Quite some time ago I noticed that my (2019) Kuhne MKU UP2424 upconverter delivers only abt. 5W RF output at 2.4GHz instead of the specified minimum of 20W. Recently discovered as well: there are also a lot of strong spurious signals in the 2.4GHz output!
Unfortunately, these problems stayed unnoticed to me for a long time. As I just found out that Kuhne cannot provide warranty/service (or willing to provide technical information…) anymore for my unit, I try my luck here in the forum!
To make a long story short: The B-upgrade TX converter-module, installed afterwards by Kuhne, has a clean 2.4GHz spectrum and sufficient high RF level. But the RF power stage with (NXP) MW7IC2725N (photo above) has only max. abt. 5W output as measured with a calibrated R&S radio tester, is unstable because it generates several strong spurious signals at a spacing of about 128MHz up- and down the wanted 2.4GHz.
Using a RF pick-up probe on my analyzer i found a fundamental oscillation of abt.128MHz with harmonics that also appear on the MMIC output around 2.4GHz as measured on top of the board without shielding, see the spectrum photos below.
After removal of the shielding cover inside, I noticed some black RF absorbing foam glued to the shielding-underside and sitting above the (input) area of the MMIC RF PA. It seems to me that these are attempts to suppress unwanted signals.
I had a close look at the very informative thread by Achim, DH2VA “MKU UP 2424A No power: repaired” describing a bad soldering issue of a FET switch. This does not seem to be the case in my unit. The +28V is there all the time and this output voltage from the internal DC-DC converter stays stable during transmission.
Changing up/down the level of bias Idq1 and Idq2 for driver and PA inside the MMIC does not help. With PTT only without drive there is no spurious detected with the analyzer. But even with a small bias set the instability occurs with any RF level of drive (144MHz) from an IC705. The spurious signals do appear on the MKU output N-connector, with everything bolted up and closed completely.
Digging on the internet I found several other amplifiers using the same RF PA module.
There are some differences in the various board layouts, perhaps the most interesting difference is that in the Kuhne PA the 28V DC feed Vd2 direct at the RF output runs via a rather short strip-line choke (too short to be ¼ Lambda?) and therefore more or less shorting the RF output to the supply?. See first photo with strip-lines to the right of output MMIC.
This issue was found in an NXP forum about the MW7IC2725N, an MMIC now obsolete by the way…
Moreover, the PCB layout for the tracks carrying the Vg1 and Vg2 voltages are long and seem not optimal as the DC decoupling capacitors are located far away from the input pins. See first photo with PCB tracks near MMIC.
Another indication for lower-than-normal RF power is that the maximum supply current @ 13.8V does not get any higher than abt. 4,2A. The specs state that the nominal current is 7A.
Some questions to other MKU UP 2424 (A- or B-version) users:
- Do other users also observe a much lower RF output and what is the supply current in that case?
- What is the supply current @ 12V …13.8V during max RF output of 20W in a normal functioning unit? Is it really up to 7A?
- Do other users also observe unwanted spurious signals? See the photo above, You just remove the bottom cover and use a simple RF pick-up probe on a 3GHz analyzer to have a safe and quick check. Good to know - No warranty seal has to be broken to check this!
- In case you opened the shielding inside; did you see a black foam pad inside the metal shielding as is in my unit?
I also would like to hear experiences of people having built 20W 2.4GHz amplifiers using MW7IC2725N regarding unstable operation (oscillations), low RF output power issues or any other problems!
Looking forward to hearing from anyone how to tackle this problem…many thanks!
Reaktionen in Deutsch Sprache sind auch wilkommen, vielen Dank!
73
Ed PA1EW