| 
H.H. Scott
TYPE 312D FM Tuner
The 312D under view here is a late model that can be used
in Europe with the flick of a switch. You need only rewire the attached cord,
or replace it with the appropriate escutcheon. As a late model, it is also equipped
with the dual 6 pole IF, which was later used in the 433, and the premier line
post-reformation in 1974. One of the tests that I subjected this unit to was a
direct comparison to the 348B, which is from start to finish, a 312D, early model.
Neither tuner was a candidate for adjustments, but both were checked closely when
I received them, and both tuners were, by this process, taken out of alignment
slightly, and then put back into alignment. This can often produce significant
results, but offered none to the early version 312D, and only a marginal level
of improvement to the late model. I was surprised when the early
312D, with its old fashioned IF strip, was the better tuner, since I had high
expectations for the latter model's IF stage. It isn't possible to do much to
the IF on the late model. Like the 342C, only the
first IF is adjustable. You can do it without too much trouble using the meter.
The tuner will only detect a signal when the first IF is beaming one across its
primary and secondary transformers. The 2nd IF is the first of two precalibrated
ceramic IF transformers. If you take the module apart, you can verify for yourself
that it has an adjustment point for each crystal assembly. It is a very expensive
version to the horribly nonadjustable ceramic filter that you will find in application
everywhere today. Although no amount of alignment put the late
model at the plateau that the early unit had established, both were about equivalent
in performance on DX tests. This indicates that the front ends were performing
identically. All other factors being the same, I knew that the nonadjustability
of the crystal circuits might be the Achilles heel of the design in the long run,
where alignment is concerned. My experience with the two other tuners (433, R74S)
that make use of the 12 pole IF suggests that my late model 312D may have slipped
in performance in a way that I have no solution to, and that the circuit should
have offered better performance. In retrospect, the old fashioned IF strip, found
in the early 312D, may be the best bet for an alignment. Properly set up, these
tuners should provide 1.7 uv minimum performance in mono for 35 dB of quieting.
Scott 312D early model on top. Final unit shown below. There
is a topological history to the Scott 312 series. The
most interesting front end, shared with the 4312, used nuvistor tubes. The most
interesting IF was found in the ultimate 312D, and if optimal, would have lended
to it a bit more of a performance edge than could be expected from its predecessor.
The 12 pole IF found in the 312D ultimate was only used in circuits for a short
while. The later version found in the 433 and the R74S both outdistance the earlier
version found here (or something is wrong, as I suspect). Although I no longer
own a 433, I did enough tests with a newly received R74S to safely conclude that
the 312D that I own is not as good as it gets. I may try to replace all critical
carbon resistors, and the filter capacitors in the hope that the better precision
afforded by new parts will bring the unit to where it should be. If
you make the assumption that all of the components of your tuner are functioning
well, but that its performance is suboptimal, then you can proceed to attempt
to improve its performance by aligning it, using instruments, or with some tuners,
like Scotts, by utilizing the meter method. On the 312D shown on top, it's simply
a matter of obtaining maximum interstation noise on a quiet portion of the band,
which will be indicated on the signal strength meter as the highest value within
the portion of the band that you are measuring. Between the amplitude that you
can hear, and the position of the meter, you can do well by this method alone.
The 312D first IF sits on the front end. It detects a signal
centered on 10.7 MHz that the front end supplies. The value that provides the
best overall performance, be it 10.7 MHz or not, will be predicated by the settings
of two IF stages, a limiter, and the detector. All of these parameters can be
all out of adjustment to the same degree, and the detector will do its job, which
will be the production of a skewed wave with what you thought was the right amount
of amplitude. But if things get too far away from the design-center values, the
result will be no stereo, or peculiar behavior on parts of the band. Tyros who
spin the values of these transformers may present you with a totally messed up
tuner that has gone antisocial on your favorite stations.
|
|
Since the 312D allows only limiter-detector adjustments, and the
aforementioned 1st IF, Mister Tyro can't get too far from the mark before things
are obviously wrong on the detector, and the limiter's functional range is sharp,
and since the 2nd IF can only function when the 1st IF is set up within a very
narrow range, no problems here. All indicators must point to 10.7 MHz, in a manner
of speaking. And that, in a nutshell is the advantage of the crystal circuit.
| I indicated earlier that
I believed that the front ends of the two tuners that I was comparing were more-or-less
spot on. Both offered similar stereo performance, and this is no surprise, since
both use the same adaptor. Also used in the 367, which is the 388B with a facelift,
the five transistor adaptor seems to offer better performance to its seven semiconductor
predecessors. It's the one to look for. If you find one of these units with the
terrible offshore copy of the five transistor unit, keep on looking. These adaptors
rarely function properly. What I would have liked to have seen in the ultimate
312D is the IC adaptor found in the 342C and later
premium Scotts, instead of the unit that got used in the modest Scottie.
In incremental terms, the IC adaptor is about 10-15% better. That's plenty, and
I doubt that many readers own a tuner with an adaptor as good as the Scott IC
model. What you get in the 312D instead is a late version of the time-switching
multiplex design that Scott first introduced in 1961. By this point, Scott had
the design down pat. It was only in 1971 that Scott put out a replacement for
the IC adaptor, (it ran out of IC's and started to reserve the IC adaptor to top
line units only) which had, count 'em, 15 transistors. It was the discrete design
that gave some insight into the complexity of the IC design. At this point, Scott
took the entire process apart, adding amplification to composite signals that
previously had gone through the entire circuit with no actual amplification. Early
adaptors used amplification to drive oscillator circuits at the frequencies that
control the detection of stereo. The five transistor adaptor, a classic time
switcher, makes do with five transistors because it does its conversion with the
tiny signal that it gets off the detector. The signal develops across a bridge
fed by one of the amplified signals, a 38 kHz oscillator. After final attenuation
that balances the channels, a miniscule signal is routed to a preamplifier. The
312D offers two separate circuits for this purpose. One supplies a conventional
(and suitable for daisy-chaining) high impedance output, the other, a low impedance
circuit. An additional amplifier supplies the scope output signals. Late model
Scott adaptors use many transistors: Scott placed all of the noise circuits that
feed the limiter in the crystal IF designs on the adaptor board. There was no
room for the noise circuit on the 12 pole IF strips, placed as they were on a
board that hadn't changed size since the 312B. The 342C style 6 pole IF is the
only one IF with some noise circuits on board. Neither of these 312D's have noise
circuits, while every premium Scott that employed crystal IFs, starting with the
342C, had them. I occasionally wonder to what extent the 312D ultimate would benefit
from the later noise circuit design, and whether the addition of an adjustable
circuit similar to that found in the 342C would have turned things around for
this tuner. IC adaptor equipped 433s had the noise circuitry tucked under
the chassis, on a separate board. Noise circuits were controlled by a variable
resistor on the 342C type IF, and this adjustability was refined on the dual pole
circuits that employed the 1971 discrete adaptor on its board. As a guide, when
miscalibrated, the circuit itself lends an audibility most discernible when moving
off and on channel. There should be no discernible noise when moving off and on
channel on a good signal with the control properly set, but there will be an increasingly
evident amount of noise when you tune to center channel, if the circuit is out
of adjustment. At its worst, this simple amplitude noise will always be present
in some measure, even right on channel, when there should be as little noise as
possible, and none at all in circuits that perform very well. You may have to
listen carefully through headphones to detect the circuit malfunctioning. It's
about 20 dB down. A diagram that indicates the evolution of the Scott IF design
is found here. Below is a reduction
of the same diagram. 
The early 312D utilizes a design in its IF that is similar to the intermediate
drawing. Scott used IC's in the place of discrete transistors, but the devices
in schematic were similar, and the cascade circuit is nearly identical. The late
312D is similar to the bottom diagram, except that where there is a coupling capacitor
in the 342C IF, the penultimate 312D places its third
IF, FL302. Note that the Active Tracking Noise Reduction circuits are Scott
Dynaural designs that were never offered. The author prepared the diagram as a
demonstration of how Scott could have used Dynaural RF circuits in its tuners.
Although employed as AF circuits only, the patented designs are all based on RF
work that Scott did for the US military during WWII. The active circuits in the
342C style Ifs were the first attempt to implement some sort of active noise tracking
as part of a tuner design. The additional circuits that are a part of the top
two schematics feed noise forward and use noise shaping and the phase lead present
between the grid and the plate of a tube (or the equivalent terminals of the semiconductor
in use) to supply a static signal of approximately 90 milliseconds that can be
substituted in the place of transient noise. This was a very effective noise reduction
circuit that could have been incorporated into an FM IF as a means to suppress
any amplitude that exceeded the bandwidth permitted to the broadcaster. Ergo,
noise. The penultimate 312D offered better performance than all of my Scott
tuners, save the early 312D and the R74S. The penultimate unit was slightly better
than my 312C and a margin superior again to the 312
and the 312A. I don't own a 310E anymore, but I recall
that it was better than all of my solid state Scotts in terms of performance quality.
I also own a 310C, a 311, a 314, a 330C, and a 370,
None can come close to the performance of the 312D ultimate. Its deficiencies
were noted only by comparison with a model offering the same specifications that
was found to be superior in practice. I can only imagine that the newly produced
dual-crystal IF's outperformed the transformer IF units that preceded it. Since
the 433 is the the "ultimate statement tuner of its time" which like
the 4310 in its time, and was made without regard to
cost, with the latest iteration of the dual crystal and the filter offered the
best hope for the IF with a perfectly on channel signal. It did, and I expect
that it supplied the 1.5 uv advertised. But with a digital tuner that used punchcards
to select a "a truly perfectly on center" signal, my own experience
suggested that a capacitivley tuned front end would be preferable from a reception
standpoint, in my locality, beyond a doubt. But there was no doubt in my mind
that for WNED in Buffalo, a weak signal one hundred miles distant, the 433 had
the quietest channel, even trumping a pair of Citation Elevens that I had in house
for a while. The more modestly priced 431 is a 433 with an analog front end.
I have never owned one, but I expect that it is the very last, and possibly the
very best of the tuners that Scott manufactured. It is the 312E. The 431 was the
last Scott tuner to be equipped with the classic front end. Along with global
use compatibility that I can almost assure, sight unseen, the front end is a symbol
of a classic approach to engineering that dates back to the 310B,
in 1955. Fitted with a front end out of the 382C,
with all of the best circuits thereafter, is the R74S.
These receivers are fairly common, and inexpensive. Like the 312D, the R74S is
switchable to European settings with a couple of switches on the back panel. The
latter is for de emphasis switching from 75 to 50 microseconds. This is... FM.
Considering how versatile the final 312D is, how it can be used anywhere regardless
of local mains provisions, I suggest that you get one if you can. Either of the
312D's is a must-have for the serious DXer. These tuners are so sensitive and
discriminate between signals with such efficacy that you needn't fear your fragile
distant signal being overpowered by off-band distortions from a powerful local
station. What you get is very clear channel delineation that is the product of
a superior front end and a full bodied IF stage that, properly aligned, will always
reject off-band artifacts. |
