Multicoupler Problem

[Squirtle]

Hello guys,

I have a question for you. In below image, i draw a block diagram of a communication system we use for our costumers. In first block diagram signals in air passes through duplexer which works at UHF band then goes to multicoupler and lastly to relays.

At output of the duplexer we measure -125 dBm power, at the output of multicoupler we measure about -90 dBm power which is not possible because the gain of our multicoupler is 5 dB at each port. Multicoupler consists of PSA4-5043+ amplifier circuit and 4 way power divider. There is a problem with our multicoupler because when we replace multicoupler with an amplfier and splitter (block diagram shown in below image), the system just works fine and we can communicate in any place where we could not do with our multicoupler.

So, it is very obvious that problem is in our multicoupler but i could not figure out what the problem might be. We do not use filter in front of multicoupler but there is a duplexer and LNA cannot pick up all signals in air because duplexer attenuates them. I wonder how signal power jumps from -125dBm to -90 dBm. What does it mean?

 

[volker@muehlhaus]

At output of the duplexer we measure -125 dBm power, at the output of multicoupler we measure about -90 dBm power which is not possible because the gain of our multicoupler is 5 dB at each port.

How did you measure the -125dBm? That is really low level, what was the exact procedure for measurement?

 

[Squirtle]

-125 dBm is what the relay measures when it is directly connected to the duplexer's receiver port, without any amplifier or multicoupler.

 

[volker@muehlhaus]

what the relay measures

Ok, I was thinking you measured using a realiable instrument like a spectrum analyzer.

So we don't really know how accurate this is. Maybe it is not accurate, and maybe the -90dBm measurement includes a lot of wideband noise or intermodulation products created in the amplifier. You mentioned the duplexer, but that doesn't means there is only your weak input signal at the amplifier input.

 

[Squirtle]

The amplifier that i have shown below must also cause intermodulation products if this is the case, right? However it works fine. The output of that amplifier shows what it is supposed to show. Gain is as it is indicated in its datasheet. However it is bandlimited from 380-430 MHz which might be the reason why it doesn't oscillate. Our amplifier PSA4-5043+ works from 50 MHz to 4 GHz and it doesn't have any filter before it. Maybe duplexer is not sufficient for attenuating other signals.

By the way -125 dBm is the noise signals coming from air to antenna.

 

[volker@muehlhaus]

The amplifier that i have shown below must also cause intermodulation products if this is the case, right?

Intermodulation can be very different, this is a property of the amplifier. Input filters also might help to reduce out-of-band signals, reducing intermodulation.

 

[Squirtle]

How can understand whether our amplifier is oscillating or not? Because when i test it on VNA it seems work fine. Is there any other test setup which i cannot observe on VNA?

 

[volker@muehlhaus]

Intermodulation is something different than osciallation. Both can be tested using a spectrum analyzer at the output.

 

[vfone]

Multicoupler consists of PSA4-5043+ amplifier circuit and 4 way power divider.

Can we see the 4 way power divider you use, and how did you connect it to the amplifier?
The gain of the PSA4-5043+ amplifier at UHF frequencies is about 18dB.

 

[Squirtle]

Actually the gain of psa4-5043+ at UHF (400 MHz) is 22dB but circuit board has a couple of attenuators on it. Here is the block diagram;

Lat-1+PSA4-5043+Lat-2+MVA-2000(Variable Attenuator)4 way PD

MVA-2000 is the output of our PCB amplifier circuit, then output of that board goes to the input of 4 way PD PCB board. As i said it has about 5dB gain on VNA when we test it.

 

[D.A.(Tony)Stewart]

Check the multicoupler output for oscillations with the antenna replaced with a 50R.
The PA drain coil may have a noisy supply causing the -90dBm output. The losses seem excessive for a 4 channel PD with no filter.

Lat-1+PSA4-5043+Lat-2+MVA-2000(Variable Attenuator)4 way PD
-1dB + 22 dB -2 dB - 1.7 dB typ (min) - 7.5 dB = 9.8 dB gain (?) not 5

 

[Squirtle]

Maybe it will be a silly question for you but how can test oscillation? Which instrument should i use (SA?) and how i detect oscillation if it is present?
I don't think the problem is with power supply part because we used suitable capacitors and inductor to make a good isolation and if this is the case, we should have seen that on VNA, right?

For signal chain, multicoupler is probably adjusted to 5dB gain using variable attenuator.

I think the problem is either oscillation or intermodulation. We will test it using a UHF filter before multicoupler tomorrow.

 

[D.A.(Tony)Stewart]

SA is the best tester.

 

[BradtheRad]

Maybe it will be a silly question for you but how can test oscillation? Which instrument should i use (SA?) and how i detect oscillation if it is present?

The grid dip meter was commonly used by HAM radio enthusiasts. One spoke of assembling coil-n-capacitor networks on several homemade cards, to detect various frequency bands. The cards fit in a jig attached to a meter.

 

[Squirtle]

We used a 414 MHz cavity filter and it just worked fine. Over gain problem is fixed. It seems that out of band signals cause the problem.