23 May 2016

Replacing an output amplifier chip in RigExpert AA-200 antenna analyzer

Hi all,

after some unsuccessful trials of repairing the AA-500 antenna analyzer from RigExpert by my good friend he decided to ask me to help him with that job last month.

As the local  (my affiliated) radio-club's analyzer AA-200 has been also damaged some time ago and I was asked to repair it and as I had to replace the same output amplifier in my own AA-200 analyzer 6 months ago I decided to make short guide and share it over here in order to help to other RigExpert analyzer users...

This article describes how to replace the Sirenza SGA-6589 output amplifier chip if the analyzer stopped working suddenly. The first symptoms are when the SWR is near 1 on all frequencies with any load. Also the flat frequency response with high noise thru wide range with no resonances or peaks during antennas tests. The Meter test mode does not show filled bars...

If the Sirenza amplifier is damaged then there are no scale deflections on both tests even does not matter if you have the test port loaded or keep it opened.

At first you need to prepare the ESD protected bench with grounded wristband (thru 1.5MOhms resistor) and ESD mat.

The best solution is to use two layers, oil, grease and solvent resistant, high temp resistant, acid proof and knife cut proof material with excellent wear hardness. These ESD rubber mats have two layers: the top one is a static dissipative rubber layer laminated to a black conductive rubber bottom layer. Surface resistance of the top layer is 10^6 – 10^8 Ohms and bottom layer is 10^5 – 10^6 Ohms.
These ESD rubber material offers excellent resistance to oil, grease and most common solvent:


After that you can start to push the lock on the back of the tester. Pull-off the battery pack and remove the 6pcs of screws from the back cover of the analyzer.

After the removing of all 6pcs of the screws you need to separate the covers very carefully because you need to remove the flat ribbon keyboard connector cable first before the removing the front cover completely.

When you are finished and the front keyboard cover is removed then you can start to unscrewing the screws to remove the CPU board.

Be careful as you need to remove the grey bottom panel (on the right side of the picture) and remove the CPU board at the same time with disconnecting the another connector on the bottom of CPU board.

It is a bit tricky but definitely not difficult. You need only to be careful and watch what you are doing... Do not use too much force just slowly and carefully pull-up the board.

Note: The CPU board is long and held by its connector in the middle of the board together with RF detector unit. Do not pull-up the CPU board on the sides BUT in the center of the CPU board close to the connector. You will avoid the stressing the board and bending it.

This is what you will see after removing the CPU board out. Remove the isolating foam and start to removing the 4 screws on the RF detector board.

After removing the 4 screws you are ready to separate the RF detector board from the back cover finally. You should pull it out a bit to the right side and then up in order to safely removing the board with its metal panel (supporting the PL-259 test port connector) from the plastic back cover frame.

Rotate the RF detector board and you can find the Sirenza SGA-6589 to be replaced.

Unfortunately in case of the unit from our club I found (except the messy amplifier place) also a bit strange damage close to dead amplifier. As you can see there is also unstuck track to L6 inductor which makes the band pass filter.

The assembly microscope Mantis Compact from Vision Engineering will help a lot in that case to repair it...

I am going to remove L6 damaged inductor and repair the lost track. It will be cleaned then refurbished the track and inductor will be replaced by Matsushita wire wounded ceramic inductor with 47nH.
The damaged L6 inductor removed and the track is back on its place. Now the residual solder need to be removed and place need to be carefully cleaned from the flux etc.
The L6 removed and to be replaced with Matsushita Corp. inductor 47nH.
After that we can revert back to the SGA-6589 amplifier itself...

The professional de-soldering SMD UNIT 60A from ERSA will be used in my case.

It is very important picture. Please give attention to the amplifier package pads/pins. The pin in the middle is soldered on the both sides! You need to heat up the both sides > in that case all 4 pads together at the same time in order to safely and quickly remove the chip from the PCB. Only in that case you will not overheat the PCB!

Note: If you have not the professional tools or at least the de-soldering unit or heating unit (hot-air de-soldering station) then ask someone around you to help you. Do NOT try to heat up the pins step by step. You will not remove the chip and you will be in risk to overheating the PCB! Avoid the strange mess around the chip and damaging your PCB in analyzer.

After removing the Sirenza chip the place need to be cleaned also from the residual solder and fluxes.

It looks better isn't it? :)

Now we are going to soldering new Sirenza SGA-6589 to the PCB and in my case also the L6 inductor as part of the band pass filter replacement.

New ceramic wire wounded inductor found. After checking it on the Agilent Precision E4980A RLC Meter it can be soldered on the PCB as same as the Sirenza SGA-6589 amplifier.

If you are not experienced with the SMD soldering then put small piece of iron to the single pad. Using the tweezer or much better sticky pen:


... hold the chip on its place and heat up the pad with piece of iron.

You have chip safely on the right place (you can heat quickly again in order to move a bit if needed) and then you can peacefully soldering all other pins/pads with small amount of iron and heat up it properly but quickly.

Do not use a lot of iron. It is not necessary and you have better control about the proper soldering and heating thru the pads/pins.

After finishing the soldering of Sirenza amplifier and cleaning the flux around if needed you can start to assembly the analyzer in reverse procedure step by step.

If you are done you should not have any kind of screws on the bench. :)

The final test of the analyzer should express the correct values for the no-load (test port open) and loaded (test port loaded with 50Ohms non-inductive resistance) cases.
The open test port test looks like that...

Another information and details you can find in short videos on Youtube (about year ago posted) with the replacement procedure here:

AA-200 Sirenza replacement - part 1

AA-200 Sirenza replacement - part 2

73 - Petr, OK1RP

12 May 2016

Alusid dummy-load 50Ohms/1-30MHz/500W in lab

Hi all,

as you know the directional power meters like NAS from R&S and many others need to be connected to proper load for their correct measurement. If you have the Insertion units (Directional coupler) loaded incorrectly (different impedance from the nominal system impedance) then the detector + following RF voltmeter will be showing incorrect values.

As I would like to arrange as precise as possible RF power measurement setup in my lab using the calibrated (NIST) NAS Directional Power Meter NAS from R&S, see over here...


so I need to use also precise and characterized dummy load for that setup.

The first candidate in the lab was an Alusid 50Ohms/1-30MHz/500W dummy-load from S. Hari, DK9FN which I used for rough checks of the different trasmitters for several years in shack.


There are the results...

I used ENA Analyzer E5061B from Agilent Technologies with full range NIST calibrations. The calibrations are valid until 18.12.2016 and according to the calibration certificate all the values are SWT (distinctly in tolerance <=20% of tol.) with no Fails.

Equipment: ENA E5061B, Port 1, 50Ohms system impedance, 50Ohms jumper RL=>35dB, 1pc of N-PL adapter used, DUT Alusid dummy-load Hari 50Ohms/500W.

Remarks: This first measurement was done with NO coaxial jumper + N-PL adapter compensation.

Ambient temp: (23.0 +/-5)deg C
Relative hum: (45 +/-25)%
All certificates for national etalons are available.

The SWR measurement done for 1 - 30MHz range.

The Z measurement done for 1 - 30MHz range in Smith diagram.

The Z measurement in R +/- jX (Ohms) units for 1 - 30MHz range.

The Return loss in Log mag for 1 - 30MHz range done.

Unfortunately this dummy-load seems to be usable for lab RF power measurement up to 7MHz only... The RL is bad for frequencies above the 10MHz and even useless for 24/28MHz ham radio band.

There you can see complete test setup on the bench in lab.

After the characterization of this dummy-load I decided to remove it from the test equipment setup (it will be used just for TX check or not measuring applications up to 20MHz/500W out) and I have to find better one...

Also I will repeat again the same measurement with coaxial jumper + N-PL adapter full compensation (there is the ENA procedure to be used) but as the RL of the jumper itself is high enough I do not believe that the results will be much different.

The uncertainties were not calculated for these values yet.

73 - Petr, OK1RP