Tom gives a technical summary of the Toshiba 3.4 GHz linear amplifier and then an explanation of how to hook it up, test it and get it on the air.
The Toshiba UM2683A 40W 3.4 to 3.7 GHz solid state linear amplifier is occasionally available as commercial surplus and on Ebay. It is a good value for the amatur who need moderate power on 3456 MHz.
|RF Range||3.44 - 3.68 GHz|
|Output Power||40 - 60 W Max (depends on device gain variations)|
|Input Power||0 dBm to 20 dBm with internal attenuator adjustment|
|Linear Gain||+42.5 dB|
|Gain Ripple||0.1 dB|
|Return Loss||25 dB in/out|
|Supply||12.3 VDC @ 15 A typical; use a 20A fuse|
|Size||5 1/8 x 8 x 15/16 inch (129 x 202 x 24 mm)|
|Weight||2 lbs, approx|
The three panel connectors are as follows:
|Output||SMA (f), right angle|
|DC||DB-15 (3 row, like VDS)|
Finally, the DB-15 D connector pin assignments are as follows:
|4||High Output Alarm; 1 to 38 dBm; 0 from 42 dBm|
|5||SWR Alarm; 1 > 12 dB RL, 0 < 6 dB RL|
|8||Low Output Alarm; 0 for Pout < 6 dBm|
|9||Bias Enable, Low/Gnd = ON (TTL)|
|15||+ LED; R820 ohm, 5 mA|
This unit requires a substantial heatsink, particularly if used for continuous duty. It is class A, so it will be drawing all its rated power immediately. You will need to dissipate about 150W! For SSB/CW use, Pin 9 provides a simple way to lower average power; just ground it for xmt on your PTT line. Max Temp 85 deg C.
Input device is a TMD0305-2 MMIC rated at 2 Watts output for 22 dB gain. Followed by TIM 3742-4L and TIM 3742-8SL Power FETs; final output device is TIM3536-60. Input and Output Circulators offer significant RF mismatch protection for the final.
Internal input attenuation is comprised of the two pots (labeled R217 & R210); these are the far two on the left with the RF deck to the top. Max gain (min attenuation) is full clockwise for 0 dBm input to full rated output of 40/50W with no power fold-back. Max input attenuation is approx 20dB for +20 dBm (estimated) input. The next pot to the right (R138) apparently sets the output stagesą bias; I set it for just under full output, which was 15A for my unit with 20 mW drive.
The internal regulator (NEC D1297) will allow limited duration/keyed SSB usage at 13.8 VDC, but it gets warm. If you prefer to use a 13.8VDC for permanent use, I suggest you put two or three high current silicon diodes in series with the supply to drop the voltage to the nominal 12.5 - 12.8 VDC, and minimize internal heating. If this regulator fails, it will likely take out the final RF amplifier stage at the same time!
The Toshiba 20W and 40 W amplifiers use a female high density 15-pin connector which is available from RadioShack:
Connector, High density, 15 position, female, RadioShack 276-1501 $1.99 ea.
Hood, 9 position (fits above connector) Metallized, Radio Shack 276-1513 $1.99 ea.
The solder pots on the above connector accept 22 AWG wire, and 4 22 AWG wires each for power input and for the power return wires will handle a bit better than 15 A comfortably. The 22 AWG alarm wire that RadioShack sells was tried first, but the insulation is too thick to fit the solder pot spacing of this connector.
Teflon-insulated 22 AWG wire, Belden Electronic Division, 83005_010 (BLK) has silver plated copper conductors with very thin TFE insulation rated for 200 degrees C and 500 V service. The diameter of this wire across the insulation is the same diameter as the solder pots in the connector.
All pins were carefully wired working from the center pins outward. Half-inch long heat shrink tubing of suitable diameter was slipped over each pin. The power connector pigtails were each 6 inches long while the remaining signal lines were 48 inches long. The entire wire bundle fits comfortably inside the hood.
Once the actual wire lengths are determined, the eight short wires going to power can be soldered to 12 AWG zip cable like that from Powerwerx, and the signal lines can be connected to the appropriate control circuits. Note that two of the control lines are not internally connected within the Toshiba amplifier and are available for bringing a couple of extra monitoring signals out.
This wiring method appears to be preferable to other wiring methods that have been recommended in that no control or monitoring functions are lost, and the Toshiba Amplifier itself is not modified.
You will need the following items:
With no input RF drive yet, connect 12 VDC to all eight of the +V and Ground leads from the DB-15 connector; current drain should be << 1A. Ground Pin 9; current will jump to 8 - 10A. Adjust R138 to about 13-14 A. >From here on, ground Pin 9 only when necessary, unless you are sure you have an adequate heatsink.
Apply 1 mW (0dBm) of RF input power; you should see a few watts output. Adjust the second stage pot (R210) clockwise to max RF output - then back off slightly. Next adjust the first stage pot (R217) to max RF output level, and back off slightly; max is 40 to 55 watts and seems to depend on temperature. If you get no output, try wiggling these pots a bit; R217 was dirty for me and I had to cycle it some for a reliable setting. The wide range I indicate for power is reflective of my current uncertainty of power measurement accuracy.
Finally, go back to the output gain/bias control R138 and tweak it for max RF output, then back off slightly for best efficiency and lowest total current. Mine seems to idle at around 13A, and rise to 14 or 15A with RF drive to max output. I use a -30 dB directional coupler and diode detector, provides 500 mV at full RF output.
Operating temp is 85° C max; if temp exceeds 90°, then auto shutdown with low power alarm to 0 (pin 8). Alarm resets after approx 10° drop.
Power Supply Voltage: This unit should be operated at 12.6VDC if possible to reduce heating of the internal power regulator. However, PTT operation at 13.8VDC is OK for normal, short time SSB usage if this amp has a hefty heat sink.