The keying circuitry
The method of switching the
T1154 to transmit can be puzzling as it's not carried out with
a single connection. The wires involved are carried on the 4-way
Plug B which is labelled on the schematic as follows:-
13 Key; 14 Keying relay contact;
15 Earth; 16 Phones. The polarising pin is between Pins 13 &
14.
Jones Plugs numbering is very
odd. The 8-way is straightforward and logical with pins marked
1 to 8....... but why is the 4-way numbered 13 to 16?
The Transmit/Receive relay is
driven to Receive from an indeterminate position when LT is applied
and the mode switch is in any position except "OFF".
Then, if the Keying relay contact Pin 14 is grounded to Pin 15
you'll hear the Transmit/Receive relay clonk over to the Transmit
position. In this situation, with only Pin 14 grounded, the HT
current meter will read zero milliamps because the PT15s are
heavily negatively biased. Then, if Pin 13 (Key) is grounded
to Pin 15, the cut-off bias to the PT15s is removed and you'll
register current on the meter. To transmit therefore, you need
to switch both Pin 13 and Pin 14 to ground. I propose to do this
by connecting these pins to a double pole centre-off changeover
toggle switch. The switch will give me Transmit in one setting,
Receive in the centre and Transmit with a morse key in the other
position. In this setting the morse key will be in the Pin 13
wiring leg.
For a full setup, including
the R1155, Pin 16 carries audio from the microphone amplifier
(V4). The microphone pressel switch merely turns on the microphone
rather than switching the T1154 to transmit. |
|
|
Odd Connectors
|
|
In a little box that came
with the Gumtree T1154 I found some strange looking connectors
which turned out to be for the high voltage lead and the aerial.
Pictures below... |
 |
The high voltage female
connector that fits the T1154, on the left is coded 10H/425 and
the spare male connector is coded 10H/430 as you can see. |
|
 |
The right hand connector
is fitted to the T1154 chassis and carries a smaller pin which
is not used to carry current but is merely a locating pin for
the detachable connector.
I suppose the high voltage connector
metal shell (on the left) should ideally be connected via a grounding
lead for reasons of safety in the event of a short between the
shell and the centre socket. |
|
 |
This is the aerial connector
coded 10H/319 which plugs into an aerial socket on the side of
the T1154 as shown below.
The metal centre socket just
floats in the bakelite case and clearly there's no provision
for more than a single aerial wire which enters via a small slot
(just visible on the left) in the end of the case. |
|
|
|
|
The next step having tested the T1154 is to build
a dedicated power supply for it which embodies features for operating
it as closely as possible to WW2 operation. The usual installations
of the transmitter included a pair of power units driven from
either 12 or 24volts but in my case I'll be running it from a
single power supply driven from 240V mains. Because of the fact
that the relay needs 6VDC and the PA valves are designed to run
from 6VDC I'm using a regulated output voltage which can be set
to supply as closely as possible the correct voltage within the
transmitter. Due to the availability of a suitable high voltage
mains transformer I'm using not 1200VDC but something lower ie.
whatever results from the one I'm fitting (550v-0-550v).
To some extent the front panel controls interface
with both the LT and HT power supplies and I'm planning on keeping
this feature although for testing purposes I'll add over-ride
facilities.
Another feature of the T1154 is the ability for it
to use either a carbon mike or an external speech amplifier (A1134).
I plan to add an equivalent later. Switching from one to the
other requires an internal switch to be adjusted.
Below I've shown a simplified block diagram of a typical
installation, and below this a simplified circuit diagram of
the transmitter. |
 |
 |
|
|
Power Supply
|
|
After a pause of a few years I'm now building
a dedicated power supply for the T1154. It will be the same width
and similar depth to the transmitter and sturdy enough for it
to act as a base. Naturally it will be built chiefly from junk
box parts and whatever materials are to-hand. Building it will
also need to fit in with circuit board and lift drive repairs
which today (4th Oct 2022) are all up-to-date. |
 |
|
The base and sides are three-quarter inch plywood
with the front and back panels each half the side of a PC case.
Amazingly the metal side was exactly 16.5 inches which is the
width of my T1154. I used a 6 inch bench vise to bend the metal
to fit.
The height of the case is 7 inches and the metal PC
side just accommodates this
The main parts I've selected are a transformer with
windings rated at 550v-0-550v 285mA plus an amp at 6 volts (this
will be for an "ON" lamp), an LFC hopefully with a
sufficient rating, and an 8 volt 4A transformer. Initially I'm
planning on a full wave diode rectifier with four high voltage
electrolytics wired in series-connected pairs (these are from
a scrap 3-phase drive unit). The 6 volt DC supply will use a
spare LT1083 regulator
(I bought two of these a few years back).
Nice to have will be a front panel meter to read either
HT or LT. I've yet to devise a way of wiring the PSU to the T1154.
I also bought a lead-acid 6-volt battery to provide low voltage
stability which I'll fit in the case.. if I can find it! |
 |
|
I decided to buy some fuseholders and toggle switches
plus a couple of panel-mounted LEDs to make the PSU look presentable.
The relay switch (S1) can only operate when its coil
is energised from the LT supply ensuring HT cannot be applied
without LT. A tentative circuit diagram is shown below. I'm not
sure about the method of protecting the 100uA meter just yet,
but I've shown a couple of clamping diodes for the moment. I
may decide to shunt the meter for further protection. The small
winding on the HT transformer I'll use for lighting LED1. I think
LED2 will be across the input to the regulator rather than the
output because the meter will indicate the latter, similarly
I might add a second relay contact in series with LED1 so that
this illuminates only when HT is developed.
There are two features to be incorporated in the new
PSU. These are start signals for the LT and HT rails which will
use relays driven from the T1154 controls. |
|
I selected from my junk box a large moving coil
meter marked 100uA chiefly because its scale matched the maximum
values of 1000 volts and 10 volts, for the HT and LT which should
end up as about 700 volts and something a little over 6 volts
(to accommodate losses). I cut a 90mm hole for the meter and
I'll be mounting three or four fuseholders and several toggle
switches viz. mains for HT (S3), LT (S4), a stand-by switch (S2)
for operating a small relay in the HT transformer centre tap,
and an ON-OFF-ON toggle switch (S5) for the meter. |
 |
|
|
In addition I still need to work out connections
to the Jones plugs where these are dealt with in the PSU, but
below is the basic PSU circuit which will be amended once details
are worked out. |
 |
|
Having ordered some new parts for the front
panel to give it a better look I now need to evaluate additional
circuits for wiring the Jones plugs so as to make the T1154 transmit.
It's now 2022 and I see above it was 2015/2016 that I last experimented
with the rig. Fortunately I'd explained a little about getting
the T1154 to actually transmit and I can make a few modifications
to the circuit shown above to deal with TX/RX. For example..
S1 switches the HT line on or off and the switch S2 which I was
going to call Standby/Off can be the TX/RX switch. Of course
one doesn't need to put the HT onto the transmitter unless its
in transmit mode. The original T1154/R1155 designers used an
HT START wire to turn on the HT rotary generator. Because I'm
using a relay to turn on the HT (incidentally the reason for
this is because the ratings of cheap commercially available switches
fall far short of the switching voltage at the transformer centre
tap. In fact a cheap relay would probably not be man enough for
the task either, but as I have loads of scrap lift drive units
that switch 600 volts I'll be able to salvage one although there's
a slight difficulty dealing with its coil voltage.... |
|
To clarify TX/RX switching I'll draw up a circuit
here.
Essentially we're dealing with three Jones plugs viz.
D (8-way), B and E (4-way) |
 |
|
This Jones plug (E) is fitted to the right-hand
side of the transmitter and has connections to the receiver HT
line (+/-220V), the HT START voltage and a voltage for setting
the aerial selector switch, nominally 12VDC.
You can also see the pair of aerial connectors and
a ground pin.
The TX/RX relay has three coils A, B and C which are
used for connecting the two aerials (MF and HF) to the receiver
or transmitter |
 |
|
 |
 |
|
Left.. Jones plug numbering and above a typical
connector pair |
|
|
|
T1154 8-way Jones plug connections:
1 220V minus; 2 220V plus; 3 Start LT PSU; 4 12V/24V
for PSU start relay; 5 LT minus; 6 LT plus; 7 1200V minus; 8
Start HT PSU; with 1200V plus carried on the single pin connector.
Below, the T1154 transmit/receve relay connections |
 |
|
As a relay coil supply voltage of around 12V
is necessary, I'll use the spare low voltage winding on the LT
transformer, this is marked 18-15-0-15-18V at 1Amp. I'll use
this to drive the relays for starting the LT and HT as in the
amended diagram below.
S1/S2 is a Test switch for turning on both LT and
HT supplies (ie. over-riding the T1154 switches) whilst S6 and
S7 are operated by the T1154 via the pair of relays labelled
LT (=START) and HT on (=START). S3 and S4 turn on the two transformers
in Standby mode with only the low voltage winding on the HT transformer
and the 15v-0-15v winding on the LT transformer providing current
(to the pair of LEDs). In fact I might add a set of diodes and
operate the pair of relays from a single toggle switch in place
of S1/S2. |
 |
|
A few years ago I tested the T1154 and
worked out that it would be convenient to use a three position
switch to select receive (centre), CW (by grounding connector
B Pin 14) or AM (by grounding connector B Pins 13 and 14). A
morse key can be fitted between pin 13 and ground. It's convenient
to fit the switch and the key jack on the front panel of the
power supply. Thinking ahead for better audio I could build a
small amplifier (a transistor version of the A1134?) and microphone
socket within the PSU also. |
 |
This circuit which saves on a switch would let
me select transmit CW or RT from Receive by grounding (via connector
B15) the key control connection B14 for CW or grounding the key
control connection at B14 plus the key connection B13 for RT.
Selecting neither leaves the TX/RX relay in the receive state
(using the +12V connection). Overall control is via the mode
switch on the T1154 panel. The morse key is across B13 and B15.
Diodes can be 1N4007. A cable from the PSU to connector B4 is
required which can also carry the headphone link at B16 from
the R1155. |
|
|
As the front panel switches are arriving tomorrow
(7th Oct 2022) I'm now planning on how to connect the PSU to
the T1154. Ideally the transmitter should stand on top of the
PSU and I have a few choices.. run a cable directly from the
front of the PSU to the T1154 Jones plug or alternatively fit
a matching Jones socket on the front panel of the PSU and make
up a cable. The latter choice is better in respect of using the
PSU for something else but the former saves some metal bashing
(I'd still need to drill a hole and fit a suitable grommet) and
the cost of the (two) extra connectors. A third choice might
be to add spacers between the T1154 and the top of the PSU and
run a cable directly from the innards of the PSU. That method
is attractive as any excess cable can be pushed inside the PSU
case and I suppose the gap would provide cooling. It would also
completely remove any need for cutting the front panel.
The final option does have another advantage as cables
to Jones connector B plus the microphone connections can be added.
This will mean adding a pair of jack sockets to the PSU front
panel. As an afterthought I'll use a rotary switch for the meter
as having three full settings will accommodate monitoring of
the 12VDC supply. |
|
It boils down to aesthetics I guess? |
 |
|
|
Nice to have would be a few more LEDs. The T1154 sends
out signals such as LT START and HT START which could be
monitored from spare relay contacts and a couple for CW transmit
and RT transmit could easily be added. It could aid fault-finding
and keeping tabs on the thing which I recall from previous testing
could sometimes be a puzzle.
Thinking ahead. If the PSU and T1154 combination works
reliably and I look for increased RF output the mains transformer
could be pressed into service using the full winding of 1100V
but as the peak might rise as high as 1540V the PT15s may suffer
as their spec tells me 1250V is the advised maximum and as the
valves are as old as me that even that voltage may be a bit dodgy.
I looked deeper into my junk box and found some RG1-240A mercury
rectifiers. A solution could be to use a half wave rectification
(an idea of G3XGW). I could use one RG1-240A with a reservoir
condenser of say 4uF and using an LFC and an 8uF smoothing condenser
the HT might be manageable. The rectifier would drop 16V and
half wave plus the choke resistance should knock the peak voltage
even lower.. I can even dispense with the reservoir condenser
and use a choke input circuit. All this is in the future though
as the initial plan is to use 700 volts.
I measured my LFC and it read 22.48H and 500 ohms
but I heard that the inductance of such a choke will reduce as
current is increased so experimentation will be the order of
the day. I guess I could measure the output voltage of the new
PSU from off-load to say 200mA with and without the reservoir
capacitors. That will give me the sort of voltage range if I
were to use the full winding with either half wave or with a
bridge rectifier. In order to complete this exercise I should
consider an HT bleeder resistor as this will give me protection
against too high an off-load voltage. I guess I could spare say
20mA making a suitable bleeder 67Kohm at 23W (that sounds excessive!).
That being so maybe 10mA is better giving me 100Kohm at 15W maybe
a big green ceramic job? Then the penny dropped when I realised
the transmitter is full of big green ceramic resistors.... I
measured the T1154 HT pin to ground and it read 20Kohm so I don't
need a bleed resistor, however for balancing the voltage across
the capacitors, and for safety (discharge of the high voltage
when the PSU is turned off) I'll fit some resistors as shown
in the final circuit. |
 |
|
Above, the front panel drilled and fitted with
most of the parts. I need to add a jack socket for the microphone,
but that won't be required in the first stage where I'll be using
the carbon mic plugged into the T1154 front panel. I decided
to use the pair of LEDs to reflect the state of the relays I'm
using for LT START and HT START. I also fitted a master switch
for the mains supply to the left of the IEC mains socket. I'm
using a 5-way meter switch to eliminate bridging contacts when
going through the settings (this is because the meter is only
100uA fsd). The switches are S3: LT transformer on/off, S4: HT
transformer on/off, S1+S2: Standby/HT ON plus LT ON and S8: CW/Rx/RT.
Later I'll amend the circuit to show how I'll activate
the pair of relays from the Standby switch. Basically I'll be
using this switch together with 4 diodes to take the place of
the T1154 LT START and HT START signals so I can test the PSU
without using the T1154. When the Standby switch is ON the two
LEDs should come on. I'll rename the switch Test/T1154 and as
a safety precaution I'll fit a third relay which will disable
the test feature otherwise full HT could be turned on when the
T1154 HT START signal is absent. The circuit above, once amended,
will show S1 and S6 as the same set of relay contacts as will
be S2 and S7 for the second relay. The LFC will be repositioned
in the earthy side of the HT supply to avoid breakdown between
its winding and ground, and as the PSU uses a wooden baseboard
the LFC case may have been live but for safety I'll be grounding
this as well as the metalwork of the two mains transformers (and
the front and rear panels). |
 |
|
It's quite nice working with wood as the drills dont
wander around as much. I hope I've orientated the big transformer so
the windings are at right angles to the others, although in this
application I guess its not important. Next, I'll be collecting
together some capacitors. I have lots of scrap lift drive units
with several struck by lightning thus writing them off but all
have banks of high voltage capacitors used for smoothing rectified
3-phase mains with a few having a high voltage/high current relay,
one of which I'll use for the HT circuit. Before I start wiring
I'll need to update the circuit diagram.
The T1154 spec refers to Plug D Pin 4 as carrying
either 12v or 24v and for some reason I assumed this was required
within the transmitter circuitry, for example to hold the aerial
relay in the receive position, but I then realised that the specific
voltage relates only to the aircraft power supply. The high voltage
and low voltage PSUs that match say 24v use 24v relays to start
up the rotary generators. This means the voltage needed to emulate
the control circuits in my PSU can be 24v and doesn't need to
be stabilised as relays aren't too fussy, but more importantly
the only high voltage relays in my junk box have 24v coils. |
 |
 |
I selected 4 EPCOS 1000uF @ 450V working with
two in series for both reservoir and smoothing. For the higher
voltage option (later) I'll use 4 in series for smoothing without
a reservoir.
The type I have use the 5-pin mounting arrangement.
Below is the latest circuit.
I drilled the panel and fitted a jack socket for the
morse key also an HT mA has been added (read at the end of the
page). |
|
|
 |
|
Rather than have to pull the PSU apart later I decided
to check the LT arrangements before fitting and wiring everything
up. |
 |
 |
 |
Here are the salient bits. I assembled the LT1083
kit which I bought a few years back and coupled it to a 10 ohm
rheostat. The spec claims 7A max output and a headroom of 2.5v
which lines up with the test results using a bench supply.
The bench supply has a max output of only 3A which
limits testing to this figure. I could run to 3A output from
the LT1083 with my rheostat set to 2 ohms and the input voltage
at 8.5v. The LT1083 module has a set of 4x 10A10 diodes wired
as a bridge but connecting across its output with a DC supply
is OK for initial testing. The 8v junk box transformer might
just about be OK.
The valves use about 4A and the Tx/Rx relay not too
much. The 8v transformer looks fairly substantial and the mains
settings will give me a little flexibility.
Max DC into the LT1083 will be about (8-1.6)x1.414=9v
so there's not a lot of slack available and bumping up the smoothing
capacitor to 10,000uF or larger from the 4,700uF fitted will
probably be necessary. |
|
|
I found this schematic that seems to match the
kit.
This is my modified "fixed" output version
where I changed R1 from 100ohms and R2 from a 5K pot. |
 |
|
|
As I'd half expected the 8v transformer failed
to provide enough voltage for the regulator when driving 4A.
Although the terminal voltage still measured 8VAC the losses
in the diode bridge (which increase with current) reduced the
regulator headroom below its minimum. Things improved with extra
reservoir capacity but only 5v at 5A could be obtained and shifting
the mains tapping to 230v only improved the output to 5.2v. A
poossible option would be to replace the diodes with power FETs
but a junk box search is easier...
This yielded a toroidal mains transformer with an
output voltage set of 200v/21v/6.5v which worked OK, giving me
an output of 6.5v at 5A with about 13v at the bridge rectifier
output, but the heatsink was running extremely hot and, as this
transformer has a 200 volt winding, I decided it might be pressed
into service for something more fitting. During this test I found
the 5k pot was not very easy to adjust so my next step was to
change the setting resistors to deliver a fixed output as close
as possible to 6.5v (this allows for a slight cable/wiring loss).
The LT1083CP output is set by a function of R2 and R1 together
with a fiddle factor ie add one and multiply by a chip constant
of 1.225 to 1.270 so I used a 120 ohm and 510 ohm to get 4.25v,
then 5.25v by adding one, and using the chip constant figure
a range of 6.43v to 6.66v which averages to 6.545v. To be safe
I'll fit a 7.5v zener diode across the output which should open
the 7A fuse if the output were to rise too high. I then ditched
the 21 volt transformer because the LT1083 got incredibly hot
and anyway it will make a good basis for an R1155 power supply..
Next candidate was an ancient Radiospares transformer
marked 12v + 12v which was the same weight as the 8v example
so about the same VA. I measured the output (fortunately two
discrete windings) as 11.8VAC for each and with a load of a little
over 2ohms I got 5A before the terminal voltage went down 10%.
I connected the two windings in parallel and found I could easily
draw 6A from the LT1083 module. The benefit of operating close
to the LT1083 headroom limit is the dissipation is minimised.
The final choice then was a pair of paralleled 12v windings which
gave me 6A into a load of about 1ohm with an output voltage around
6.25VDC. In the circuit above the LT1083 now has fixed resistors
of 120 and 510 ohms. |
 |
Mostly assembled and now completing the wiring. |
|
|
To keep the LT1083 extra cool I added a larger
heatsink and additionally bolted it to a piece of aluminium (below
left). Just out of interest I checked continuity and as I'd half
expected the heatsink was sitting at the output pin voltage.
In my case the side of the case where it's mounted is plywood
so as long as nothing touches it the thing will survive. The
live heatsink could have been avoided by using an insulating
pad between the LT1083 and the heatsink but this reduces the
overall efficiency of the heatsink so I didn't bother.
Although the complete circuit looks relatively simple,
wiring it up is taking ages. I decided to mount the HT output
fuse on the rear panel close to the HT wiring. Next, I'll carry
out some tests on the meter circuit using external variable power
supplies, then fit suitable fuses and test the LT, HT and relay
power supplies. |
|
Top left is the 24v relay supply. The control
wiring isn't done yet and is chiefly adding two more relays then
wiring the morse key jack socket and the transmit/receive switch.
I'll fit a protective cover over the HT rectifier
diodes later as they're a bit exposed and the bleed resistors
take several minutes to leak away full charge.
Top right is the high voltage contactor which switches
the centre tap of the secondary winding to the LFC fitted in
the earthy side of the HT supply. |
 |
|
|
The LT voltage measured 6.4v at the meter and
6.57v on my multimeter. The relay supply measured about 31 volts
off load and the HT registered 780v at the meter. I'm switching
not only the positive connections to the meter, but the negative
connections as well. The reason being the HT isn't grounded because
it feeds T1154 bias circuits, and in order to prevent accidental
damage to the 100uA meter I didn't use the switch setting next
to the HT position. This leaves a guard position to prevent the
HT negative spiking the meter as it's replaced by ground for
the low voltage supplies.
I added a forward-biased diode across the meter terminals.
Because the meter is capable of reading 50mV full scale the diode
has to conduct as closely as possible to something like 100mV.
I tested various junk box diodes and found an OA10 gave be less
than 200mV at a few mA of forward current. Not ideal but better
than nothing. I was prompted to do this (and leaving a guard
position) because the test meter in my CJD
Receiver is open circuit almost certainly because of weak
circuit design.
I now need to wire up the test switch. Once the T1154
is connected the LT Start and HT Start signals will automatically
turn on the power supply, but I need to force on the LT and HT
supplies independently of the T1154 for testing. The circuitry
includes 1N4007 diodes which perform an "OR" function
for the Start relays, spike suppression at the relay coils, and
logic functions around the transmit/receive switch. Note that
RL3 relay is used to prevent or turn off local control of the
HT supply if an LT Start signal is received from the T1154. |
 |
|
A couple of small changes.. LED1 and LED2 will
indicate the presence of voltage fed to T1154. LED1 is driven
by a low voltage derived from R7 and R8 to avoid running HT to
the front panel. The drain is set at about 3mA (about 2mA to
LED1 and 1mA to ground) and will aid the discharge of the capacitors.
The resistors are fitted after the HT fuse whereas the meter
measures HT before the fuse. LED2 monitors the LT to the Tx but
another (blue) LED is fitted in the LT1083 module. I used a strip
of Veroboard on which to mount the two LEDs and their feed resistors
R11 and R12 both 1.2Kohm.
All done! Final testing is next on the agenda. Test
the LT and HT Start signals to see if the correct relays kick
in then check the LT and HT under full load conditions then I
need to work out the length of the cables from the PSU and T1154
plugs.
Below, I tested the LT and HT Start logic and it worked
OK. The final checks are to load the LT and HT to see the full
load voltages.
Centre, you can see the termination strip for the
cables to the T1154 Jones plugs (Left to right C, B/15-B/13,
and D/8-D3) |
 |
|
Probably the second most important thing I need
to consider (the first being maintaining a perfect filament supply
to the transmitter) is the HT voltage. Calculating the characteristics
of this is a bit of a black art but suffice it to say one hardly
ever sees the HT voltage expected. There are two fundamental
things one should consider.. these being the AC side of things
and the second, the DC aspects. Obviously if a transformer has
its voltage marked on the side then one can expect to see this
when the primary is run at its design voltage.. but not so. A
decent transformer will have a marked "VA" rating and
this is a measure of the rated voltage and current at the input
when its producing its rated output. A transformer designer has
practical and financial considerations to consider such as its
physical size and the cost of materials used in its construction.
The key thing I'm considering is the resistance of the output
circuit. As output current increases the losses due to heating
for example will reduce the terminal voltage so it stands to
reason that the no-load terminal voltage will be higher than
the figure marked on the case and in my example the rating 550v-0-550v
at 285mA will result in a higher figure (which I measured as
588v-0-588v). The true figures are related however, not solely
to DC conditions (mere resistance) but to AC conditions. So,
without venturing into equations, I'll suggest the inductance
of the secondary winding, in fact suitably modified by effects
of the primary winding, plus the diameter and composition of
the secondary winding wire will determine the AC and DC aspects
of the secondary.
Once the cicuit is constructed actual measurements
can be made. If one uses best practice, such as the use of "suitable"
capacitors and an LFC then one can hope to get the best from
our junk box components. If we reconsider the effects of the
components some tweaking might be possible but not the general
practical outcome.
I measured the output voltage under only a tiny load
(the set of balacing/bleed resistors at the capacitors) and found
it to be 780v on the meter and 796v the next day (it depends
on the mains voltage). This figure nicely equates to the expected
result ie. 550v x 1.414 = 778v. But was it 550v and what exactly
was the primary voltage at the 240v tapping? I'll check these.
The readings today are 588v x 1.414=831v and a reservoir capacitor
voltage of 796v with a load of just the bleed/balancing resistors
suggesting some ripple which is affecting the DC multimeter.
Mains voltage was 239V and quite wobbly.
The transformer markings declare 285mA at 550v, but
the winding is 550v-0-550v and the way the winding is designed
to be used is with a full-wave rectifier so I reckon the full
winding (0-1100v) will accommodate not 285mA but half this or
142mA. DC-wise the resistance might be 1100v/142mA=7.7Kohm, but
this is clearly rubbish as the DC dissipation would be 155watts,
so forget about DC. The transformer wouldn't actually work properly
if DC is allowed to flow in the windings. In fact the secondary
winding DC resistance is 57+65=122ohms (why are the halves different??)
and the primary winding a mere 5.4ohms. The secondary losses
will be a factor of 122ohms and the characteristics of the AC.
Obviously the primary winding uses 50Hz in the UK but what about
the secondary? Full wave rectification runs at 100Hz but uses
each half of the secondary alternately. Anyway you can see the
theory is not straightforward except we can more or less assume
we can see 550v across either half of the secondary at the rated
full load of twice 142mA.
Turning to the DC side of things. The rectifiers are
essentially dumping power into the reservoir capacitor and under
very little load the capacitor will charge near to the maximum
(=peak) voltage from the rectifiers. Because the 550v is quoted
as RMS the peak voltage is 778v (under the rated load) and this
is essentially the same whatever the capacitor value, however
the whole reason for building the PSU is to supply power to an
external load so as quickly as power is dumped into the reservoir
capacitor the quicker we may be removing it. The design must
therefore use a large enough capacitor to keep it charged as
high as possible under load. Practical considerations decide
the size of the capacitor... Next we look at the sort of DC voltage
at the reservoir capacitor. The rectifier, the capacitor and
the design of the transformer will result in a DC supply at the
reservoir which will vary from almost clean to horribly hummy
(100Hz ripple) so we need to add some filtering. The filtering
will need to minimise the 100Hz ripple and now we're back to
AC considerations, which means not just DC resistance but also
capacitance and inductance (if an LFC is used).. My LFC has a
DC resistance of 500ohms which isn't too bad but not good.
Drawing our 285mA through 500ohms results in a loss
of 285mA x 500 = 142v so my full load output voltage cannot exceed
778v-142v=636v. Here I'm assuming that the transformer output
figures take into account the secondary losses. A valve rectifier
would drop a fair bit of voltage but I'm using silicon diodes
whose DC forward resistance is tiny. I'm using a smoothing capacitor
equal in value to the reservoir capacitor for convenience and
this serves two purposes. Firstly it acts with the LFC and reservoir
capacitor to filter out the 100Hz ripple and secondly it acts
as a second reservoir for our DC output.
How about working out the filter values for say 50Hz
or 100Hz? A ready reckoner tells me I need 20H or 10H and a couple
of microfarad, so my capacitors (say 4 x 470uF in series.. worst
case) will be fine. The LFC I'm using measures 22.48H with my
meter but of course the current through it will be mainly DC
and as this increases the inductance drops so as more current
is taken by the load the more unfiltered ripple will appear and
of course as the inductance decreases we need more capacitance
to re-establish our 100Hz filter. A rough and ready calculation
tells me the HT should be clean enough.
How do I check the HT? At 285mA and 636v I'll need
a dummy load having a rating to handle the full output of 550v
x 285mA= 157watts. If I see 636v then the max current should
be say 157w divided by 636v=247mA. A resistor of about 2.5Kohm
rated at 150w will do. Fat chance of that so let's look in the
junk box... but I thought of a better idea.
A simple option is three 240v x 60w tungstem lamps
in series if indeed I see something like 700v, but ideally I
should use a variac and crank up the mains voltage to get the
lamps glowing rather than place what is effectively an instantaneous
short across the HT output (three 60w lamps = 3 x 75ohm when
cold). Left.. ensuring they weren't touching the biscuit
tin and below my rather battered and thrice blown up and repaired
Variac now mounted in an old computer case. |
|
|
|
I connected my Variac and increased the setting
from zero but nothing happened. I then realised that the PSU
test feature uses 24v relays and as the 24v transformer was only
seeing a low input voltage the relays weren't going over hence
the table below starts at 400v which is where the test relays
activated. If you do some calculations on the maximum output
voltage achieved, the LFC loss is 500ohm x 0.225mA = 112V which
added to 707V gives 819V (peak) or 579V RMS. This means the transformer
is running at a slightly elevated output voltage at a lesser
current. This is likely to be because the load impedance of the
lamps isn't quite matched to the 2.5Kohm optimum load (in fact
it works out at 3.14K). |
|
Voltage across load |
Current in load |
Total power |
|
403.5V |
168.5mA |
68W |
|
510V |
189.9mA |
97W |
|
597V |
206mA |
123W |
|
707V |
225mA |
159W |
|
The load was three 60w 230v lamps wired in series.
The Variac setting was halted in roughly 100v HT steps
up to maximum.
The HT transformer is rated at 157w |
|
|
Next, I checked the LT. This was achieved using a heavy
duty 10ohm rheostat although the ammeter leads also figured in
the load. |
|
LT Current |
LT Voltage |
LT Power |
|
1.6A |
6.51V |
10.42W |
|
2.06A |
6.50V |
13.39W |
|
3.035A |
6.46V |
19.61W |
|
4.15A |
6.42V |
26.64W |
|
5.03A |
6.38V |
32.09W |
|
5.75A |
6.17V |
35.48W |
|
6.42A |
5.85V |
37.56W |
|
These are the test results for the 6v LT supply.
The LT1083 module is working from two paralled 12v RMS windings
which for some reason were only 11.8v off load. As the test current
increases there are increasing resistive losses in the wiring
and in the relay driven from the T1154 Start signal (or in this
case the Test switch). These, together with the diminshing headroom
for the LT1083 chip and the 4700uF reservoir mean that the output
voltage drops off. The loss is 5% at 5.75A and 10% at 6.4A.
If necessary I could always add 10,000uF additional
reservoir capacitance to improve regulation over 5A. |
|
|
For practical reasons as well as for safety
I fitted a lid, on which the T1154 will stand. Talking about
safety... I ordered a 500gm reel of 60/40 Tin/Lead multicore
solder and it arrived today complete with 14 pages of EU safety instructions but as far as I
can tell as long as I'm not pregnant or breast feeding a baby..
which I'm not.. I should be OK. To be really certain I'd have
to read a further couple of dozen directives including the effect
of the flux on various sea creatures including algae.
As far as the recommended respirator, goggles and
chemically resistant gloves are concerned I'm not going to bother...
Below right the cable method I chose which allows
for enough freedom to permit transmitter testing with the PSU
set to one side.
I left a reasonable space inside for cable excess
(around 12 inches more than is visible here) to be stowed away. |
 |
 |
 |
|
|
All that remains is to plug it all together
and press some switches...
Not so fast though.. I put this project on one side
for various reasons and now having reread this blurb in March
2023, I realised that I have no idea what the switches or LEDs
are for.. so.. my labelling machine will be pressed into service
and this will put things right.
By April 2023 I'd decided to add an HT milliameter
and, as I'd just discovered the perfect one for the job in a
box of old radio parts, I cut a hole in the front panel and fitted
it.. just the additional wiring, producing a few labels, and
I'll be ready to fire up the T1154. |
 |
|
The left meter reads 3mA so will be shunted
to give 300mA full scale for the EHT line. The right meter is
switchable and reads the 6 volt, 24 volt and 700 volt rails. |
|
|
|
Key to switches and relays:
S1= turn on EHT trnx primary only, S4= turn on mains supply and
24VDC control voltage, S2= turn on 6 VDC trnx primary supply,
S3= test EHT when LT start present (=RL3 activated), S5 transmit/receive.
RL2 inc RL2a= activated on LT start command. RL3= safety relay.
RL1a= EHT on from HT start command. RL3a= control for EHT on/off.
S6= voltage rail meter select |
|
Just summarising.. The power supply looks over-complicated
because I've attempted to emulate a complete R1155/T1154 system
rather than merely just power the transmitter. For example, a
number of system components in say an airbourne rig include special
features such as the provision for conserving battery life plus
not wearing out electromechanical items. The numerous "frills"
in my PSU mean that operating the mode switch on the T1154 will
produce the results expected in an airbourne rig. Of course I
might have made the odd mistake so the circuit shown above might
get modified. An interesting bonus is the 24 volt transformer
extracted from the junk box and which is bolted to one side of
the case hums. I was going to swap it but it does let me know
when the PSU is turned on and in stand-by mode so I'll leave
it.
The last task... fitting the (basically 3mA) HT current
meter was a bit iterative. I realised that inserting it in the
HT positive lead was not the best plan (much like the T1154 designers
change to the HT fuse position). I moved it to the negative feed.
Testing such a meter is easy if one has a low voltage power supply
with a current limiter. One makes up a shunt resistor with a
length of insulated thinnish wire.. say 18 inches and solders
this across tags on the meter terminals. Set the PSU test voltage
to a nominal figure such as 2 volts and the current to say 100mA
and connect to the same solder tags. Once the reading and shunt
resistor (suitably pruned) produce the correct 100mA (=1mA) meter
deflection increase the current limit, in my case to 300mA (=3mA)
and confirm this is OK. Then disconnect the shunt and coil it
up on say a length of quarter inch plastic tube, glue it to the
back of the meter and solder the coil ends to the meter tags.
Gluing it in place helps prevent the thin wires from breaking.
Now I must make some labels for the controls as my
memory is not what it used to be... |
|
Now that I finally decided to retire from repairing
lifts I'm tidying the workshop to restore it to my radio museum
plus shack. In the process I noticed a couple of big transformers
that had been hidden away, one is a huge autotransformer but
the other has a message from my father-in-law G3AQY to his XYL
G4GOJ (alas now both silent keys) saying "Do not throw away
Gill.. spare for Heathkit SB1000 linear". Well.. would this
do for a full power T1154 HT transformer? I found details on
the Net suggesting the secondary has a winding of 790 volts with
a pair of lower voltages. One is rated at 5.2 volts 15 Amp and
the other I'm not sure about.. maybe around 12 volts ? The SB1000
uses a full wave doubler to develop a whopping 3100 volts and
this drops to 2700 volts at 500mA. This means the SB1000 transformer
will be a perfect option... providing it will fit in the case
that is! If not then perhaps I'll be building a Mk2 PSU. Peak
output = 790 x root 2= 1117 volts and with its huge current rating
I can't see it dropping too much from this level.
Size: 5.5 inches x 4.5 inches x height 5.25 inches
Just enough room in the box. |
 |
|
|
Just one thought though... it was common for this
particular transformer to fail in the Heathkit SB1000. Rumour
has it that they would develop a leak between the high voltage
winding and one of the other windings. Why would this be a common
problem when transformers are not exactly state of the art components?
It was also said that a new version was produced viz. the "54-1063"
made by a different manufacturer. Mine is the original type so
is it going to be reliable? Could the fact that the SB1000 used
a full wave voltage doubler have had a bearing? This produced
a whopping 3100 volts off-load. At least in my application I'll
use a half wave rectifier. Below a section of the Heathkit construction
manual showing the alternative US//UK mains options. |
 |
|
I tested the SB1000 transformer and it all turns out to
have been a waste of time. Despite the output being 790V on the
drawing above it was actually 1190V off-load and in fact no different
to the old transformer marked 550-0-550 (= 1100V) |
|
pending |
