Using the California Microwave 11-026700 transmitter assembly

Dave Robinson WW2R, G4FRE

Introduction

A considerable number of the California Microwave "Transmitter assembly 11-026700-08(h)" have been seen at hamfests over the past 5 years. Many people have bought them, but it is hard to believe this, judging by the scarcity of "on-air" models. This paper details how to get them going on 3456MHz along with details of one method of incorporating it into a transverter and will hopefully encourage some to dust them off and help populate the band.

Equipment Description

In commercial use by AT&T for 3.7 - 4.2GHz microwave links, the amps were powered from positive ground 24Volts via a rack mounted California Microwave "Power supply 52-090095-0". The Power Amplifier unit consists of a 12.5 x 5.5 x 1.5 " aluminum can, containing the electronics, attached to a 13 x 6.75 x 1.5 " heatsink. The input connector is SMA, the output is via waveguide, but this is easily converted to SMA. A 6 core cable terminating in a 6 pin plastic plug supplies the operating voltages from the PSU.

The Power Supply unit consists of a 19 x 6.5 x2.5" box with one 2 pin connector, for applying -24V and a 6 pin connector for outputting the regulated voltages to the PA. It provides outputs of +10.25 and -5.3 Volts. It has an on/off switch and 4 test points.

The observant will have noticed the PSU is bigger than the PA, in addition it's -24 volt requirement makes using it portable a problem. The first problem to be tackled therefore is to build a PSU to build the amp off 12V. In its original state the amplifier's performance is poor at 3456MHz and will need retuning.

Power Supply Unit

If the 6 screws securing the aluminum can of the PA are removed, along with the 6 smaller screws securing the aluminum screen below, it will be seen that the 6 core cable is connected as follows:
 
 
Orange and Purple Drain Supply
Brown and Blue Gate Bias
grey Ground
red rectified RF Sample

The orange and purple wires supply +10.25V at 1.7 and 1.5A respectively from isolated sources in the PSU. The 10.25V requirement precludes the use of the 78H series of regulators when running the PA off a lead acid battery as they require a minimum of a 2.5 Volt drop across them. A better idea is to use one of the LT108X series of low dropout voltage regulators which typically require 0.5 to 1V across them to maintain regulation. A single LT1084 regulator, which has a short circuit current of 5A is used. These regulators, along with a whole range of linear technology Inc products are available from Digi-Key (www.digikey.com)

The brown and blue wires supplying the -5.3V need around 10mA with the 10.25V supply on , 20mA with no 10.25V applied. This can be supplied by a LT1044 IC. The DC-DC converter chip is supplied from a 7808 regulator. The more usual ICL7660 is not used as it has been found more prone to expiring. The output of the LT1044 feeds an LM337T adjustable negative voltage regulator set to give -5.3V. The resistors around the LM337 have been increased, without adverse effect, from the usual recommended values, to reduce the current loading on the LT1044 IC.

It was decided to include extra circuitry to protect the amplifier from the loss of the negative rail, which has been known to destroy the amplifier. The circuit used COMPLETELY removes the positive 10.25V supply in the absence of the gate bias, rather than the more often published design such as (1) and (2), which only shuts down the regulator to 1.2V output.

The complete circuit diagram is shown in Fig 1. A PCB was designed for the circuit, excluding the components for the 10.25V regulators which are mounted directly on the heatsinks, and the power relay. The PCB layout Fig 2 and the component overlay in Fig 3. The component listing is shown in table 1.

Correct operation of the PSU should be ensured before proceeding. Normally the relay should operate and +10.25V and -5.3V should be measured at the appropriate points. If the -5.3V rail is removed (for example by disconnecting pin 8 of IC2) the relay should drop out and +10.5V rail should disappear.

Power Amplifier

Firstly the output connector has to be changed, to allow the usual RF connectors to be used on the amplifier. Carefully unsolder the brass pin from the output track of the amplifier. With a hex wrench, undo the 6 set screws holding the waveguide assembly in place and the four screws holding the tube to the edge of the PCB. It is possible, with a careful sawcut to leave part of the waveguide assembly with the 5 feedthroughs and the earth lug and 4 mounting holes. For the faint hearted the safer option is to unsolder the 6 wires connected to the bracket and remove the whole assembly. This is replaced with a short length of 0.5" aluminum angle fitted with 4 bolt in feedthroughs and a ground lug. (The brown and purple wires are connected to a common feedthrough). This method also shrinks the amplifiers area, allowing a changeover relay to be connected directly to the output connector.

The mounting plate left on the edge of the PCB has the correct fixing centres for a 4 hole SMA chassis mount female socket of the type with the long centre pin and the extended ptfe insulation. The centre pin and insulation are cut to length. The two holes for the lower screws are tapped and four long screws and 2 nuts hold the connector in place.

As a guide to the correct operation of the onboard PA regulator board measure the gate and Drain voltages of the devices:-
 
Device
Vgs
Vds
TR1
-1.1V
7.3V
TR2
-1.1V
8.1V
TR3
-1.2V
9.8V
TR4/5
-1.5V
9.8V

These voltages, are approximate and vary slightly between amplifiers, but give good indications of potential problems, and are measured with a 50 ohm load connected to the input and output of the amplifier to avoid instability corrupting the results.

Connect 10W power meter rated at least to 4GHz to the PA output. If a 3.9GHz source is available (such as a Midwest microwave "brick" awaiting conversion to 3456) lower the output to 1mW using attenuators as required (beware, some models produce over 250mW!) and apply to the PA input. The PA output should be at least 7.5W

The next step is to apply a 1mW of 3456MHz to the amplifier. The output could be as low as 3W, so retuning is necessary. It will be observed that the board has extra printed stubs on the board, mostly on the inputs and outputs of the devices. These should be connected and disconnected with solder "blobs" AFTER DISCONNECTING THE POWER SUPPLY BEFORE EACH ADJUSTMENT to maximize the output power working in sequence TR1-5, optimizing the gate matching first then moving onto the Drain. If necessary extra pieces of copper foil, available from hobby shops may be added to the board. A recently discovered alternative to soldering the foil is to use adhesive backed copper foil which is available in various widths from stained glass craft shops. This was discovered to be the vital ingredient in my wife, Meg's N2NQI's newly mastered hobby and has meant many trips to supervise the correct widths being purchased. When properly tuned the amp should give at least 7W, it has been consistently found to give slightly less output at the lower frequency, despite repeated optimization efforts. Finally the inner cover should be replaced over the circuit boards, this has RF absorbing foam under it to maintain amplifier stability. The outer "can" was discarded.

The so far unused red wire drives the negative terminal of a 10mA meter via a series resistor to ground to indicate relative output.

Transverter Integration

Having built the amplifier it needs integrating into a transverter. I originally used the no-tune design board with the 540MHz oscillator (3,4) but transverter was large, even without a PA. A new system was built, designed around a transverter by DB6NT published in (5). The complete equipment layout is shown in Fig 4. The transverter module contains a TX converter (200mW output), RX converter (1.4dB nf/26 gain) and a 144MHz interface (for the FT290 originally, but easily modified for an IC202) contained in a 6 x 3 x 1.2" box. The only modification I needed from the original design was to solder rigidizing lengths of 12AWG copper wire across the ground plane of the board, otherwise it was found that the PTFE board flexed, detuning the transverter module. The Local Oscillator is separate; continuing the trend of miniaturization the 3312MHz LO employed was the design by G4DDK (6) which generates 10mW in a box 6x1.5x1.2". The Preamp is the single stage WB5LUA design bought as a kit from Downeast microwave (7) using an  ATF36077 device  yielding a noise figure of 0.45dB system noise figure (excluding antenna changeover relay).

The 20dB attenuator was put on the input to the PA to avoid the risk of overdriving. The PA input filter was one of the "Collins" copper filters obtained at the 1991 Microwave Update in Arlington; this ensures an ultraclean output spectrum as shown in Fig. 5. The 24V for the relay is generated by the NE555/complimentary pair arrangement published in (8), this is totally enclosed in a 2x2x1" screened box with feedthroughs for input and output to stop power supply noise.

The complete transverter is housed in and on a 19x7x2" aluminium chassis. A hole 1" narrower and shorter than the PA heatsink is made in the base and the PA assembly is mounted , by holes tapped in the heatsink, through the hole, heatsink on the top.

Conclusions

Hopefully this article will encourage more activity on 3456 over longer distances. Remember the old adage of the ham bands "Use them or Lose them"

References

1. Power Amplifier for 13cm. E.Gobel, DUBUS 2/94 pp22-29

2. Dual power supplies for GaAsFets. B. Troetschel, Proceedings of 37th West Coast VHF/UHF Conference pp41-51

3. A no-tune Transverter for 3456MHz. J.Davey, Proceedings of Central States VHF Society 1987 pp 51-57 (also QST Jun 89 pp21-26)

4. A Clean, low-cost Microwave local oscillator, R.Campbell QST Jul 89 pp15-21

5. Transverter for 3.4GHz. M Kuhne, DUBUS 4/91 p23-33

6. A high quality source for the 2.8-3.5GHz range- G4DDK009, S.Jewell, RSGB Microwave Newsletter March 94.

7. www.downeastmicrowave.com

8. An IC202 interface circuit. D Robinson. Proceedings of Microwave Update 1992, pp95-98
 
 

Figure 1. CIRCUIT DIAGRAM

Figure 2. PCB TRACK LAYOUT

Figure 3. Component Overlay

Figure 4. Integration of Amplifier into Transverter

Figure 5. Output spectrum of Complete Transverter

Table 1 Component Listing



 
 
 
 
 

Component Value
C1 0.1uF 25V TANT 
C2 1uF 16V TANT 
C3,C4  22uF 10V TANT
C5  0.1uF Ceramic
C6 10uF 16V TANT
C7 22uF 16V TANT
D1 1N4001
IC1 7808
IC2 LTC1044
IC3 LM337T
IC4 LT1084CT
R1,R6 470
R2,R5,R7 1k2
R3 4k7
R4 3k3
R8 867(820+47 in series)
R9 120
TR1 2N3904 or similar
TR2 BD132 or similar
RL1  DPDT 12V coil 5A contacts