PA 23cm 500W 4x MRFE6S 9160

Assembly Tips & Tricks

Philippe Borghini / F5jwf

Last Update: 18.01.2019

Up to date docs to build the PA assembly, last update 18.01.2019: PA_23cm_500W.zip

7.02.2016 S11, S22 measurements.

Figure 1: PA 23cm S11	Figure 1b: S11 Smith chart

Figure 2: PA23cm S22

30.01.2016 Adjustement and test.

Before applying +28V to the drain of the LDMOS pre adjust the gate voltage to around +3.3V. Then +28V can be applied and current adjusted to 1.4A.

For this purpose, measure the current on your power supply or on the wire feed with current probe clamp and adjust current stage by stage. Jumper can be shorted (Figure 1) or not to apply gate voltage and adjust only one LDMOS at time.

I have seen pretty large dispersion in the gate voltage for 1.4A drain current: Vgate: 3.0…3.3V

	Idrain=1.4A Vgate~3.0 … 3.3V
Figure 1

Once the 4 transistors properly biased, low level RF can be injected (typically 100mW ) and measure the gain. In principle, there would be nothing to adjust or retouch and it should peak to about 15 to 18 dB gain. It could possibly be interested to check residual power on the load, using an RF probe and verify that we have between 10 to 15 dB isolation.

When good test results at low level are done then tests at sustained power can be started, typically between 5 and 10W drive. This is the tricky part and obviously there are several precautions to take:

• Ensure that the output load 50 Ohms hold that such power (max 600W)

• Provide cooling accordingly. Efficiency of these devices is only about 40% @1.2GHz. This means that for 1400W (28V 50A) drawn off the power supply, 560W are the RF and 800 are heat that must be dissipated.

• You have to find solution to measure such power. I use a 40dB directional coupler and a 1.2GHz 50Ohms load rated for 1kW.

• Prepare heavy duty (for serious guys) 28V power supply, the saturation power reaches 630W for a current of 65A.

• At the beginning, start with brief transmissions to detect potential problems before the disaster.

• Cautions on your fingers, eyes and other attributes, the RF burns and such high levels are not trivial. For my part, I never stay close to PA when running at full power.

Optimization is the major difficulty of this assembly because the four-way interact with each other and mismatch occur into saturation region. It is therefore necessary to optimize at maximum power. The position of the caps C107 (C207, C307, C407) play on the output power and on the balance of 4 transistors. Playing with position of these caps to optimize the linear region as large as possible. Figure 17 corresponds to the gain curve after optimization. The 1dB compression point is more than 500W with a good linear characteristic. This is the target after optimization.

Figure 2

Figure 3: Automatic Test setup measurement Pout vs Pin

15.01.2016 Build of the first PA with released PCB.

The copper plate has to be drill according to mechanical drawing.

Figure 1: Drilling the copper plate

For accuracy, the PCB is used to locate the holes in backplane plates.

Figure 2: Mechanical parts	Figure 2b: Input flange in brass bended

It is a good practice before soldering the device to check them with DVM

	When device failed I usually seen: Input Vj~0V R~20…50Ohms Output Vj~0.6V R>10MOhms
Figure 3:Internal model of LDMOS and LDMOS Ohmic Test

The LDMOS soldering is the FUN time…Here are the step to follow:

1. Clean up the copper plate carefully with Isopropylic alcool (Now we have in stock tnx JP).
2. Cut the leg of LDMOS to around 3mm of the package to get potential desoldering much easier.
3. Apply a small smear of solder paste. Use only Edsyn CR11 (Sn42Bi58) see BOM for availability. Quantity as shown below (Figure 3) is much enough. Be careful too much paste and short circuits to ground could occur on gate and drain legs. On the other hand not enough paste produces poor thermal conduction. Good practice is placing the LDMOS on the past, pressing firmly, and remove overflow on the edges.
4. You have to find a trick to maintain LDMOS in position and in contact with the copper during soldering. Otherwise they drift above the soldering bath. I use a holding leg with M8 screw (Figure 6), DF9IC use a small weight on each device.
5. Prepare your hot plate. I use regular hot plate from kitchen (tnx Honey) which is regulated with PID temperature regulator with PTC100 inside the item to solder. Temperature can be adjusted to 140°C +/- 10°. You must avoid to heat up above 180°C otherwise you will destroy the parts.
6. Place your copper plate on the hot plate and switch on. Depending of thermal inertia and your setup it takes about 5..7 minutes to be a 140°…150°. Then the paste melts and you have usually nothing to do except switch off heating. If the quantity of paste is fine you should see a small excess going outside the device (Figure 4). If you estimate that you don’t have good quantity of paste you can add or remove now.
7. When the copper plate is cold check with DVM that you don’t have short circuit to ground. If you are not lucky, then heats up again, remove the defective device, clean up and re solder it.

Figure 4: Good quantity of solder paste is important	Figure 5: Device soldered

Figure 6: Solder paste preparation	Figure 7: Hold the devices during the soldering process

The output aluminum plate has to be prepared to hold the 50ohms load.

Figure 8: Output Aliminium plate

All components can mounted on PCB before final mechanical assembly.

Figure 9: Two part PCB equipped with component	Figure 9b: Details of the coils

Input connector, brass edge and PCB backplane can be soldered together (Figure 9) for a perfect RF transition.

It is difficult to do the same on output connector because the aluminium plate. Solution could be to use a small piece of copper sheet (0.1mm thick) bend it around PCB edge (Figure 11) solder it on the backplane with minimum paste to avoid thickness. Then you are able to solder it to the brass edge from top of PCB for a perfect RF transition. Other solution are also possible as 1.2GHz are not so critical.

Figure 10: Input connector details	Figure 11: Input connector details

For the output connector, it is difficult to do the same as input because the aluminium plate are close to edge of brass flange. Solution could be to use a small piece of copper sheet (0.1mm thick) bend it around PCB edge (Figure 12) solder it on the backplane with minimum paste to avoid thickness. Then you are able to solder it to the brass edge from top of PCB (Figure 13) for a perfect RF transition. Other solutions are also possible as 1.2GHz are not so critical.

Figure 12: Output connector backplane detail	Figure 13: Output connector

Before final assembly check every fit well together, then add thermal compound paste on the output aluminum plate (Figure 12)

Figure 14: Thermal compound paste on output aluminium plate	Figure 15: Thermal compound paste on 50Ohms load location

At the end remaining item can be mounted:

· Brasse edge

· Jumper

· +28V bypass

· Output 50ohms load

· Solder the gate and drain of the LDMOS

Figure 16: Final assembly ready for test