter than a 6K7,
which is a "remote cut off pentode".
The remote cut off tube was invented to be used as a
variable mu tube, since the spacing of the grid wires
is not a constant pitch.
The result is that when the grid bias goes negative, the gain of the tube
reduces,
and this is useful in radio sets where the signal is small,
but the amount of amplification is controlled by the DV bias
at grid1.
Used as a driver for the 2A3, the voltage output required is consideable,
and the maximum peak voltage is equal to the bias voltage on the 2A3.
Used this way, the 6K7 will have considerable 2H distortion, which in theory
should cancel that produced in the 2A3.
With the 6J7, the 2H is a lot less, and the amount of 2H cancellation is small.
This don't much explain what you hear.
At low levels of operation, the cancelations of 2H in the amp may be negligible,
and the extra linearity of the 6J7 is making it sound better.
If I had any choice, I would use the 6J7.
I have a few, and I might build a retro looking amp with them one day,
with all top capped tubes, and with 807 at the output.
The 6J7 is one of a line of pentodes which evolved from the first
pentodes, and they got better and better as time went on,
and as the RF and IF frequencies went higher.
The 6J7 is a fairly low Gm tube, only 1.2 mA/V,
but sufficient to get a gain of around 100 into a load of 100k.
From the 6J7 came the 6SJ7, Gm 1.7 mA/V,
6SH7, Gm 4.9mA/V, and so on.
On page 511 in the RDH4, it says that a 6SJ7 has less thd
up to 30 volts of output and above this the thd rapidly increases
above that produced by the same tube connected as a triode.
The 6SJ7 in pentode appears to make less than 1% thd at 10 vrms output,
but around 2% in triode.
To find out which tube, or which mode sounds better,
try what you can, but its only by measuring that you would know
what the numbers really are.
A 6J5 triode may be more linear than a 6J7 pentode.
I have never used pentodes in signal amps, such as a line level amp,
because the gain is quite high, often over 150, so
one is forced to use feedback to reduce it, and
its simpler to use a 12AU7 with low gain, and not have to use any shunt FB.
In a phono amp the pentode noise isn't nice, and triodes are better.
In power amps, the EF86 became common choice for input tube,
used in Mullard 510,520, and Quad II.
But before that became popular, the 6J7 was often used,
sometimes with signal feedback applied from one of
the output stage anodes to the screen grid of the input pentode.
Patrick Turner.
Fabio Berutti
original metal/ceramic envelope). I'm not sure if it will be "quiet"
enough, but from its data sheet the 6J7 seems a good tube for audio, not so
different from the widely used (in phono stages, too) EF86.
As a driving stage as in Your case I'd go for the 6J7 triode-strapped. Low
distortion, partly cancelled, but - most of it - MUCH lower output
impedance. The 2A3 triode THEORETICALLY does not draw grid current until
Vg=0, but all DHTs in fact "suck" current from the driving stage which
"slows down" unde charge. The final result is a combination of ALL
intermediate stages...
Ciao
Fabio
"brent" <brn...@home.com> ha scritto nel messaggio
news:9AkBb.620663$pl3.237288@pd7tw3no...
Patrick Turner
Fabio Berutti wrote:
> I'll be trying to make a phono (MM) preamp using 6J7s (old Svetlanas with
> original metal/ceramic envelope). I'm not sure if it will be "quiet"
> enough, but from its data sheet the 6J7 seems a good tube for audio, not so
> different from the widely used (in phono stages, too) EF86.
I'd strongly recommend you use a well done DC heater supply.
The EF86 has special low hum heaters, but not the 6J7.
Leak and Quad both used the EF86 for MM carts, and they
are barely quiet enough, and same perhaps could be said of the 6J7.
And then you have to think of what sort of RIAA eq you wish to use.
It was often done with a shunt FB network around the pentode tube,
which helpd reduce screen noise on the pentode.
One could use a pair of 6J7 cascaded, then have another as a trioded
cathode follower output buffer and have a series voltage feedback RIAA network.
You should have the external metal shields fitted to the 6J7
grounded to the chassis.
Patrick Turner.
Yves Monmagnon
"Patrick Turner" <in...@turneraudio.com.au> wrote in message
news:3FD735ED...@turneraudio.com.au...
>
>
> Fabio Berutti wrote:
>
> > I'll be trying to make a phono (MM) preamp using 6J7s (old Svetlanas
with
> > original metal/ceramic envelope). I'm not sure if it will be "quiet"
> > enough, but from its data sheet the 6J7 seems a good tube for audio, not
so
> > different from the widely used (in phono stages, too) EF86.
>
> I'd strongly recommend you use a well done DC heater supply.
> The EF86 has special low hum heaters, but not the 6J7.
>
> Leak and Quad both used the EF86 for MM carts, and they
> are barely quiet enough, and same perhaps could be said of the 6J7.
>
> And then you have to think of what sort of RIAA eq you wish to use.
> It was often done with a shunt FB network around the pentode tube,
> which helpd reduce screen noise on the pentode.
> One could use a pair of 6J7 cascaded, then have another as a trioded
> cathode follower output buffer and have a series voltage feedback RIAA
network.
> You should have the external metal shields fitted to the 6J7
> grounded to the chassis.
I do like the RDH4 idea to ground the plate so that it acts as a shield,
taking output
signal from the screen.
Did you try ?
Yves.
John Byrns
<fabio....@libero.it> wrote:
> I'll be trying to make a phono (MM) preamp using 6J7s (old Svetlanas with
> original metal/ceramic envelope). I'm not sure if it will be "quiet"
> enough, but from its data sheet the 6J7 seems a good tube for audio, not so
> different from the widely used (in phono stages, too) EF86.
AMPEX used 6J7s in the first stage of some of their tape replay
preamplifiers, to amplify the output of the tape head. The play only 350s
that were in the main control room of the radio station where I worked
used 6J7s in this capacity, I believe the heaters were supplied with DC.
RCA also used the 6J7, well actually a premium version with a four digit
type number which I forget, as the input stage in many of their microphone
preamps, and the heaters were run from AC, they seem to have worked fine
even with those old low output RCA ribbon mikes.
Regards,
John Byrns
Surf my web pages at, http://users.rcn.com/jbyrns/
Patrick Turner
Yves Monmagnon wrote:
This makes the tube into a low power triode.
RDH4 says it has lower noise, presumably shot noise,
which is elecron noise hitting the anode.
But still you are left with grid input noise, which dominates the
noiise production of a phono stage triode, and since the 6J7
is a low Gm tube, the triode noise won't be any better than 1/2 of 12AX7.
> Did you try ?
I did once, but I wasn't able to get any great noise reduction.
But RDH4 recommends it for a microphone amp.
In my phono amp, I use a j-fet, the 2SK369 as a cascode driver for 6EJ7, which
is a much later made frame grid pentode, which has much
more Gm than a 6J7, and a mu of around 60 as a triode.
The noise is low enough to use MC with no step up tranny.
The gain of the casode stage where the Gm of the fet is much higher than the
tube
is simply fet Gm x RL of the top tube being cathode driven,
so if the RL = 16k, as it does in my passive RIAA filter,
and the fet Gm = 40 mA/V, the gain is 640,
so a 0.3 mV input signal gets amplified to 192 mV
at the input to the RIAA filter,
then the filter reduces it to about 21 mV at 1 kHz, then
the next bootstrapped follower amplifies it 45 times using a 12AT7
to 945 mV, so the output is around a volt from the phono amp,
and no further gain stage is required for the signal before it is sent to the
gain pot and the
power amp.
The 6EJ7 could be any twin triode as long as 5 mA of idle current flows,
since that's the ideal operating current for the fet.
12AT7, 6DJ8, 12AU7 are all OK, and there
is very little change in cascode gain, because the gain still remains
Gm x RL of the tube.
A 12AU7 or 6DJ8 could be used as the bootstrapped follower
second phono stage, if less gain is required.
Tyipical input noise with a j-fet in cascode with a tube is
12 to 20 dB less than a 12AX7.
I have not tried the 6C45pi high Gm triode yet, the tubes sit there
waiting for a test, but I ain't had time to get around to it.
Patrick Turner.