Thread: 6CA7 in AB2.
View Single Post
  #3   Report Post  
Patrick Turner
 
Posts: n/a
Default



john stewart wrote:

Patrick Turner wrote:

Anyone have any experience in driving 6CA7
in either tetrode, triode, or UL, into class AB2?
The 6CA7 is supposed to be equal to EL34.
EL34 don't much like AB2; but since the
6CA7 is a tetrode, not a pentode, perhaps, like the 6L6,
AB2 operation is perhaps fine.

What sayest thou, and they of years challenged,
and who accompanied history's beginning, and who must
have memory of a bygone age?
What noble efforts at AB2 were sustained,
leading to much music well listened to?

Patrick Turner.


See ABPR for the attachments referenced in the following.

Hi Pat- Never ran a 6CA7 / EL34 in AB2 but last year built an
experimental 6V6GT rig running AB2. It managed 26 watts in burst
mode & 19 watts continuous after the PS sagged to steady state.
The 6V6 grids are driven by a PP 6BQ7 CF. That helps to minimize
distortion as a result of G1 current.


Indeed, I know about the drive requirements needing to be CF, and low
impedance,
because the grid I reduces R in to the output tubes.


IMO should be OK for a music amp since the grids will be driven
+ve only on peaks which are but a small percentage of the time.
See the attachment for test results. This all part of an article on
Amplifier Burst Testing to appear in the November issue of
AudioXpress magazine.


I wanted to try 6CA7 in a circuit meant for KT88.
The limitation, especially in triode mode, is the Vg = 0V line.
With say 807, you often see the data curves including for a grid swing
up to +15v, allowing 30 watts AB2.



Also built an AB2 amp using a PP pair of 33 two volt pentodes.
It runs triode or UL. The driver is a triode connected 33, which
in turn is transformer coupled to the following grids.
That one is published in the "Glass Audio Projects" book from
AudioXpress.com.

The attached graph is a composite
shewing the drive signal to the grids (LHS with X10 Probe) &
the resulting G1 current (RHS) measured across a 10R sampling
resistor in the grid circuit. This one manages 4.85 W in UL &
2.5 W in Triode mode.

Cheers, John Stewart


I couldn't see any item under your name at abpr,
maybe the currawongs and magpies haven't flown in yet with the info.

The data curves for 6CA7 are usually given as EL34/6CA7,
and there is a difference, the former is a beam tetrode, the latter is a
real
pentode, and the couple of 6CA7 I do have in operation ARE different to
EL34,
with lower Ra and gain, but higher Pd.
I like the 6CA7 more than the EL34, and they sound well.

If I buy 50 of these now from New Sensor, they will only cost US $280.

The driver for the AB2 amp will have a choke with CT grounded,
and each end going to the grids of the output tubes.
The cathode circuits might be a new AC regulated cathode bias type,
which has all the benefits of excellent self regulation of the cathode
bias R,
but when going into class AB, the cathode cap won't charge up
due to the presence of one transistor bypassing the high positive peak
currents
seen in the cathode circuit, in all AB amps.
The choke won't allow the grid bias to change, as it does with a normal
RC coupled stage driven into grid current, because of the charge up in
the
coupling caps.
The charging up of cathode biasing and the grid coupling cap
both cause a tube amp to become paralysed when pushed a lot.

The output stage will be acoustical, so the CFB will INCREASE
the input impedance, despite the grid current, in the same way as in a
McIntosh.
Thus the AB2 amp is possible, and the extra headroom good for dynamics
such as loud drumbeats, where the slightly higher distortion after going
from
class AB1 into AB2 don't matter.
The acoustical prevents severe distortion developing in the driver LTP
stage,
because the voltage demanded still isn't too high for RC coupling.
Probably, I will have a direct coupled CF driving the choke, and then
the gain LTP can be a 6CG7 which is RC coupled to the CF.
Then the usual SE input stage.

Two 6CA7 work OK into a 6.6 k RL, without much AB2,
with B+ = 450v, and 10k is really nice.
But if the load has a dip to 3.3k, then the voltage
swing is limited, even in tetrode mode.
I was just after a little bit more performance.

Patrick Turner.