Croft Charisma Pre-Amp
|
 |
This modern pre-amplifier
arrived in March 2025 for an "overhaul". Although people
tend to be quite excited by it I found it was a bit amateurish
in build quality. Then again, I think the same of Roberts radios
which are pretty clever designwise. This example hasn't got the
build quality that one might expect from going market prices
because all solder joints are of the "just held and touch
soldered" style, although testing a component takes only
seconds because of this. Wiring is neat but wire diameters pretty
small and many wires running close together are uninsulated. |
|
 |
Valves are unusual, having
"high quality" coding (even the voltage regulator).
Resistors are a mixture of exceedingly high tolerance types,
wirewound ceramic and standard 4-ring types. Presumably these
relatively expensive parts are used to keep overall operating
tolerances to a minimum although drift from design figures due
to valve ageing might make this aspect a waste of time and money
(static voltage measurements proved this to be indeed the case).
A circuit diagram is hard to
find so I've drawn one up from my examination of the thing, below.
Oddly there are a few design anomalies. the heater supply is
a regulated 12 volts which measured 11.94 volts in this example
when the filament voltage should ideally be 12.6 volts. The addition
of an extra diode or two and this could be fixed. Another mistake
is to have missed a decoupling capacitor at the DC heater supply.
In addition, although very high tolerance resistors are used,
there are a few examples of +/- 5% coupling capacitors, both
100nF and 1uF type MKP1841 which defeats the object when it comes
to precise audio balance.
|
|
 |
Below is the power supply
schematic. The rectified HT can be a trifle high considering
the regulated voltage level of only 85 volts. A separate stabiliser
heater voltage is used because of the elevated cathode voltage
at V7a. Strictly speaking, no capacitor is required at the output
of the LM340T12 but surprisingly I found no decouplers in the
heater circuit at all. |
|
 |
The pre-amp seems to be
OK electrically. The only static powered test I did was to measure
the grid voltages at V1b, V3b, V4a and V4b. I found V1b was sitting
at greater than 125mV whilst the others were close to zero. I
had a high quality "7025A" which is identical to the
12AX7 except it uses a "long" anode against the 12AX7WAs
"short" anode so, in the interests of balancing audio,
I swapped the new 7025A with V4 (for the reasoning see the way
the triodes are arranged in the circuit diagram above) and put
V4 in the place of V1.
I'll be testing audio performance
in a day or two. Scroll down to see the results.
|
|
|
A few comments on this
pre-amplifier. When I started looking around the Net for details
on it I became absorbed in the comments. It seems it's not a
static design or that it suddenly appeared in the form above.
For example this one uses (in its audio section at least) two
types of valve viz. 12AU7WA and 12AX7WA. These are relatively
new versions of the 12AU7 (1951) and 12AX7 (1954) which are basically
the same as the "British/European" coded ECC82 and
ECC83. If we go back in time before those B9A all-glass based
valves were introduced the equivalents of theSE were the 6SN7GT
(1941) and 6SL7GT (1942), the short tubular style or without
the "GT" being the bulbous style envelopes. The 6SL7
in turn was a development the earlier 6SC7. One distinguishing
difference between the various types is the heater voltage/current
so for example the newer ones can use either a 6 volt or a 12
volt power supply (Pin 9 is a heater centre tap... which isn't
used in this equipment.. but allowing for a 6 volt supply). The
newer of the valves mentioned are designed to be run from either
6.3 volts 300mA or 12.6 volts 150mA. This can be DC or RMS AC.
Generally speaking the emission of a valve cathode reduces with
use and if the heater is run at a reduced voltage then ageing
will be accelerated.
I guess the important question
is whether or not the choice of valve has any effect on audio
fidelity? Well, it certainly does, but many other things come
into play also. Ideally whatever sound is listened to should
have been recorded and subsequently reproduced to give one a
precise copy of the original sound. When one listens to a sound
and turns up or turns down the level it should get louder or
quieter but should have the same characteristics as the original.
The end result will depend on loads of things but I'm afraid
the valves used in a pre-amplifier will not figure much if they're
doing a proper job. I guess, for a pre-amplifier such as this
the most noticeable weakness will be hum or sound distortion.
You'll be able to find examples of the "Charisma" pre-amp
using the earlier valves I mentioned above but why were these
used exactly? You'll also find comments about the make of volume
controls and strange comments regarding differences in sound
quality. My comment on this sort of thing is that the volume
of left and right sound channels can be independently twiddled
so why bother with careful matching and high tolerance resistors?
Finally, if true sound reproduction
is really important, then forget about FM radio which is a mish-mash
of processing, CDs which use bandwidth restricted digital data,
cassette tapes which are subject to recording media failings
and Dolby processing and vinyl records which rely on mechanical
interaction between a needle and bumpy plastic.
|
|
Test results:
Tests were pretty basic with
an audio generator driving L & R inputs in parallel with
a twin channel scope at either L or R output and the test input.
I was using a direct coax lead with no attenuator on the signal
generator for Blue Channel 2. For Yellow amplifier output Channel
1 used a 10dB probe which catered for the voltage readings below.
An audio level of 1 volt is
considered to be 0dBV, with a domestic standard of -10dBV representing
316mV.
A VU meter is generally said
to read "0" for 775mV RMS so tests were carried out
using voltages slightly in excess of VU zero.
Both L & R volume controls
set at max for all tests
Despite its physical size and
overall complexity the amplifier is using only a couple of triodes
for amplification of Tuner, CD and Tape with an extra three triodes
for additional amplification of a gramophone pickup.
|
|
 |
 |
Tuner input 1Hz Left output
Gain 1.38/0.347=4
|
Tuner input 1Hz Right output
Gain 1.39/0.346=4
|
 |
 |
Tuner input 25Hz Left output
Gain 5.55/1.31=4.2
|
Tuner input 25Hz Right output
Gain 5.52/1.31=4.2
|
 |
 |
Tuner input 1KHz Left output
Gain 5.68/1.34=4.2
|
Tuner input 10KHz Right output
Gain 6.03/1.35=4.5
|
 |
 |
Tuner input 20KHz Left output
Gain 6.04/1.35=4.5
|
Tuner input 50KHz Right output
Gain 5.80/1.35=4.3
|
 |
 |
Tuner input 100KHz Right output
Gain 4.92/1.31=3.8
|
Tuner input 200KHz Left output
Gain 3.29/1.25=2.6
|
 |
 |
Tuner input 1KHz Right output max just before
clipping
Gain 22.6/5.44=4.2
|
Pickup input 1KHz Right
Gain 39.6/.0795=498
|
 |
 |
Tuner input open. Left output
Unterminated input socket residual 50Hz hum of 65mV.
Note that HT hum would be 100Hz so this reading is
from stray earth currents in the test environment plus possible
microphony in V4/V5 valve electrodes from the 50Hz mains transformer.
|
Tuner input open. Right output
Unterminated input socket residual 50Hz hum of 46mV.
|
|
All the tests were carried
out using a single test tone at various frequencies and levels.
I didn't attempt to carry out tests using simultaneous test tones
or to check linearity or harmonics. Neither did I attempt to
listen to the results, merely relying on oscilloscope traces.
I didn't bother to simultaneously compare the left and right
channels because there are twin volume controls making that exercise
irrelevant. Finally, although improvements could have been made
(eg the valve heater circuitry), I left the equipment as originally
manufactured. |
|
|