|
|
|
|
|
|
Solid-state replacement for electromechanical vibrator (EMV)
- under construction
In this article we propose a generic solution
for replacing the electromechanical vibrator (EMV) in vintage
valve-based radios by a robust modern solid-state alternative (SSV),
which we've named Zerhacker — the German name for an EMV. The
design consists of two PCBs with SMD parts, and is small enough
to be fitted inside the cylindrical aluminium enclosure of the existing old EMV.
|
Author
Marc Simons
Publication date
25 December 2024
Version
1.00
Keywords
Electromechanical vibrator, spy radio set,
valves (tubes), Solid-state vibrator
|
|
|
|
In old radio's and transceivers with valves (tubes), an electromechanical vibrator
(EMV) is often used to convert the battery voltage into a higher voltage for
the anodes of the valves. This voltage is usually in the 75-300V range, and is
commonly known as High Tension (HT).
In practice, especially in mobile use, battery voltages between 2 and 12V are
common. In most cases, this voltage is also used to power the filaments (heater)
of the valves. This is known as Low Tension (LT).
|
The battery voltage (LT) is then used — with the help of an
electromechanical vibrator (EMV) and a transformer —
to create the HT.
An HT voltage between 75V and 300V is quite common,
depending on the design of the radio and the valves.
The EMV first converts the DC battery voltage into an AC voltage,
which is then fed to a transformer. This transformer converts the LT/AC
voltage into a HT/AC voltage, which is then fed back to the EMV,
where it is converted into a double-phase rectified HT voltage of the desired
magnitude. Most EMVs run at 100-150 Hz.
|
|
|
The problem is that EMVs are quite old, some even more than 80 years.
After so many years, the contacts are worn out and are in most cases badly
corroded. The latter is caused by the ozone generated by the switching contacts,
in combination with the harmful gasses emanating from the rubber gaskets of the
EMV. This renders many EMVs useless and beyond repair. And even if they
work today, they will certainly meet their end of life at some point in the future.
This is frustrating for collectors and living museums, who want to demonstrate
the equipment to the public.
In many cases, the actual radio set is still operational and many hours
have often been spent on its restoration, only to find that the EMV is
broken and renders the device useless.
This problem can be overcome by replacing the EMV with an electronic
alternative in the form of a solid-state vibrator (SSV). Such SSVs
have many advantages over the old EMV, such as:
|
- No audible noise
- Glitch-free switching
- No RF switching noise
- No contact wear
- Energy efficient
- Small form factor
|
|
In this article, we propose a generic solid-state vibrator (SSV), that
can be used as a replacement for an existing vintage EMV, and that can be build
inside the existing enclosure of the old EMV.
|
|
In the Crypto Museum collection we have quite a few
vintage spy radio sets that
can be powered from the mains as well as from a low-voltage DC source,
such as the battery of a car.
The latter was important when the device was used in areas where
no mains power network was available.
|
For that reason, such radios often contain an EMV to convert the battery
voltage into the HT voltages needed to power the receiver and the transmitter.
More often than not however, the EMV is broken, corroded and/or beyond repair.
Examples of such spy radio sets in the Crypto Museum collection,
are the wartime
British Type 3 Mark II (B2),
Type A Mark 3 (A3)
and the American SSTR-1,
but also post-war (Cold War) radio sets
like the Mk.122,
12-WG and
RS-6.
We think it would be great if we could demonstrate these device to the public
in the years to come.
|
|
|
The idea to replace an EMV by an SSV is not new. Numerous circuits have
been published in electronics magazines over the past decades, but unfortunately
most of the designs are old and only solve the EMV problem for a single
type of radio. This prompted Crypto Museum to rethink the problem
and come up with a more generic solution that can be made with modern components.
Before we started, we defined the following design critaria:
|
- It should work with an input voltage between 2V and 15V DC.
- Use the existing transformer.
- It must be suitable for generating an HT voltage between 60V and 300V.
- All parts must be fully isolated, so that the LT input voltage as well as the HT output voltage can be wired any way around.
- It should cope with any existing scenario.
|
One of the problems we identified is that it is uncertain how an EMV is
wired inside a random radio. There are 2 ways to connect the LT voltage to the
transformer: with either with the (+) terminal or with
the (-) terminal of the battery connected to the centre contact of the primary
winding of the transformer. The same uncertainty exists at the HT side,
where the centre contact of the secondary transformer winding goes to the
+HT rail or to the -HT rail (Ground).
Some possible configurations are shown in the simplified
circuit diagrams below. The diagrams show a simple valve-based circuit
around V1. At the left is the transformer (TR1) which has a centre contact
at the primary and at the secondary side. The outer contacts of the primary
side are connected to the (no) and (nc) contacts of S1a.
The outer contacts of the secondary side are connected to the (no) and
(nc) contacts of S1b. Switches S1a and S1b are operated synchronously
at 100-150 Hz
by an additional coil inside the EMV which is not shown here for simplicity.
|
| |
Configuration 1
Centre contact of primary winding connected to (+) terminal of the battery.
Centre contact of secondary winding connected to HT+ rail.
|
|
| |
Configuration 2
Centre contact of primary winding connected to (-) terminal of the battery.
Centre contact of secondary winding connected to HT- rail.
|
In the above configurations, switch S1b is used as a double-phase rectifier.
As a result, the HT voltage will have the same polarity as the LT side.
Reversing the battery terminals will therefore also reverse the polarity of the
HT voltage. It is important to verify this, prior to connecting the valve-based
circuit. Note that the secondary switch (S1b) is not used in circuits where a
double rectifying diode is present in the circuit. This situation is shown in the
simplified diagram below.
|
| |
Configuration 3
A double rectifying diode (V2) is present at the secondary side of the transformer.
In this case the battery may be connected either way around.
|
|
If we want a generic solution, it must be able to cope with the above situations.
This means that there must be a full isolation between the LT and HT switches.
Furthermore, there must be a full isolation between the LT switch and its
control circuit.
The real design challenge is the switching at the primary side of the transformer.
We solved this by using a pair of N-MOSFETs in back-to-back configuration, whith their
gates and their sources connected together, as shown in the diagram above.
Both MOSFETs will be switched on at the
same time to form an 'electronic contact', independent of the direction of the
current. If we use two such circuits, we can emulate a regular
Single-Pole Double-Throw (SPDT) switch, like the one that is present in the
original EMV. The gates of the MOSFETs must be driven in an isolated way.
Implementation of the HT switch is relatively simple and can be solved by using
a pair of high-voltage diodes. Note however that, depending on the circuit, it might
be necessary to reverse the polarity of the HT voltage, as we will see later.
|
The diagram below shows the oscillator circuit. IC5 is a HEF4060
Ripple Counter/Oscillator.
It has been dimensioned in such a way that the Q3
output delivers ~76kHz and Q9 delivers ~1200Hz.
This oscillator starts running as soon as proper battery voltage is connected.
The 1200 Hz output at Q9 of IC5 is fed to the clock input CP0 of a HEF4017
Johnson Counter (IC2). Output Q0 goes hight first, followed by Q1, etc.
Only one of the outputs of the Johnson Counter is high at any time – much like
a chasing or sequencing light — stepping at the clock rate.
As soon as Q8 is reached, it will be set, and – via Diode D11 – the Master
Reset pin (MR) is driven. This will reset the Johnson Counter and ensure
that the scanning cycle will be started over and over again. The frequency
at each of the 8 outputs is 1200 Hz / 8 steps = 150 Hz.
The outputs of the HEF4017 (IC2) are wired as a sequencer in such a way
that we get two non-overlapping driving signals. The first signal is created
by combining outputs Q1, Q2 and Q3 to drive MOSFET T3. This simulates the
upper contact. The second signal is created by combining outputs
Q5, Q6 and A7 to drive MOSFET T5. This simulates the lower contact.
Outputs Q0 and Q4 are unused and provide a dead time to ensure that
the outputs from T3 and T5 will never overlap.
Overlapping would cause a short circuit between the upper and lower contacts,
which must be avoided.
We also need the dead time to avoid abrupt changes in the
transformer's magnetic field.
In the next stage, we must isolate the driving electronics
from the actual N-MOSFET switches. This is done by using two small pulse
transformers (TRA1, TRA2), but in order for this to work, the signals
fed to the transformers must be modulated with a relatively high frequency.
This is done by adding two gate drivers (IC3, IC4) to drive the two
transformers. The non-inverting input (pin 3) of each gate driver is fed
with a 76 kHz signal, taken from the Q3 output of IC5.
The oscilloscope image above shows the modulated outputs of IC3 (green)
and IC4 (yellow}. It illustrates that both signals are modulated with
~76 kHz and that there is no overlap between these two modulated signals.
|
|
|
MOSFET transformer driver
|
|
|
IC3 and IC4 are MOSFET Driver IC's. Normally - via a small resistor -
the output of these IC's drive the gate of a MOSFET . But here we
have a trick up our sleeve: we use IC3 and IC4 to modulate DC to AC,
which enables us to isolate and drive the gates of the pair of N-MOSFETs.
The modulated square wave outputs from IC3 and IC4 are strong enough to
drive the windings of both pulse transformers (TRA1, TRA2).
At the secondary side of the transformers is a schottky rectifier that
produces a DC voltage from this transformed energy.
This voltage is applied to the two gates of N-MOSFET pair T2:A
and T2:B. When the square wave is present, this 'contact' switches on.
Note that the gate capacity of each N-MOSFET pair acts as a small
smoothing capacitor. Each N-MOSFET has a Cgs of approximately
1 nF, so two gates give us 2nF smoothing capacitance.
The same is true for the N-MOSFET pair T4:A and T4:B.
In the dead time or during switch off, the energy from the pulse
transformer will be gone. The 47k resistors over the gates (R9, R16)
will discharge the Cgs capacitance in the N-MOSFET pairs
to let them switch off. The zeners (Z1, Z3) are there to protect the
gate voltage at all times.
A snubber network (C13, R11) and a varistor (VAR1) over the N-MOSFET
pair protects it against excessive ringing and avalanche overload.
An identical circuit (R14, C21) is used for the bottom half
of the 'switch'.
Note that due to the non-overlapping sequences, only one N-MOSFET pair
will be on at any given time. In addition, there is enough dead time
between the pulses, to allow each pair to switch off gently, thereby
discharging the gates of the N-MOSFET pair via a 47k resistor.
|
Our proposed circuit can be used to replace very low voltage EMVs.
This is necessary, e.g. when used with some WWII German equipment,
of which the LT was sometimes as low as 2V.
It is not possible however, to drive MOSFETs at this low voltage.
For this reason we've added a step-up converter (IC1) which converts the
low input voltage to 8V DC. For safety, the extra circuit around ZREF1 is
added. It forces a reset of the Johnson Counter as soon as the battery
voltage drops below 2V and the step-up converter is no longer able to
provide a stable 8V output. LED1 lights up if the input voltage is high enough.
This is an indication that the circuit works as expected.
|
To make sure the SSV can be fitted inside the existing enclosure of all
known EMVs, we've divided the circuit over two equally sized PCBs
that we've named (1) Control board and (2) Switch board. Note that the latter
carries the HT voltage which can be potentially dangerous. These two PCBs
will be mounted back-to-back by means of four 2-pin solder headers.
...
|
This image shows the Zerhacker connected to a 240V lamp, to
demonstrate that it can run on just two 1,5V batteries.
At the top right are the two Zerhacker PCBs mounted back-to-back.
At the centre is the transformer. Note the green LED on the PCB, which
indicates that input power is OK.
|
|
|
Indentify the current confguration
|
|
|
Before modifying your EMV, you have to identify the connection method
of the EMV used in your equipment. The possible configurations are
discussed in the chapter 'Configurations'.
In most radio's and transceivers, the (+) terminal is connected to the
centre contact of the primary side of the transformer.
Note hoever that this is not guaranteed. Please check this out before
you start.
For the control electroncs this is not important, as long as the GND
is connected to the (-) terminal of the battery, and the (+) terminal
of the battery is connected to the +Vol1 (yellow) or +Vol2 (red) terminals
of the PCB, depending on the battery voltage.
This enables the oscillator and the drivers for the two N-MOSFET pairs.
In this situation, the green LED will light up.
|
|
Most EMVs can be opened by folding back the sealed aluminium edge
at the lower end of the EMV, just above the contact pins. This can
be done with a screwdriver. Other EMVs can often be opened by removing
a couple of screws and taking off the enclosure. Note that this can be a
delicate job. It is necessary to take your time, so that the enclosure
can be reused for the SSV.
|
Once the enclosure has been taken off, remove the mechanical switches and
the drive coil. Remove the existing gaskets and any rubber parts that you
may find. They are no longer needed and we don't want them to gass-out later.
Next, add a proper isolation to the inside of the empty enclosure, to avoid
short circuits between the PCBs and the enclosure.
Use thick isolation paper or a thin expoxy layer. Such things can easily
be found on local HAM radio flea markets.
|
|
|
Now solder the wires to the correct terminals of the EMV plug.
Note that the pinout of most valve plugs are are specified when viewed
from the bottom . So check and double-check.
It is best to make a drawing before you start.
Switching at the secondary side is achieved with two 400V diodes.
Again, check the configuration to ensure the type of HT circuit
you are dealing with. For a positive voltage, set both slide switches on
the PCB towards T2 and T4. Likewise, set both switches to the opposite
position for a negative voltage. Each switch is a DPDT, that simply
reverses the high-voltage rectifying diode.
Verify that the output voltage has the correct polarity before connecting it
to the equipment.
Test the Zerhacker in your equipment before refitting the enclosure.
Install it in the EMV socket of the equipment and test it thoroughly to
ensure it works as expected. When you are satisfied, refit the enclosure and
re-seal the bottom end of the aluminium can (when applicable).
---
Figure 6a and 6b are examples how to use this Zerhacker-design with old German W Gl 2,4a and MZ 6001 MVs. We hope that by publishing this article that lots of old equipment will work again.
|
|
|
12V Zerhacker — W.GL.12a-b-e
|
|
|
|
The diagram below shows how to convert a German W.GL.12a-b-e
vibrator. Note that the contacts of the vibrator are shown as seen from the
top (i.e. the solder side of the contact pins).
|
- Connect K1 to K2 by means of a shorting bridge.
- Solder a 33Ω resistor from the shorting bridge (K1, K2)
to the +2 terminal (VCL2, 10-15V). Ensure that the legs of the
resistor are properly isolated, so that they do not come in contact
with the rest of the circuitry.
- Put both switches in the lower position.
- Connect the wires as shown in the diagram on the right,
using the table below as a guide:
0 → X Brown/green
0 → - Black
1 → Q Grey/white
3 → A Brown
4 → B Brown
5 → C Grey
6 → D Grey
|
|
|
- Check the configuration of your circuit and place the
switches in the corresponding position.
For configuration 1, put both switches in the lower position:
For configuration 2, put both switches in the upper position:
- Connect the wires as shown in the diagram on the right, using the table
below as a guide:
0 → 
Shield
1 → A Brown
2 → X Brown/green
2 → +1 Yellow
3 → B Brown
4 → C Grey
5 → Q Grey/white
6 → D Grey
7 → - Black
8 not connected
|
|
|
|
|
|
Any links shown in red are currently unavailable.
If you like the information on this website, why not make a donation?
© Crypto Museum. Created: Friday 26 July 2024. Last changed: Monday, 03 November 2025 - 07:51 CET.
|
 |
|
|
|
