Problems with AOR AR8600mk2 receiver (RF switches?)

[borjam]

Hi

I have an odd problem, maybe someone can help.

Some time ago I purchased a used AR8600mk2 receiver. Turns out it was damaged, its sensitivity was really low.

I sent it for repair to a very reputable guy here and indeed when it came back it had improved a lot. He found a damaged RF switch (SPM3204) and a transistor. Seems it was damaged by RF applied to the input (maybe a nearby transmitter, who knows). He also told me that the problem was a real bitch because there was some intermitent problem which he determined to be one of the SPM3204 switches. It uses a lot of them.

When it came back I noticed several surprising things: At least one or two more switches are damaged. If I turn on the receiver it takes more than 30 seconds to be able to receive. And if I activate and disengage the input attenuator it takes the same 30 - 40 seconds. Those switches should be fast.

I also noticed something odd. There was some kind of impedance adaptation problem. It's difficult to describe, but when running on an attached telescopic whip it was way too sensitive to the whip length. Of course I know that changing the length will tune the antenna to a frequency, but the effect was way too sharp.

So I tried to determine if something else was wrong. I attached the receiver to a miniVNA Tiny and begun looking at the RL.

I have observed several curious effects.

First: The first time I did it I did a wideband scan and, surprisingly, the RL curve came flat. This receiver has (as far as I know) a rather strong front end filtering and I expected to see a dip in RL. There wasn't any and the RL was low, maybe in the order of 4 dB. Bad.

Second: After doing several tests it begun to behave differently, like something "woke up". And I begun to see more reasonable RL curves. (More on this later)

The surprising thing was, after the "awakening" it worked way much better for a week or so. The "awakening" effect remained in effect despite the fact that I power off the receiver completely when not in use. So it's not a matter of reaching a certain temperature, it can be powered off for 12 hours in between.

I also noticed something strange. Some of the front end filters are tunable (75-118 MHz, 118-174 MHz, 174-300 MHz, 300-470 MHz) and indeed, each of these filters has a varicap diode. The thing it, the filters seem to be somewhat detuned. I attach some examples.

So, I am wondering. What is the failure mode of these RF switch MMICs? Could it be that they are adding some parallel capacity to the circuit? I guess that would detune the filters. The other scenario, a problem with the control voltage used to tune the filters seems to be more unlikely.

[borjam]

I forgot, block diagram and schematic of the RF part.