Evaulating weak signal LNA performance - 144MHz

18 January, 2017 - Reading time: 6 minutes

I conducted some recent tests with VK7MO who was running a low powered transmitter on WSPR (100mW). This gave me the opportunity, among other things to evaluate how well my Minikits LNA masthead (preamp) was performing on 144MHz.

I must admit when I put my preamp up the tower, I was a victim of the "all knobs to the right" rule, setting it for maximum gain (+18dB). The ham advice I read at the time specified to turn up the gain until the background noise increases slightly "1 to 2 S points". But I live in a relatively remote and quiet location, so the background noise did not lift all that much - so I thought. More on that later.

Anyhow, testing with 7MO with my preamp on/off made no discernible difference in my decodes from him which were relatively weak (-28). As the local 2m beacon is off air, I didn't have anything else to compare it with at installation time. Turning the LNA on resulted in the WSPR audio level bar graph increasing by 7dB. Some searching was done through various posts on the VKLogger, and some valid questions were raised by VK2OMD on the subject of masthead preamps and LNA's.

1. (When) Is a preamp worthwhile?
2. Can a preamp degrade SNR? How?
3. What is more important, Gain, or Noise Figure?

To answer some of these questions, I needed to calculate my overall noise figure (NF) of my system. A good article to start with is here.

I decided to calculate the NF of my system without the preamp inline. This would give me a baseline to see if there was any improvement. Below is the diagram from the above link with my data in it. The NF of my rig (FT897D) was a rough guesstimate as I don't have the ability to measure it. However QST magazine post reviews of rigs Minimum Discernible Signal (MDS) figure, from which, you can work out the NF value. I couldn't find the FT897 review, but I could the 857 which is essentially the same radio.  MDS is -140, so the calculated NF is 6.9dB. A local amateur did some measurements on the 857 and concluded it's NF was 7.3dB, so pretty close.

Below is an example of what to expect when there is no preamp in the line. I use a length of RG213 from the antenna and LDF-450 down the tower to the amplifier, bias tee, radio etc. I haven't included those items in the loss path as it is negligible. There would also be some induced noise in these components, again I think negligible. We can see that receivers SNR is degraded only by half a dB due to the coax loss.

Quoted from another article I read on the internet - "When amplifiers are cascaded together in order to amplify very weak signals, it is generally the first amplifier in the chain which will have the greatest influence upon the signal to noise ratio because the noise floor is lowest at that point in the chain". My Minikits preamp claims it's NF is typically 0.7dB at 145MHz.

Below is with my preamp in the line. Note there is now extra loss of 0.3dB between the antenna and preamp for the coax tail. It's CNT-240, about 4m of it to the preamp which has to be mounted vertically on the mast. My last preamp had water enter when it was mounted horizontally. The closer to your antenna terminals the better, as any loss before the preamp degrades the overall SNR. My SNR is enhanced by 6.15dB due to the preamp in line and the losses before that to the antenna being only 0.3dB. As my cable loss is only 0.55dB after this, the gain has been set too high (18dB). The gain should be set to only offset the losses plus 10dB. Additional gain has little positive effect on the SNR and increases the risk of intermodulation and overload on strong signals. It also probably has the potential to degrade the SNR as it's just amplifying additional noise, maybe masking our weak signal.

Conclusions 

• Overall gain needs to be reduced to the 10dB level. There is not much loss between the LNA and the radio.
• Is the preamp worthwhile? Potentially in my quiet location - or for EME where every dB counts
• Can a preamp degrade the SNR?  Too much loss before the pre-amp will result in no improvement, and even a degradation. Too much gain can have the same effect.
• Noise figure is more important in overall system performance. Perhaps the biggest improvement in my already existing system is using a radio with a lower NF. I've had a number of amateurs raving about the performance of their FT847 rig. This has a reported NF of 4.9dB and a lower MDS (2dB improvement over the 897D). IC7000, IC-910H and 706MKIIG have similar specs. An IC9100 has a reported NF of 3.9dB and a MDS of -143dBm! A 3dB or twice the improvement, but with this improvement comes the price tag!
• You are probably going to notice at least *some* improvement in a properly designed system. However improvements in the feedline loss, using good quality coax and connectors would be paramount over having a preamp. Also there would be negligible improvement in noisy locations. Another more important component not mentioned above is improvement in the antenna, use a high gain yagi with good sidelobes and a low antenna temperature. Noise is everything! If you can't reduce the noise, you can't hear the signal masked by it! 
• When picking a LNA, look for a low NF over gain. You can improve the feedline from the LNA to the radio so extra gain is negated. Remember the first amplifier in the chain will have the greatest influence on the SNR and system performance.