|
A brief word of
explanation if you'd like to read on.. This example of an R1155
had been factory modified to cover the Trawler Band; something
which I'd overlooked before discovering and attempting to fix
certain tuning discrepancies. |
|
I've been getting a T1154 transmitter up and running for
use on the amateur bands and I decided to pair it with one of
my R1155 receivers. I have two, of which one has been too far
modified in terms of the front panel, but the second, although
much modified inside, still retains most of the original front
panel. It's a version "E" and has the early tuning
knobs and is the same as a receiver I used to use back in the
mid-1950s. The first task is to examine the chassis and see exactly
what's been done to it and what work is needed to get it running
again. The plan is to get the receiver working with the minimum
of work. Some people would strip the whole thing down and rewire
it, but the result is no longer a 70 plus year old receiver.
Below are some pictures showing the condition at Day One. As
this is the first R1155 I've worked on, the job will be mainly
a learning exercise.
This example has been modified
to produce a working receiver, I'd guess in the 1960s and thankfully
the modifications were done sympathetically to preserve the look
of the receiver. See another
conversion that took no account of this !
Having struggled with the circuit
diagram in a copy of AP1186 I decided to PhotoShop it and produced
a version suitable for anyone attempting to carry out basic refurbishment
of an R1155. This divides the overall circuit diagram into four
parts together with resistor and condenser parts lists. The pages
are A4 printable.
Printable R1155 schematic plus parts lists |
 |
|
Pretty well original,
just missing the 4-pin Jones Plug for the DF function and connector
securing posts. I cleaned up the dial some time ago so it looks
quite acceptable. The headphone jack socket was fitted to a blanking
panel which I removed when I cleaned the dial. |
|
 |
|
Here you can see evidence of fairly
extensive modifications: The DF parts have been removed and a
mains transformer fitted together with a rectifier socket and
smoothing condenser, but no choke. I suspect this was there but
had been removed as there was evidence of cut wires at the smoothing
condenser tags and no continuity from the rectifier to the receiver's
HT line. |
|
 |
|
Underneath, the RF compartment
is missing its cover. Apparently added is the black transformer..
see later. |
|
The added mains transformer |
|
|
Missing V6, also the metal
cans for V8 & V9. V6 should be a VR100, equivalent to a KTW62,
which is what's known by some as a high gain 6K7. I didn't immediately
have one so I fitted a KTW63 which has a slightly lower gain. |
|
 |
|
Left the added rectifier socket located
in the original location of V2, a valve used in the DF section. |
|
Above you can see the
receiver which is a version type "E" carries the label
showing No.1 Signals Depot. This was located at West Drayton
during WW2. |
|
|
|
It seems there's not too
much to do. I'll remove the mains power supply, fit any missing
or non-functioning valves, tidy up the Jones Plug area and check
the audio output circuitry so it reflects the original design.
Then power it up and sort out any problems then maybe make a
new screen for the RF compartment. |
|
Removing the mains transformer
was a bit tricky because two securing screws were inside the
RF compartment and I needed to detach the row of trimmer condensers
C57-C61 to get at them. Looking further I spotted that V9 looked
odd and removing it showed it was marked 6K6GT and the penny
dropped. The mods included loudspeaker output. The jack socket
wasn't for headphones, it was for a loudspeaker. I decided to
leave this modification in place and maybe fit the jack socket
to the left of the two 8-way Jones Plugs.
|
Right.. the old transformer which
I may use later for an external PSU. This view is the underside
of the transformer which was fitted using long 4BA screws each
with a pair of brass spacers to keep the tags clear of the chassis
Below: Empty space after removing the
power supply. |
 |
 |
|
Right.. the 6K6GT
where V9 should be..
Below.. parts from the old power supply |
 |
 |
|
I also found that switch S1 at
the top right corner had been rewired as a mains switch. The
other two DF switches were in place but had been disconnected.
See an article written in 1946 which shows a typical R1155 conversion
for amateur radio use. Wireless World
1946
The article is useful as it shows a
nicely simplified circuit diagram of the receiver and only the
circuitry of interest. I reckon there's a mistake in the power
supply and output circuit. The circuit diagram shows a KT63 without
an output transformer and, as most people with a knowledge of
valves will realise, the screen of the valve will draw excessive
current and ruin the valve if the anode is left open circuit.
This article may well have been the one used by the chap that
modified my example. |
 |
This looks rather odd doesn't
it....
To make room for the mains transformer
this can has been detached from the chassis and lifted upwards
and sideways so it's now situated above the tuning condenser.
A result of lateral thinking.
It carries filter components to notch
out any commercial stations operating close to the IF of 560Kc/s.
I'm not too sure about the unscreened lead which may introduce
broadcast signals. |
To check heater continuity I
connected a 6-volt battery to pins 3 & 4 of Jones Plug P1.
The valve heaters came on. I'll also check the HT connection
(the easiest way is a smoke test). As always, when tackling a
receiver with an unknown history the first step is to confirm
the viability of proceeding. Any complicated mechanical parts
need to be serviceable and the yaxley switches need to work properly.
If all is well I'll proceed to swap any bad components, then
align the IF strip and front end. So far everything looks OK,
although I'm a bit unhappy about the rubber covered wire which
is used in the set. Some of this is brittle and liable to shed
insulation and short out. This in itself may not be a show-stopper
because lots of receivers in the 20s and 30s were wired up using
bare wire.
Next, I made up a cable using
an 8-way Jones socket and plugged it into P1, put the receiver
on its right hand end, and connected the cable to my home
brew power supply. The connections reflect those in the T1154
into which the cable from P1 usually terminates.
The pin numbers are marked on
the Jones socket body and are labelled: 1: HF Aerial; 2: Trailing
Aerial; 3: LT+; 4: LT-; 5: 220V Rx HT+; 6: Headphones; 7: 220V
HT+; 8: 220V HT-. As I understand the position, there's no reason
why the R1155 can't use 6.3V AC for it's valve heaters, but this
voltage has to be DC in the T1154. Receivers using an internal
home-brew PSU invariably used AC for the heaters so are not entirely
compatible with the T1154. The 220V HT supply at Pin 5 is switched
to the R1155 from the T1154 as is the DC heater voltage which
is routed via the Mode Switch on the T1154. The HT connection
at Pin 7, is unswitched by the T1154 and is provided for using
the DF function which is independent of the transmitter. Another
point of note is connection of HT negative, As with the 1200V
HT negative supply to the T1154, the 220V HT negative of the
R1155 provides bias supplies and is therefore not directly connected
to the chassis. This means that the two supply voltages cannot
share the same transformer secondary winding if full wave rectification
with centre-tap return is used.
If I continue to make changes
to my home brew PSU I'll add a 220V independent output.
Initially the heaters didn't
come on but that turned out to be oxide on the plug pins. Setting
the HT to 230 volts, I connected the HT wire. The loudspeaker
burst into life with lots of white noise so something's working.
Turning the wavechange switch didn't produce any crackling and
after 30 seconds smoke began to rise from the top of the chassis
from somewhere around the wavechange switch gear wheels. HT current
was showing as around 100mA, which is a bit high so maybe a leaky
condenser and a burning resistor? Thinking back... does this
fault explain the missing HT choke? It may have gone open circuit
when faced with lots of HT current. Did the owner get as far
as diagnosing the open choke, then removed it and didn't get
any further?
|
The origin of smoke
in any unrestored R1155 will be this HT negative feed resistor
designated "R1"
|
|
I moved the HT input from my
home brew PSU to one made by Solartron. This let me raise the
HT from zero and control the current resulting in the smoke.
I found the resistor responsible was R1 which is fitted above
chassis on top of the block condenser C1/C92/C94. The manual
is a bit vague about R1, its value being given as either 2,000
or 4,700 ohms. But, why would R1 get very hot? Its purpose is
to help provide bias voltages, however it will carry most of
the HT current so if the current is excessive the resistor's
rating might be exceeded.It looks like 2000 ohms so if 100mA
flows through it the thing will dissipate about 20 watts and
it looks like maybe a 2 watt resistor, hence the smoke. If it
was a 2 watt resistor then an HT current of 40-mA will result
in 3.2 watts. So, it seems something is drawing too much HT current.
First guess is perished rubber covering (of which there's a lot),
second is a badly leaking condenser and a prime candidate is
whatever has been used to drive the 6K6 grid (making it draw
loads of current), but most likely is all three. Simple enough
to check, just unplug the valves. I also need to check the modified
circuitry because the audio amplifier V8 is still present, so
that whoever did the mod must have retained this and connected
its anode to the grid of V9 via new condenser. |
 |
|
Above... Once I'd got the refurbishment
underway I decided to build a simple power supply using some
of the parts previously removed from the chassis. One reason
is the way the R1155 receives its HT supply. HT negative is connected
to a network of components rather than the chassis. This is done
to develop various negative bias voltages and makes it awkward
to use a lab bench supply. As you can see the new PSU is very
basic. Connections are made to the 8-way Jones plug via a choc
bloc and mains is fed in via an old computer power supply socket.
The front panel is part of an old computer case. I might add
a meter and a switch or two to the front panel later. |
|
|
Further rummaging around
the R1155 chassis revealed several poor condensers and swapping
these resulted in some reception. All the bands were alive but
very patchy indicating alignment and overall gain were way out.
I settled on Radio 4 on 198KHz. While I prodded with my multi-meter
the HT current suddenly shot up and more smoke appeared. It seemed
to be coming from a small red resistor at one end of the RF box,
at least this looked charred when I'd switched off power and
it measured open circuit. The 2.2Kohm resistor, R42 which is
the anode feed resistor for the RF amplifier section of the frequency
changer valve was open circuit. Checking its decoupling resistor
proved this was to blame, having no capacity and a DC resistance
of less than 1Kohm. I replaced the two parts and turned on the
set. This time it came on with lots more audio and much more
lively. |
|
Above, testing a typical
old condenser (this one is supposed to be 0.1uF and was made
in 1943)
|
|
|
|
Alignment was next, but
soon came to a dead stop because the slugs in the IF transformers
except the last one, L21, were damaged and immovable. I turned
to aligning the five wavebands. I'll set down the details here
because it may be useful for anyone thinking about this task.
First the wavebands (restricting myself to the R1155E) together
with their relevant coils and trimmer condensers. |
|
Tuning |
Tuning |
Aerial |
|
|
Aerial |
RF Amp |
|
RF Amp |
Osc |
|
Osc |
|
Waveband |
Low |
High |
Coil |
Value |
Code |
Trimmer |
Coil |
Value |
Trimmer |
Coil |
Value |
Trimmer |
|
1 |
7.5Mc/s |
18.5Mc/s |
L2 |
0.74uH |
|
C61 |
L7 |
0.77uH |
C66 |
L13 |
0.70uH |
C72 |
|
2 |
3.0Mc/s |
7.5Mc/s |
L3 |
5.64uH |
|
C60 |
L8 |
6.21uH |
C65 |
L14 |
5.83uH |
C71 |
|
3 |
600Kc/s |
1,500Kc/s |
L4 |
158uH |
10D/161 |
C59 |
L9 |
163uH |
C64 |
L15 |
70.6uH |
C70 |
|
4 |
200Kc/s |
500Kc/s |
L5 |
1462uH |
10D/162 |
C58 |
L10 |
1374uH |
C63 |
L16 |
248uH |
C69 |
|
5 |
75Kc/s |
200Kc/s |
L6 |
11.426mH |
10D/163 |
C57 |
L11 |
10.5mH |
C62 |
L17 |
699.2uH |
C68 |
|
|
Now, because the R1155
is not that different from other communications receivers, the
location of the coils and trimmers is not logical, being dependent
on mechanical layout, and therefore cannot be guessed. If you're
fortunate to have the RF compartment metal cover, the various
range markings will be present. In my example the cores of the
coils have been removed and cuts made in their ends so they can
be adjusted from the space between the front panel and the RF
compartment. The designers intended the coils be adjusted from
the rear of the receiver through a tangle of wires. |
 |
|
A view of the RF compartment
with its array of trimmers and coils (minus its metal cover)
The local oscillator (right),
RF amplifier 2 (centre) and RF amplifier 1 (left)
|
|
Below are the locations
of trimmers and coils corresponding to the 5 wavebands (ignore
coils in square brackets at this point). Access is tricky for
some of the coils. and if they've been messed about too much
some cores may need drilling out and replacing. |
|
|
|
Pictured below are two
further coils located at the rear of the RF compartment which
are not easily identified in the official drawings. I guess they're
for lower frequency ranges. Also, I've noted some coils above
which are not concerned in basic alignment. These are L1, which
is described as "Dummy Loop", L12 "IF Filter"
and L18 "Choke, Ranges 1 & 2". L12 is wired into
Range 3 and is used to provide a notch at the IF, 560Kc/s. I'm
still looking for L4, L5 and L6 which are the Aerial coils for
the three lower frequency bands and there's also a discontinuity
between the aerial connections on the Jones plug and the front
end RF amplifier, V3. L18 is wired across L13 and L14 (the two
short-wave local oscillator coils) and is tapped to provide coils
L18a & L18b which will provide rejection of any strong local
signals close to the IF. |
|
Centre of box...
what's this coil; is it part of the suppression of interfering
MF station circuitry?
|
|
|
End of box... what's
this coil, wired to Range 3 trimmer but with connection to a
coupling winding cut? What's that 2.2Kohm resistor?
|
|
|
|
Anyone contemplating working
on an R1155 will need to understand the wavechange switch. The
various published circuit diagrams show the wavechange switch
to be chiefly responsible for selecting one set of coils from
the five sets corresponding to Ranges 1 through 5. The switching
is not altogether straightforward however because of the two
major functions of the receiver viz. normal reception and direction
finding and the two different aerials found on aircraft (What
if one aerial is not available? Well, there's an aerial switch
in the installation which can provide suitable connections so
that if one aerial is not available, another can be used). You'll
see in the circuit diagrams that the wafers making up the switch
are identified by a series of letters. The four wafers are "z",
"y", "x" and "w", starting from
the geared end. Each wafer has a front and rear; "f"
is at the geared side and "r", logically, towards the
back of the switch. Each side of a wafer can carry two single
pole 5 way switches each comprising a "wiper" plus
5 "outlets" making 12 possible connections or a total
of 24 possible connections per wafer. These possible 24 are not
all fitted to all wafers. The circuit diagram shows 7 sets of
switches: FS.zf, FS.zr, FS.yf, FS.yr, FS.xf, FS.xr and FS.wf.
The set FS.wr is not included in the circuit diagrams. There
are two basic types of switch, a simple one of five "select"
plus a type used to short-circuit coils that are not part of
the set used for the selected waveband. The tags on each side
of a wafer are numbered 1 to 12, including two "select"
tags "6" and "12".
To aid legibility the circuit
diagrams duplicate wafers FS.zf and FS.yr . Also somewhat confusingly,
you may note that connections between front and rear wafer tags,
where applicable for the "shorting" wafer switches,
is not shown in the circuit diagram. This feature is supposed
to be indicated by a circle, instead of a letter, shown in the
centre of the wafer drawing as shown in FS.yf, however FS.zr
omits this information.
Ordinarily, anyone contemplating
restoration of an R1155 will not need to know any of this detailed
information unless there's a problem with the switch or, as in
my example, there have been extensive modifications that have
gone wrong. This might happen if the DF circuitry has been removed
without taking account of its effect. Below is a summary of the
wavechange switch functions. Published circuits show the wavechange
switch in the Range 1 (the highest frequency) position. Note:
I need to check the following table for errors. |
|
Wafer |
Function |
|
FS.zf |
V4 osc grid circuit C35 connects via
tag 12 to oscillator reaction winding tuned circuits L13 to L17
on tags 1-5 |
|
|
L13-L17 tuned circuits on tags 7-11
connect to tuning condenser on tag 6 (see also FS.zr) |
|
FS.zr |
V4 osc anode on tag 6 connects to
tuned circuit L13 to L17 taps on tags 1 to 5 |
|
|
Tags 7-11 are wired to FS.zf tags
7-11. Osc coils L13 to L17 common (R35) on tag 12 shorts unused
coil reaction windings on tags 1 to 5, plus R35 connects to unselected
higher frequency tuned circuit taps. |
|
FS.yf |
Tuned cct common on tag 6 shorts
out unselected coupling coils on L7 to L11 on tags 1 to 5 |
|
|
|
|
FS.yr |
V3 anode on tag 12 connects to RF
output coupling coils on tags 1 to 5, |
|
|
V4 grid circuit on tag 6 connects
to RF amp tuned circuits on tags 7 to 11 |
|
FS.xf |
Mode switch on tag 7 connects to
aerial coupling on HF coils L2 & L3 |
|
|
V3 AGC line on tag 1 shorts out unused
aerial coupling coil inputs |
|
FS.xr |
Top cap, grid of V3 at tag 1 connects
to tags 7-11 aerial coil tuned circuits L2 to L6 |
|
|
Mode switch on tag 7 connects to
aerial MF tuned cct taps |
|
FS.wf |
Mode switch on tag 6 connects to
MF aerial coupling A on L4 to L6 on tags 8 to 11; |
|
|
Mode switch on tag 12 connects to
MF aerial coupling B on L4 to L6 on tags 2 to 5 |
|
|
Looking for the nth time at AP2548A
to clear up loose ends I spotted L4, L5 and L6, which are hard
to miss being contained in large cans originally located in the
space occupied by the mains transformer. Are they important?
Well, they're used for the medium and long wavebands so in terms
of amateur radio operation are not vital. Studying the mystery
coil fitted on a bracket at the end of the RF box, it might be
a medium wave aerial coil added to replace L4? Adjusting trimmers
C59, C58 and C57 should reveal whether or not any coils are present.
I suppose I could always find suitable replacements and fit these
because I find the LF bands fairly interesting. If anyone has
an R1155 less the DF coils here's what you need (L5 and L6 below)... |
|
|
Aerial Coil |
L2 |
L3 |
L4 |
L5 |
L6 |
|
Inductance |
0.74uH |
5.64uH |
158uH |
1462uH |
11.426mH |
|
|
While I'm on the subject
of tuning I'll briefly set out the local oscillator settings.
If the receiver hasn't been messed with you won't have any trouble
but if a core has been twiddled or been replaced then you need
to be very careful. A table below shows the possible configurations
for the R1155. The lowest frequency band is interesting because
the local oscillator sweep is very small. Also, look at the negative
frequencies for the two lowest bands. Take Range 5, with the
tuning condenser set at near minimum capacity and the dial reading
200Kc/s. The IF of 560Kc/s can be arrived at by subtracting 360Kc/s
from 560Kc/s or by subtracting 560Kc/s from 760Kc/s. However,
as the tuning condenser is meshed the 360Kc/s local oscillator
frequency will drop to say 300Kc/s and the resulting response
would be 260Kc/s. In other words the dial is reading one way,
lower in frequency, and the receiver's response is reading higher.
This means that you've tuned the local oscillator to the wrong
setting. Quite easy to do if the receiver has been messed with
or a tuning slug has been replaced. One way to check is to tune
a second receiver (or a spectrum analyser) so it can hear the
R1155 local oscillator. If you don't do this check you may spend
hours of fruitless twiddling. This also applies to the higher
frequency bands if you try and align the receiver listening to
the image response by mistake. You can see the effect if you
input from a signal generator a strong signal, say 10mV at 10Mc/s
with the R1155 set to Range 1. Turning the tuning knob will result
in lots of responses but the two strongest will be 10Mc/s and
either 11.12Mc/s or 8.88Mc/s. These are twice the IF or 1.12Mc/s
apart from the dial reading. As the R1155 local oscillator should
always be tuning higher than the indicated dial frequency (for
all wavebands) the image should be 11.12Mc/s. If you hear strong
responses at 10Mc/s and 8.88Mc/s you've set the local oscillator
to 9.44Mc/s by mistake. This is quite easy to do and will result
in reduced overall sensitivity. |
|
It reminds me of
a time many years ago when I was calculating the local oscillator
and image responses in the design process for receivers in the
Skynet satellite anchor station at RAF Oakhanger. I wasn't responsible
for the calculations for the X-band receivers in the satellite
itself, but I recall there was a major problem once the thing
was in orbit and too late for twiddling. I mean, who has a plastic
screwdriver 22,200 miles long? |
|
Range |
Coverage |
Local Osc-LOW (WRONG) |
Local Osc-HIGH (CORRECT) |
Image Response |
|
1 |
7.5Mc/s to 18.5Mc/s |
6.94Mc/s to 17.94Mc/s |
8.06Mc/s to 19.06Mc/s |
8.62Mc/s to 19.62Mc/s |
|
2 |
3Mc/s to 7.5Mc/s |
2.44Mc/s to 6.94Mc/s |
3.56Mc/s to 8.06Mc/s |
4.12Mc/s to 8.62Mc/s |
|
3 |
600Kc/s to 1,500Kc/s |
40Kc/s to 940Kc/s |
1160Kc/s to 2060Kc/s |
1720Kc/s to 2620Kc/s |
|
4 |
200Kc/s to 500Kc/s |
-360Kc/s to -60Kc/s |
760Kc/s to 1060Kc/s |
1320Kc/s to 1620Kc/s |
|
5 |
75Kc/s to 200Kc/s |
-445Kc/s to -360Kc/s |
635Kc/s to 760Kc/s |
1195Kc/s to 1320Kc/s |
|
|
Before I proceeded to
deal with shortcomings in the front end, I'd noticed horrible
double-humping of the tuning of strong test signals so clearly
not all was well with the IF amplifier. There are three IF transformer
cans and each has two coils tuned to 560Kc/s. All are chewed
up, some more than others. The usual method of swapping a damaged
iron dust core is to break it up and remove the pieces then fit
a new one. Sometimes, if the thing can be removed without further
damage it can be turned round and the slot at the other end used
instead. In my set the cores were locked almost solid and I decided
to drill one of them out. I did this with the screening can in
place and I inadvertently damaged the winding so I had to remove
the screening can to replace it. As it was the can in the corner
of the chassis it came off easily after removing a pair of 4BA
nuts. I measured the undamaged coil and found it was 200uH without
its core. After winding a new coil on a suitable former I fitted
it and checked its inductance then removed turns until it was
something like 200uH. The R1155 AP gives the primary coil as
250.5uH and the secondary as 251.5uH, presumably with the cores
in place.
Fitting it was tricky as the
new coil former needed to be positioned so that its core could
be accessed through the hole in the screening can. After everything
had been reassembled I turned on the power supply and the signal
generator and commenced to align the IF amplifier. It started
well, but after a few minutes smoke billowed out from the vicinity
of the next screening can. To cut a long story short, I cut the
three live leads to the can containing condenser C29 which is
adjacent to the IF transformer and fitted new condensers C29,
C30 & C31. The hard part was fitting a new anode feed resistor
R30 (2.2Kohm) because this was fitted inside the IF can and detaching
this was not easy because one nut is hidden away. |
|
I removed the triple condenser
(3 x 0.1uF) which was bolted to the chassis and tested it. First
I measured its capacitance. The three sections weren't too bad
measuring 0.16uF, 0.16uF and 0.2uF, and their DC resistance seemed
fine with the multimeter registering over 30Mohms. However as
one of the condensers had seen off the anode feed resistor I
checked the three using an HT supply. Setting the voltage to
300V I fed this across each condenser in turn with a series resistor
of 1Kohm so I could test for leakage. The first condenser was
OK showing only about 2mV across the resistor, but the other
two leaked badly and the longer I left them the higher the leakage,
disconnecting them at 50V across the resistor. I then re-tested
the supposedly good condenser and this too, after heating up
from the others, was just as bad.
Having got the IF strip adjusted
to what I thought was 560Kc/s (previously there were at least
two clear tuning points ie. a double hump) I attempted to align
the front end. Here I had difficulty as can be seen when I measured
the local oscillator frequencies. To do this I used a small search
coil connected to my spectrum analyser and turned the tuning
knob on the R1155 to each end stop and waveband in turn. Although
every waveband had previously produced broadcast signals I'd
had trouble aligning most of them. |
|
Range |
Measured Kc/s
Osc Low & High |
Resulting Kc/s
Tuning |
Correct Kc/s
Dial Reading |
Correct Kc/s
Osc Low & High |
Inductor
& Trimmer |
Problem |
|
5 |
652 |
92 |
< 75 |
635 |
L17 |
Osc frequency too high |
|
792 |
232 |
>200 |
760 |
C68 |
Osc frequency too high |
|
4 |
784 |
224 |
<200 |
760 |
L16 |
Osc frequency too high |
|
1090 |
530 |
>500 |
1060 |
C69 |
Osc frequency too high |
|
3 |
1749 |
1189 |
<600 |
1160 |
L15 |
Osc frequency too high |
|
3600 |
3040 |
>1500 |
2060 |
C70 |
Osc frequency too high |
|
2 |
3370 |
2810 |
<3000 |
3560 |
L14 |
Osc frequency too low |
|
6840 |
6280 |
>7500 |
8060 |
C71 |
Osc frequency too low |
|
1 |
7750 |
7190 |
<7500 |
8060 |
L13 |
Osc frequency too low |
|
18170 |
17610 |
>18500 |
19060 |
C72 |
Osc frequency too low |
|
|
Looking at the above table you can
see that whilst ranges 1,2, 4 & 5 are alignable range 3 looks
badly wrong. Because the local oscillator was so far out I tried
unscrewing and screwing the core fully out and then fully in
to check its tuning range. In the table below you can see that
it's impossible to tune the medium waveband to line up with the
dial markings (as per the readings 1530 and 840 which represent
the minimum local oscillator frequencies achievable) . In fact
it almost looks as if someone in the dim distant past has modified
the set to tune top band instead of the medium waveband. In most
superhet receivers the designers use what are called "padder"
condensers to get the local oscillator and tuning condenser to
set the desired waveband edges. The padder condenser for range
3 may therefore be faulty. The R1155 local oscillator is different
to many receivers because it's the anode of the oscillator that
carries the tuning rather than the grid. Range 3 coil is L15
and the padder is C75 and this must be too low in value. If not,
as all the other wavebands are within spec, either coil L15 is
faulty or the dust core is not increasing the inductance of L15
sufficiently. C75 should be 537pF. Very odd, I disconnected one
end of C75 and it measured 1,002pF so I removed it and discovered
it had 1000pF 5% marked on it. |
|
Tuning Core |
Measured Kc/s
Osc Low & High |
Resulting Kc/s
Tuning |
Correct Kc/s
Dial Reading |
|
Fully out |
2090 |
1530 |
<600 |
|
Fully out |
4410 |
3850 |
>1500 |
|
Fully in |
1400 |
840 |
<600 |
|
Fully in |
2950 |
2390 |
>1500 |
|
|
Clearly the
inductance isn't high enough to allow alignment of range 3, but
as the padding condenser will effectively negate the effect of
inductance, I decided to shunt it. Adding more and more capacity
almost achieved the minimum frequency but shorting out the padder
did the trick. I was able to align range 3. As I tuned the various
stations on the medium waveband however it was plain that they
were not on their normal dial settings. Classic Gold in Bournemouth
on 828Kc/s was nearer 780Kc/s so I investigated the IF strip.
Sure enough it was peaked on 589Kc/s rather than 560Kc/s. I tracked
down the reason to a coil sitting away from the main set of coils.
This is a notch filter which is supposed to stop strong signals
near to 560Kc/s from getting into the receiver. Previous attempts
at IF alignment had been skewed because of the notch filter.
Once I'd temporarily added an extra iron dust core to the coil
the receiver responded to 560Kc/s at the aerial and IF alignment
was possible. I used a spectrum analyser with a tracking generator
and set up the IF response. When I checked Classic Gold again
it was sitting at the right place on the dial.
Below are the inductance values
for the IF coils and the BFO coil just in case you need to replace
one as I did. |
|
COIL |
PRIMARY |
SECONDARY |
|
L19-IF1 |
244.5uH |
251uH |
|
L20-IF2 |
250.5uH |
251.5uH |
|
L21-IF3 |
132uH |
254.5uH |
|
L22-BFO |
272uH |
|
|
|
Below is the response curve
for the R1155 IF with maybe a little distortion to the shape
due to a little loading by the probe at the rectifier diode?
The vertical divisions are 20Kc/s apart and the centre line is
560Kc/s. By my reckoning the skirts are about 40Kc/s wide at
-50dB or 20Kc/s at -20dB. My aim at this time was to correct
the centre frequency rather than to get the narrowest curve. |
 |
|
After finally managing
to get Range 3 roughly aligned I decided to find out why the
padder was 1000pF. Checking the circuit, I found that by disconnecting
the padder and the trimmer I could measure the capacitance of
the tuning condenser. It turned out to be 85-560pF including
the trimmer and strays and 55-530pF including strays rather than
the 660pF (plus strays) version used in later models. This explains
the larger Range 3 padder as the smaller one wouldn't give a
wide enough tuning range. I also investigated the 0.1uF condenser
which decouples the oscillator anode feed of 22Kohm. As this
was almost completely inaccessible I broke off the wire leading
to it and connected this to a new 0.22uF. As far as I could tell
the old condenser was serviceable. I still have to strap out
the 1000pF padder to get correct alignment so I'm left with the
probability that the Range 3 coil hasn't enough inductance. I
wonder if the tuning core is the problem because the coil looks
original? Another slim possibility is the coil primary and secondary
windings are connected back to front ie. tuning the lower inductance
primary instead of the larger secondary.
As the oscillator works normally
and I don't have any iron cores that increase the inductance
of L15 more than the one fitted it will be relatively straightforward
to just add a second coil in series with L15. To work out the
details:
Target frequency for tuning
600Kc/s = 600+560=1160Kc/s; Capacitance 550pF in series with
the padder of 1000pF= 354pF.
Correct inductance of L15 is
70.6uH but calculated inductance is 53uH given 1160Kc/s and 354pF.
Shortfall 70.6-53= 17.6uH. This
is achievable with 60 turns on 0.5" former wound over 1"
using 27 or 28SWG enamelled wire, however when I tried this the
receiver became unstable so I removed the trimmer panel, then
removed the coil, L15. This is done by removing the two 6BA nuts,
easy because the former has 6BA screws held in place with locknuts.
I measured the coil and found it was 53uH as calculated. Rather
than modify the old coil I found a suitable former in my junk
box and wound it with a 70uH coil. To do this I used 37SWG wire
and prepared the former by wrapping it with double-sided tape.
This prevented the wire unravelling. After around 70 turns of
the 115 required I made a loop in the wire for the coil tapping
point then finished winding the turns. It measured 70uH as predicted.
I then taped over the coil with more double-sided tape and wound
about 20 turns for the feedback coil using 32SWG wire. This measured
about 7uH and much to my surprise the new coil worked perfectly.
Pictures below... |
The old Range 3 coil which looks
original and whose main winding has an inductance of 53uH and
a reaction winding of about 3uH.
On the right is the new coil
with a main winding of 70uH and a reaction winding of about 7uH.
The new coil has a tapped main winding and I made this about
the same percentage as the old coil. The left pins are connected
together and are wired to the main winding and the reaction winding
like the original.
I first wrapped double-sided
sticky tape along the former then wound 37SWG wire to achieve
a single layer. The reaction winding was made with 32SWG wire
in the same manner. |
 |
Below, a view of the oscillator
coils after removing the trimmer panel, and right after removing
the Range 3 coil. |
 |
|
|
The new coil, not quite
finished being tested in-situ to check its tuning. |
 |
|
View after replacing the
trimmer panel. |
 |
|
I was listening to our
local radio station on 1359Kc/s when the sound went muffled and
hissy, a puff of smoke arose from the front of the chassis and
a resistor burnt out... yet another condenser has bitten the
dust!
Further testing indicated that
the three higher bands all suffered from a sharp dip in oscillator
output at about two thirds mesh of the tuning condenser. This
is a bit puzzling. Maybe one of the rotor plates has a blob of
solder stuck to it? Further investigation required. I think it
was due to too large a negative bias voltage or too low an HT
voltage due to leakage in the HT line to ground. |
|
As there's no connection
between the Jones plug aerial pins and the RF compartment, I'll
need to sort this out. Originally the aerial (there were two..
the HF aerial and the MF aerial) passed via the mode switch and
the DF circuitry so someone might have dispensed with this and
just connected an aerial to the common circuitry associated with
the grid of V3? As I plan to use the R1155 with my T1154 I'd
like the aerial connections to be relatively original. I'd noticed
the end wafer of the wavechange switch has had most if not all
its wiring disconnected.
I decided to disconnect the
receiver from the power supply and take it where there was plenty
of light so I could trace the aerial circuitry. It seemed to
me that there were two different points for different wavebands
where an aerial brought in stations. At the left of the dial
is some white lettering which I've deciphered as "RANGE
3 SCALE X2". I looked at the wavechange switch and noticed
the lettering for Range 3 was much darker than the other lettering.
Cleaning it revealed "3 - 1.2". I looked at the main
label. It tells me its an R1155E but the Air Ministry code is
10D.NIV.400 (see below). So that explains the strange coverage
for Range 3. This particular set has been modified to cover the
Trawler Band and the coilset for the R1155L trawler band has
presumably been fitted in place of that for the medium waveband
of the R1155E.
My T1154 declares itself to
be a Model "N" but it's not. It has four wavebands
and just about covers Top Band, so I think it would be a good
idea to swap back the receiver oscillator coil, then I can use
the set-up on Top Band. That's if I can find the old coil.... |
 |
 |
|
Above, the unusual label.
The US government issued a publication of almost 200 pages which
covers probably all the UK abbreviations and includes the code
"NIV" on page 161. Above, you'll notice that the label
engraver baulked at drawing the crown between the letters "A
M". Somewhere, in a locked filing cabinet in a dusty basement,
there's a register which includes the definition of "400".
Marked on the rear of the chassis is that same serial number,
73736, so at least that's correct. |
|
|
|
I suppose, to preserve the integrity
of the old receiver, which may now turn out to be unique, I'll
have to refit the old oscillator coil and, while I'm at it, swap
those rusty 6BA screws!
I found the original coil after twenty
minutes. Gravity had acted and it had fallen to the workshop
floor where stuff is arranged in an archaeological fashion in
strata. I spent ages working out the connections but finally
turned on the power to find all was well. Tuning the dust core
brought the band edge in line at 1.2Mc/s with the dial showing
600Kc/s and, after adjusting the trimmer brought 3Mc/s onto the
1.5Mc/s marking. The RF amplifier coil responded and aligned
correctly but the aerial circuits have a problem which I discovered
previously so the next step is to sort this out and check that
the aerial feed is taken via the Jones plug to the T1154 output
connector. A quick test using a long wire brought in several
medium wave stations at the LF end of the band plus lots of signals
that sounded like Russian jammers between 2 and 3Mc/s.
The next step was to replace
lots of rusty screws with new brass 6BA screws. When I removed
the screws securing the label shown above I found it was a label
for something completely different that had been flipped over
and re-used.
 |
This is the back of the R1155 label.
This may be the only such label in existence as I've never heard
of such an equipment. Pyrotenax was a copper clad cable. |
Next, I was determined to trace
receiver deafness. I found the coil I'd previously found mounted
on a brass plate was a new Range 3 aerial coil and its return
connection from its tuned winding was disconnected. This explains
the deafness on Range 3. But if that coil is for Range 3 what
are the three coils mounted in the coilpack? Certainly two are
aerial coils for the top two HF ranges but what is the third?
Perhaps the old range 3 coil was too tricky to remove so was
just disconnected and left in place?
At least I've fitted some new
screws to replace rusty ones. I suppose I should blacken the
heads sometime...
|
|
|
I'd guess that lots of R1155 receivers
are missing their coil-pack covers so I've added these pictures
of my home-made aluminium cover. It's held in place by seven
6BA screws rather than eleven in an original because I simplified
the bending. I didn't drill holes for the coil adjustment at
the rear because these coils are mainly inaccessible on my set
and cores are all adjustable from the front. A cover is not essential
but tuning and tracking will be optimum if it's fitted and final
tuning made with the cover in place. I use a 4BA nut spinner
for setting the trimmers so the holes are drilled to suit its
diameter. Apart from the trimmer holes, I drilled one for adjustment
of the IF rejector and one for the Range 3 aerial coil which,
on my set, is fitted on a bracket at the rear. |
 |
 |
|
|
Having got the original Range
3 coil refitted and working, I wonder if the original receiver
conversion (probably back in the 1940s) was comprehensive? Was
the complete set of Range 3 coils fitted at the time? It may
depend on the reason for the modification. Did the set need to
be really sensitive? If the answer is in the affirmative, then
coils for the aerial and the RF amplifier would have been fitted
so that perfect tracking could be achieved. A clue might be the
additional coil mounted on a small brass plate near the aerial
switching wafer of the wavechange switch. This coil looks very
similar to the oscillator coil and is indeed connected to the
correct tag on FS.xr. Oddly there is a small 2.2Kohm resistor
also fitted close to the coil and that looks to be fitted amateurishly.
Possibly the chap that removed the DF parts and fitted the internal
power supply was confused? The coil connections are cut and roughly
resoldered which might support this idea.
I already mentioned that I need to construct
a power supply for the receiver as I want to free up my workshop
power supply and also because of the bias wiring which places
the chassis at an odd voltage in respect of HT negative. I made
a rough and ready chassis from a piece of chipboard and piece
of computer case and wired up a simple power supply (with 6.3V
AC for the valve heaters) using the mains transformer removed
from the modified R1155. I connected this up and the receiver
worked OK but, after about 30 minutes, the set had warmed up
and I noticed the chassis voltage was slowly drifting upwards.
Clearly there's HT leakage probably due to one or more of the
remaining old condensers. I imagine the only practical solution
is to remove all the old tubular condensers and fit modern parts,
but read on.... For interest I checked the HT current. With the
set just warmed up the HT current measured 68.6mA and with the
valve heater feed disconnected this dropped to 30.3mA. The HT
voltage measured 251.9V and the chassis measured 36.8V to HT
negative.
Note that these voltages and
currents result from my home brew power supply so will be representative
only. Not only that, but later you'll read that I found a leaky
condenser responsible for some of the HT current.
|
I
had my breakfast and then returned to the set. There was a rather
nice warmish smell, like the inside of a 1950s government surplus
shop so I gingerly put my finger on the side of each of those
metal tubular condensers screwed to the chassis. All were cold
so I checked the HT current and it had risen to 114.3mA from
68.6mA. I checked the HT voltage and found it measured 172.5V.
The chassis was resting upside down and I turned it on its end
with the Jones plug uppermost so I could check another one of
those tubular condensers for warmth. I didn't have to though
because I could feel a warm glow coming from the grey metal clad
paper condenser screwed to the end of the chassis. The sides
were bulging slightly and it was clearly on its last legs. I
clipped the wire feeding its HT terminal and the HT current dropped
back to 62mA, the HT voltage jumped to 252V and the chassis measured
only 20.8V. Tuning to Radio 4 showed the set was a lot more sensitive.
The condenser is C93 (10C/979) and is marked 4uF. It had accounted
for an HT drop of around 20 volts when cold and a whopping 80
odd volts and 50mA when hot. Oddly the working voltage isn't
marked, neither is the manufacturing date. Its resistance measured
with an ohmeter is around 8Mohms (which I guess is a bit too
low and was rising as the condenser cooled which is ominous)
but strangely the capacity and ESR are normal (picture below). |
 |
|
|
As the condenser is a
substantial part of the R1155 I decided to keep it. Many metal
condensers like this one are oil filled and anyway, whatever
is their construction they seem to last for ages but maybe 70
years is condenser "end of life" ? Cutting it open
was tricky but I have a large old Solon soldering iron and I
found by cutting the ends of the base where damage wouldn't be
visible with a hacksaw enabled me to unsolder the terminal panel.
Inside I discovered a lump of aluminium and paper exactly the
same as that in the failed metal cased condenser in the T1154
(see link below). It's also very similar to the construction
of the standard tubular condensers used in this and other receivers
from the 1940s. Clearly, something in their construction is degrading
after 70 years and resulting in rising temperature and a sort
of avalanche type failure.
Below, the same 4uF condenser,
but re-stuffed with two 2uF and a 1uF rated at 275VAC which should
be OK up to about 380VDC. I drilled out about half the contents
which was sufficient to accommodate the new components. See
other bad condensers (these in an R206 Receiver) and
another in a T1154. |
 |
|
|
The R1155 has several triple 0.1uF
350 vw condensers (like the one below) which are in tubular aluminium
cans bolted to the chassis. These will all warrant investigation
as they eventually develop an electrical leaks which reduce whatever
voltages they're decoupling as well as not properly performing
their decoupling function. I've described
a method for restuffing these cans here. The cans are not
always easy to remove from the chassis because of accessibilty,
but it's a worthwhile project to replace them with new parts.
The method I've described has a bonus. To refit a can you can
pull out the new capacitor assembly which is held inside a section
of drinking straw, screw the can back in place, then push the
new capacitor assembly into the can. This considerably simplifies
can replacement. The leads can then be trimmed to length and
soldered in place. The capacitors are 0.1uF, size 1210 (3.2mm
x 2.5mm) and are type X7R rated at 500 volts. The wires are high
voltage/high temperature. Pricewise, these or similar ones vary
considerably from 3p to 10p if you shop around.. but most suppliers
offer lowest prices in quantities of a couple of thousand. I
was lucky and bought a couple of hundred for 3p each. |
 |
 |
|
Before I'd decided to
tackle these triple condensers the refurbishment had been going
well, but suddenly there had been a sizzling noise and smoke
when one section of one of the cans had failed. It seemed sensible
to replace all six of these cans before proceeding. This had
the advantage of being able to remove a large number of new,
bright orange, decoupling capacitors I'd been fitting as work
proceeded. In addition to the tubular cans there's a triple condenser
in a cube-shaped grey block adjacent to the tuning condenser
and a couple of metal-cased condensers screwed to the chassis
that I should replace. |
 |
These old condensers are
sometimes still in perfect condition but as the larger version
in the R1155 had become red hot I thought this example might
be on the way out.
I identified the connections
which are labelled "C" for common, two marked "2.5"
and one marked "1" and checked them with my ESR60.
Readings were:-
2.62uF @ 0.10 ohms
2.86uF @ 0.4 ohms
1.23uF @ 0.16 ohms
Next, I'll check them by applying
HT.
Across a series resistor of
100kohm connected to 300 volts DC I measured 29 volts for the
1uF, and 17.5 volts and 4 volts for the 2.5uF. These voltages
represent less than a mA of leakage, but are symptomatic of a
problem that can only get worse. |
 |
The leakages were not
too bad but in the interests of longevity I decided to fit new
innards in place of the waxed paper/aluminium. I guess the condenser
is rated at no more than 350 volts so found some plastic capacitors
rated at 275 volts AC (about 389 volts DC). The old condenser
has a pretty small volume for a 1940s component and fitting the
new parts wasn't easy.
The top of the condenser is
fitted with a soldered lid, with a lip of a couple of mm, carrying
four terminals. Each terminal comprises a threaded brass rod
with a hole through the centre, a solder tag and a special dished
4 BA nut. Condenser wires pass through the holes and are soldered
into the nuts. I had to remove as much solder as possible then
drill out the holes so wires would pass through then soldered
to the nut.
I found in my junkbox a number
of 1uF 275VAC capacitors and managed to squeeze in eight of them.
I needed eight because all were lower than their marked values
(a cost saving ploy by the manufacturers??). I fitted two groups
of three and a pair, giving me two 2.7uF and one 1.5uF.
Bearing in mind the larger 4uF
block condenser had failed suddenly this one, unmodified might
have done so in the future.
Fitting is easy as it's attached
to the chassis with threaded holes by 4BA screws. |
|
|
After refitting the new
old condenser the R1155 worked perfectly, but I'd forgotten to
add R1, the 2.2Kohm resistor. It didn't seem to bother it. I
measured 65 volts between the chassis and C1 and 45 volts after
I'd fitted R1. Presumably there's enough resistance in the bias
chain with AVC on or off to let the set work OK and it's only
essential in other modes? I suppose I should make resistance
checks to confirm this. I've now finished stuffing two more triple
condensers. It took only minutes. I used a piece of copper wire
an inch long, soldered three chip capacitors to it then added
three red wires to their ends and a black wire to the coper wire.
Squashed 3 inches of plastic drinking straw and pushed the assembly
into it and capped it with the end of a cotton bud. With the
can drilled out using a drill fitting the hole I can push the
capacitor assembly into place once the can is refitted to the
R1155 chassis. |
|
|
Up to this point in proceedings
I've been connecting my aerial to a capacitor sticking out of
the coil pack. Whoever modified the receiver back in the 1950s
probably added this to bypass the switching arrangements. If
I intend to use the R1155 in conjunction with a T1154 this modification
is not ideal so I need to plan exactly what I'd like to do in
respect of connecting an aerial. The first thing to consider
is the switching arrangements in the T1154 as that is where the
send/receive switch is located and this switch handles antenna
switching. Operating on shortwave bands is relatively straightforward
but an option is open to use new amateur bands on 136KHz and
472-479KHz (600m) so this might be taken into account in any
tidying up of the switching within the receiver.
After studying the circuit diagram I
found that the main aerial inputs to the receiver, ignoring all
the DF arrangements (all that circuitry has been removed anyway),
are tappings on the RF amplifier coils for the two long wave
bands and the medium wave band and a coupling coil for each of
the two short wave bands. These five inputs need to connect to
Pins 1 and 2 of the 8-way Jones plug. Pin 1 is designated "Fixed
Aerial" and Pin 2 "Trailing Aerial" and it might
be useful to include an extra switch on the front of the receiver
to combine these so a long wire can be used for general reception.
It's likely that the capacitor connection I'm currently using
merely connects to the grid of the RF amplifier as it appears
to work on all five wavebands. In the original circuit two condensers
C100 and C102 are used to isolate any DC present at the R1155
longwave and short/mediumwave aerial circuits (respectively)
from the T1154.
Missing from my receiver are coils L5
and L6 so I'll need to find some and fit these for optimum results.
From the table (above) these were 1462uH and 11.426mH respectively.
To make things simple I may use a coupling coil for the aerial
rather than tap these.
Summarising:-
Jones Plug Pin 1 is wired to C102 (0.001uF)
which goes to FSxr via the Mode Switch MSBf and MSdf with R63
(2.2Mohm to ground).
Jones Plug Pin 2 is wired to C100 (200pF)
which goes to FSxf via the Mode Switch MSBf and MSdf.with R62
(2.2Mohm to ground). |
|
I decided to make a compromise
by removing the RF wiring from the mode switch and instead have
five aerial tuning coils selected by the wavechange switch with
each one having a small winding for aerial feed then connect
two feeds from these through to the Jones plug via C100 and C102
as before. This meant that I had to wind a pair of coils.. one
each for Ranges 4 and 5. These two will connect to the trailing
aerial connection with the other three to the fixed aerial. I'll
also arrange for a spare panel switch to connect the Range 4
and 5 coil aerial windings to he fixed aerial connection. This
will let me tune the reiver to any range using a single long
wire aerial. It will also let me use the R1155 with the T1154
using two separate aerials.
I couldn't find a suitable coil for
Range 5 in my junk box. The handbook tells me this should measure
11.4mH. So that it can track between 200 and 75KHz I had to use
a former suitable for a core so that the inductance can be set
correctly for 75KHz. I found a suitable former and wound 1100
turns of fine wire on it. This measured 9.2mH and I reckoned
a dust core would easily increase this to the desired 11.4mH.
I tried this out on the bench but couldn't get it to tune across
a broad enough range, although by reducing couplings to the test
equipment it improved somewhat. I reckoned that in-situ in the
reiver it would be OK, but after adding it I discovered it wouldn't
track. It would almost tune at the high end but there seemed
to be excess inductance for 75KHz. Wondering if the coil had
too much self capacity I did some tests and found this was about
47pF. I imagine this is higher than the original coils because
mine uses a "scramble winding" of single thin copper
wire. Traditionally low frequency RF coils use litz wire in a
low self-capacity winding. I don't think the error in tracking
is important but I'd like to investigate further.
Interesting is L4. This is shown mounted
on the chassis adjacent to L5 and L6, but has been replaced by
a small aerial coil mounted on a brass plate at the rear of the
aerial coil compartment. Bearing in mind that this receiver carries
a modified Range 3 it's possible this coil was added as part
of the modifications to cover the new range 1.2-3MHz. |
 |
 |
|
Above are the two new
aerial coils, L5 (left) and L6 (right). Each has a coupling winding
which is connected to the wavechange switch at wafer FS.wf together
with that for the Range 3 aerial coupling coil. The return path
for these coupling coils is the receiver chassis, but the return
for the two tuning coils is connected to the bias voltage carried
on an 0.1uF condenser mounted underneath L1-L3. Connecting the
aerial coupling windings for two new coils to the wavechange
switch was tricky because the wafer contacts were buried in wiring
and not readily identifiable because original wiring had been
removed by the last owner and he'd never completed the modifications.
L4-L6 had all been removed and discarded so that the added power
supply parts including the mains transformer would fit on the
chassis.
L5 has an inductance of 1.4mH
and L6 9.2mH and because both home-made coils carry a fairly
high parasitic capacitance their tracking is not perfect, but
satisfactory. I looked around the Internet and found a really
interesting and easy way to discover coil information. Basically
it uses the fact that a square wave is rich in sine waves and
one of these will be close to the resonant frequency of any coil
under test. You need a square wave having an amplitude of something
over say 5 volts, a 10Mohm resistor connecting the square wave
output to the coil and an oscilloscope with a 10:1 probe also
connected to the coil. My scope has a square wave test output
and I guess that might do in place of the separate square wave
generator? When the coil is connected and the oscilloscope adjusted
you'll see a set of sine waves decreasing in amplitude from the
square wave edge. The frequency of these waves represents the
coil resonating with its self capacity so if you know the coil
inductance (measure this on an inductance meter) you can calculate
the value of self capacity. Also, the Q of the coil can be determined
by counting the number of cycles to the point where their amplitude
is exactly half the maximum value. I checked my new L6 and found
it was 9.2mH + 47pF and my new L5 measured 1400uH + 27pF. L6
had 1100 scramble wound turns.
In the above picture (the new
L5-left and L6-right) you'll see two 2Mohm resistors. These are
R63 and R62 connecting to Jones plug pins 1 and 2, fixed and
trailing aerial inputs routed from the T1154 transmitter where
they connect to the transmit/receive relay. To make the R1155
independent of the transmitter I intend to add a coax connector
and a switch for combining these for long wire use in general
reception. I fitted the coax socket and a headphone jack socket
in the space adjacent to the two Jones plugs (still remaining
in my example).
I also took a look at the slow
motion drive (the early type 13 variant) because it had some
backlash. I noticed the cork washer was loose so this needs to
be re-glued to the underside of the part still on the spindle,
then grease applied to the various parts after cleaning. |
 |
|
The receiver was working perfectly
before I added a small plate carrying an aerial socket and a
headphone jack socket, removed the original toggle switch above
the mode switch and fitted a double pole double throw type. I
connected the aerial input connections at the Jones plug to the
new switch and added a short between the two on tags and a wire
to the new aerial socket. This lets me use the R1155 independently
from the T1154 with a long wire covering all five wavebands.
The headphone jack socket was then connected to the secondary
of the audio output transformer and the new panel fitted over
the unused Jones plug and the space to its left where a small
Jones plug should have been fitted.
I turned on the set but could only hear
a faint tinny sound from the loudspeaker. Plugging in headphones
initially let me hear Radio 4 on 198KHz but this gradually faded
out to just a faint hiss. My first guess is that one of the few
remaing old condensers had failed. There can't be many left now
but I suspect a screen or anode voltage is now very low... maybe
around the 6K8 mixer/oscillator? The voltage across the bias
resistor R1 measures 47 volts and HT to chassis 170 volts, but
it turned out that all the components were fine and the 6K8 had
stopped working. I plugged in a 6A8 that I had on the bench from
another job and it worked OK except it didn't oscillate at all
on the 7.5-18MHz band so I'll have to dig around for a 6K8. |
|
new plate |
|
I tested the performance of the
front-end (with the 6A8 plugged in). To do this I connected my
spectrum analyser tracking generator to the aerial and my high
voltage probe to the anode of the mixer. When you do this with
a tunable receiver you'll see two or three well defined signals.
See a scan of the RF25 unit. The local
oscillator and either or both main mixer products ie. Fosc +/-
Fif. If the scan is broad enough to include the IF of the receiver,
this also will be visible. I prefer to scan just the band in
question as that is what you need to see to the check ganging
of circuits. What you see is the performance of the receiver
at whatever frequency the dial is set. With a receiver having
two tuned RF circuits you may see either a nicely defined peak
to the left or right of the local oscillator (depending on whether
the design has the oscillator higher or lower than incoming RF),
or a fat signal or even a double hump indicating the two RF circuits
are not ganged properly. As the higher short waves are tuned
you'll also see the image at the opposite side of the local oscillator.
Below, in Range 3 scans there were two humps 1.01MHz and 1.07MHz.
With a single conversion receiver (like the R1155 which uses
an IF of 560KHz) reception in the highest shortwave band will
be pretty poor in terms of image response.
Because of the higher proportion of
capacitance to inductance at the low end of each band the width
of the response ie. the selectivity will be better there and
as the LC ratio increases so does the selectivity worsen. This
is unfortunately a physical fact: to achieve a wide tuning range
the band will have poorer selectivity at the high end and of
course the image response will suffer.
With the scan extending to more than
double or triple the local oscillator frequency you'll also see
its second and third harmonic. Because of these extra (false)
signals and the use of a long wire, the highest shortwave band
will be full of signals, many of which are transmitting on frequencies
miles different to the dial setting. In practice, unless one
of these falls inside an amateur band these false signals are
unimportant. Alas though, interference from low energy lamps,
network cameras, computers and a whole host of unregulated electronic
apparatus results in a high background noise level not helped
by receiver-generated extraneous mixer products shifting extra
noisy parts of the spectrum to the IF amplifier.
Below... scan results of the five ranges
where the measured amplitudes reflect the tracking generator
amplitude (constant over all tests) |
|
Frequency band |
Range 1: 75KHz to 200KHz |
Range 2: 200KHz-600KHz |
Range 3: 600KHz-1.2MHz |
Range 4: 3-7.5MHz |
Range 5: 7.5-18MHz |
|
Low end |
-28dBm |
-39dBm |
-35dBm |
-29dBm |
No osc |
|
High end |
-22dBm |
-4.7dBm |
-23dBm |
-31dBm |
No osc |
|
Spot frequency |
150KHz= -25dBm |
300Khz= -24dBm |
|
|
|
|
|
I'll repeat these measurements
once I've fitted a new 6K8.
I fitted an equivalent X65 and
made various measurements on Range 3 which is unusual in this
particular version covering 1.2 to 3MHz (although the dial still
retains the original marking of 600KHz to 1.5MHz). I moved the
dial from the low end then to the high end as you can see below. |
 |
Dial set to 600KHz (x 2) showing the RF from the tracking
generator peaking at about 1.2MHz but with double humping indicating
poor tracking between the aerial and mixer circuits.
|
 |
Dial set to 600KHz (x 2) showing the local oscillator
at 1.757MHz indicating the correct IF of 560KHz.
You'll note the spike on the right which is the secpnd harmonic
of the local oscillator, 15dB down on the fundamental. Although
the two peaks look to be in the same ballpark the scale is logarithmic
making the second harmonic about 33 times weaker than the fundamental.
|
 |
Dial set to 1.5MHz (x 2) showing the RF from the tracking
generator peaking at about 3.02MHz.
You'll note that the RF response at the high end of the band
is broader (x2) than that at the low end because the capacitance-inductance
ratio is smaller making the receiver less selective.
|
 |
Dial set to 1.5MHz (x 2) showing the local oscillator
at 3.455MHz.
The RF should be tuning at 3MHz but the oscillator isn't
tracking correctly.
To be correct the oscillator should be 3.560MHz. This can
be corrected by unscrewing the local oscillator trimmer condenser
so increasing the LO frequency.
Once this has been done a second check needs to confirm the
low end settings are still correct. If not the Range 3 oscillator
coil needs to be adjusted followed by a further check at the
high end.
|
|
|
During testing I noticed something
odd and reminiscent of my R208 investigations. As I tuned across
the highest frequency band I noticed the local oscillator amplitude
dramatically dropped at something like a third of the way from
minimum frequency and the exact frequency of the reduction was
affected by the tuning of the mixer coil. The R208 local oscillator
was killed off on its highest band but the R1155 local oscillator
doesn't actually fail, it just drops to a low value before increasing
again to its original level.
Checking further I found that when ther
mixer coil is tuned near to the same frequency as the local oscillator
the oscillator amplitude falls and then stops oscillating when
the mixer is tuned to the same frequency.
All is now resolved. The issue was this...
the mixer coil for Range 1 (7.5 to 18MHz) was jammed in its coil
in a retracted position with the core end flush with the top
of the former. This meant the tracking was severely affected.
Although the trimmers could peak the coils at both ends of the
band the mixer coil was tuned alternately to the main signal
or the image and crossing over at some point in the band, typically
9MHz. At that point the mixer and local oscillator coils were
tuned to precisely the same frequency causing the 6K8 to block
oscillation.
I couldn't retract the core but by gently
easing it back into the former it slackened enough to get better
tracking. By trial and error I worked out that the local oscillator
should be on the high side of the incoming signal. Once the mixer
was tracked to the oscillator I was able to track the aerial
coil which was also miles out of adjustment. Checking at the
top end of the range, an 18MHz modulated signal at 40uV had the
same audio output as the image signal at 19.12MHz at a level
of 1mV which represents an image suppression of 28dB. Further
tweaking should improve this. The local oscillator amplitude
is now pretty well the same amplitude across Range 1. |
 |
|
Above, the receiver just about completed
with its power supply. The left meter reads 62mA and the right
about 230 volts between HT+ and HT- rather than the chassis which
sits at about 50 volts above HT-. All the wavebands are now working
pretty well using a long wire aerial and a final alignment check
plus repair of the slow motion drive are all that's required
before the R1155 can be tried out with the T1154. The power supply
above is really for using the receiver in stand alone mode or
as a general purpose HT/LT PSU for other valved sets. It's now
relocated above the T1154. |
 |
|
When I tested the receiver on a long
wire I found the sensitivity on the two long wavebands was much
less than another example that I have. I put this down to the
home-made coils L5 and L6. These have excessive parasitic capacity
and are wound from single enamelled copper wire whereas the correct
coils have litz wire and are wound to keep parasitic capacity
low. I've now obtained a set of three original style coils L4/L5/L6,
two of which (L5 & L6) will replace the
scramble-wound versions I fitted. It remains to be seen if
sensitivity is improved once these have been fitted.... |
|
|
