Information On The
CTE TR45 Transverter

Photo of the front panel of the CTE TR45

Introduction To The CTE TR45 Transverter

Welcome to my page on the CTE TR45 transverter.
I felt the need to put some information on the internet regarding this unusual piece of equipment, as I was recently trying to research it and was surprised how little mention there is of this device, apart from other people asking for information.

Please be aware that it is illegal to use this equipment with the factory settings; The information on this page is for those with an interest in radio and electronics, and is for educational purposes only - I can not condone the use of transmitting equipment like this on any frequencies other than those assigned to amateur radio, and then only by licensed operators.

The TR45 is a transverter that was marketed during the 1980s by CTE International, although the circuit board bears the name 'electronic systems', presumably the company responsible for designing the circuit. A transverter takes one frequency and converts this to another, doing so for both transmit and receive. This particular unit is designed to accept the output of a typical CB radio and convert that to an output within the 45 metre band. It also receives signals on the 45 metre band and converts these to a frequency the CB radio can accept.

Photo of the rear panel of the CTE TR45
Rear panel of the CTE TR45 showing the warning about misuse

The 45 metre band, also known as 6.6MHz or 'Echo Charlie', is allocated to en route aircraft communications which take place on the upper sideband, with channels spaced at 3KHz intervals, along with a few Volmet transmissions. This band has also historically been used by illegal operators and suggestions are that these people are often licensed radio amateurs who have become dissillusioned with the tight restraints on the amateur bands, and what they perceive as a rather silly radio ham culture which has grown up such as reporting every signal as '5 by 9' regardless, the exclusive persuit of QSL cards and the snobby attitude of a few towards newcomers. Transmitting outside the constraints of the license allows for a more interesting conversation, and as a long-term SWL on the amateur bands I can fully understand the attraction.

Operation of the CTE TR45

The unit I had here to study was fitted with a 20.6920MHz crystal, and to describe its operation in the simplest terms, the unit works by taking the input of the CB radio and subtracting the frequency of the crystal from it. What remains is amplified as the transmitted signal. In theory this would mean that by inputting a signal at 27.355 (channel 35) you should get 6.663MHz out. This isn't the case in practice although it is quite close, and this discrepancy is due to both stray and deliberate capacitance around the crystal which slightly reduces its frequency of operation, and indeed there is a control provided to vary the frequency a little. In practice I found that an input of 27.355 produced an output more or less on 6.670MHz, the calling frequency of the 6.6MHz band. By going up one channel on the CB, the frequency is increased by 10KHz, as was the TR45's frequency, so knowing the one channel allows you to calculate the rest.

The transmit bandwidth of the TR45 is fairly narrow and only just covers the 45 metre band with output falling off very rapidly at the band edges, so the CB input does have to be within the right range of channels to get any output - typically between channels 22 and 39, or 27.225MHz and 27.395MHz. It should be noted that the receive capability is fairly wideband and not subject to the narrow bandwidth of the transmit stages - this means that activity on the 40m amateur band can be monitored using this unit, as long as the CB radio has SSB capability and features the 'high' channels, as the amateur bands require the CB radio to be set to receive on 27.685 and higher, this frequency corresponding to the lowest part of the 7MHz band - 7.000MHz.

Front panel controls include a simple on/off switch, which when in the off or 'bypass' position allows direct connection through to the antenna. There is also an RF gain control, but I found this to be fairly useless using my resonant vertical antenna, as during the evenings especially the large signals received from the poweful broadcast stations above the 40m band caused the RF amplifier to overload and develop intermodulation products, therefore the only useable setting was being turned anti-clockwise for minimum gain. Presumably with a mobile or other inefficient antenna the RF gain would be a more useful feature. The remaining control is a frequency shift control, enabling the frequency to be varied a few KHz.

Inside The CTE TR45 Unit

The photo below shows the interior of the unit, with the wideband coupled output stage and its two transistors mounted on a reasonable heatsink. I must assume the output stage is designed like this to allow operation on other frequencies besides the 45m band, and indeed the adustable RF transformer, T8, allows tuning to suit other frequencies. Output power is similar to the CB input, so for SSB you could expect around 12 watts PEP.

Photo of the inside of the CTE TR45

The other tweakable controls are RV1 and RV3.
RV1 trims the amount of CB input presented to the circuitry, which is taken across a large 52 ohm resistor acting as a dummy load for the CB unit. This means the unit presents an almost perfect SWR to the CB. Although this control can decrease or increase the output power, there must be a reason for its setting, so best leave it alone unless you are using a CB with a very low power output. RV3 is labelled 'MIX LEVEL' and therefore it is best not to adjust this, as there is a risk of upsetting the operation of the circuit.

Modifications To The CTE TR45

Improving Stability

After messing with this unit for a while it quickly became apparent that all was not well with the design - it suffered terribly from frequency drifting, especially during transmitting sessions. A bit of investigation with the cover off revealed that the component responsible was the varactor or varicap - basically a diode acting as a small capacitor, whose capacitance can be varied over a small range by applying a reverse bias to the diode, used for pulling down the frequency of a crystal slightly.

All diodes exhibit this effect to some degree, but varactors are designed to maximise this capacitive effect. It is also a recognised characteristic that the capacitance changes slightly with temperature. The problem in this case was so severe that I could make the unit drift by 500Hz merely by breathing onto the area of the board where the varactor was situated, so something had to be done. Below is a diagram of the circuitry designed to provide the variable bias to the varactor in order to allow the frequency to be shifted.

Diagram of the unmodified circuit
Diagram of the varactor bias circuitry before modification

As can be seen, the regulated 9 volt supply is passed to a 47K potentiometer via a 100K resistor. This configuration means that the maximum available reverse bias to the varactor would be about 3 volts, and the minimum would be zero. Now, on the understanding that the temperature stability of a varactor is at its worst with very low values of bias and therefore highest capacitance, it seemed a good idea to change this section. I bypassed the series 100K resistor with a 1.2K resistor as I happened to have one to hand - this now enabled the voltage to swing almost to the full 9 volts. There was still the problem of the minimum voltage being zero, also the voltage range was too great so I cut the board around the earthed end of the potentiometer to isolate it, then soldered in a 47K resistor to bridge the isolated region. This effectively lifted the lower end of the potentiometer from earth potential, also reducing the range of voltages available. Extra decoupling capacitors were added to the top side of the pot and to its output - they may not be strictly necessary but seemed a worthwhile enhancement as there seemed to be the equivalent of potentiometer crackle which the capacitors damped out nicely.

Diagram showing modifications
Diagram showing modifications, including the small signal diode in place of the varactor

I'll make a note here that the varactor did in fact appear to be connected the wrong way round on the board, with the bias voltage not reverse biasing it as should be the case, but forward biasing it. The circuit diagram provided with the unit also shows it connected forward-biased. Surprisingly enough, soldering it in in the conventional manner made little difference!

With the new modified bias supply created I tested the unit again. The rate of drift was greatly reduced but still a problem, so I decided to simply do away with the varactor altogether and unsoldered it from the board in disgust. In its place I soldered a very low value variable capacitor to enable the unit to be pulled onto frequency, and tested it again.

This time the frequency was very stable and breathing onto the crystal region of the board produced little change in frequency. However, I found that I missed the ability to vary the frequency, as the shift on the CB rig was a bit coarse making it difficult to properly resolve SSB transmissions. I decided to try an alternative approach and soldered a small signal diode in reverse connection across the variable capacitor. My theory was that the capacitance change of the small diode would be very small compared to a varactor and any temperature instability would be correspondingly small. The picture below shows the actual components soldered to the underside of the board, and the region of the board around the leg of the potentiometer which had to be isolated. A bit untidy, but easy to modify and experiment with.

Photo of the stability modification I performed on the TR45
Actual reverse side of the board showing the rather untidily added components

Finally, a correctly working unit!
The frequency shift now operated over a satisfactorily small range - about 700Hz in total at a rough estimate, meaning I was able to use it as a very effective fine tune for voice clarification. More importantly however, the temperature stability problem seemed to have been all but completely resolved. The variable capacitor is still needed to pull the crystal onto a convenient frequency and as it now represents the bulk of the capacitance in circuit the effects of the small diode are flooded by it.

In conclusion, there is still a small amount of drift as the unit heats up during transmitting but this is relatively small and can be easily compensated for between overs by a slight tweak of the frequency shift. This modification has at least now made the unit usable.


Overriding The Automatic RF Switching

One of my pet hates with add-on equipment like this, linear amplifiers included, is the automatic RF switching facility which makes the unit switch over to transmit when it senses RF energy at its input. Almost always there is a delay before these units switch back once the transmit signal has ceased which I find annoying, and also if there is a short pause in the speech whilst transmitting in SSB mode the unit will often switch back causing irritating relay dropouts.

Although this automatic RF sensing is necessary if the unit has been attached temporarily, if it is to be used in the same setup for any length of time I like to disable this feature, using the microphone switching to directly control the auxilliary equipment. This can easily be acheived by fitting a relay in the transceiver in such a way that it activates when the microphone is used and taking a pair of the relay's normally open contacts to a socket on the rear panel - the details of this modification will vary from rig to rig, but the principle is the same. This modification to the TR45 assumes you are prepared to do this to your rig - there is no point in doing this mod otherwise.

The diagram below shows the part of the circuit within the TR45 responsible for sensing RF at the input, and energising the changeover relays.

Circuit of the RF sensing section of the TR45
Diagram of the Rf sensing and switching part of the circuitry

The component numbers are as they appear printed onto the top side of the circuit board, for ease of identification. Q6 is the transistor nearest the two long enammelled wire coils, and the other components are situated nearby. C24 is a 22 microfarad capacitor responsible for introducing the slight delay in the unit switching back once the input signal has ceased - important for SSB use - and this modification centres around removing this component from the board.

NOTE:
The same could be acheived by removing R20 from circuit, but this would render all automatic switching inoperative. By removing C24 at least the unit can still be switched by the presence of RF f need be, but without the delay or 'hang'. In an ideal world a switch would be fitted to give the option of either having the capacitor in or out of circuit rather than removing it completely - some linear amplifiers have this facility in the form of an 'AM/SSB' switch.

A socket of some sort, perhaps a phono socket or an earphone jack, needs to be fitted to the rear of the unit to enable connection to be made to the modified transceiver's relay contacts. A connection is then made from the collector of Q6 (marked on the board) to the centre connector of the socket, and from the other connection of the socket to ground, which is any part of the large area of conductor running around the perimeter of the board and to which the outer of the antenna sockets are connected, or the metal frame of the unit as this is not isolated.

Once this is done, when the PTT relay you have fitted within your transceiver closes, this will ground the relays within the TR45 causing it to immediately switch regardless whether there is actually RF present or not. The unit will stay activated for as long as the microphone is operated.


Valid HTML 4.01!