Tuner Signal-to-Noise (S/N) - how to improve?

Whitehall

Super Member
I love my Marantz 4300 but its noise floor is higher than other tuners I own even on the same antenna system. I do love that "blackness" between notes. I suspect it is a design tradeoff as the manufacturer did not provide a specification for S/N. This level of noise is not unbearable and is barely noticable, just that it gets wondering about the following questions:

1) what design features affect S/N?

2) can better maintenance reduce the noise floor?

3) are there easy mods to improve S/N?
 
Whitehall said:
I love my Marantz 4300 but its noise floor is higher than other tuners I own even on the same antenna system. I do love that "blackness" between notes. I suspect it is a design tradeoff as the manufacturer did not provide a specification for S/N. This level of noise is not unbearable and is barely noticable, just that it gets wondering about the following questions:

1) what design features affect S/N?

2) can better maintenance reduce the noise floor?

3) are there easy mods to improve S/N?

Well my RF theory is pretty rusty but I seem to recall the S/N ratio is most limited by the NF, noise figure, of the first RF amplifier active device. Because that stage adds gain the noise figure of the next stage is improved by the gain of the first stage and so on done the active stages. Bandwidth also has a large effect on the S/n ratio, with narrower being better then wider for total noise, however then there is a tradeoff of audio frequency Vs bandwidth so you can see that minus any component failures, the S/N ratio of a given tuner is an engineering tradeoff that gets designed in. A low noise mast mounted RF preamp can help is that it then becomes the first active amplifier stage and has the biggest effect on the system S/N ratio.

Lefty
 
I've read elsewhere that the mixer is the largest contributor to noise within the tuner.

I was thinking that there might be a modern, low noise FET for the front end (currently a pair of 2SK19Y and a 2SC535B local oscillator) but I haven't found one.

I've posted extensively on the possibility of adding a GaAs pre-amp up antenna - that's certainly lower noise gain than the 30 year old stuff inside my box.

However, once the signal gets to full quieting, signal strength doesn't seem to matter - the noise is still there, albeit near the limit of discernment.
 
Contributors to the noise floor:
1. Front end amp.
2. Mixer
3. LO phase noise
4. IF amps
To find the source you need a spectrum analyzer and a good rf generator. If you have a generator, you can check with a CW signal. The noise should drop above the quieting level and reach a minimum. If it's quieter at that point than when it's receiving broadcast signals, then maybe it needs alignment or filter mods to bring the signal level up. If there's no difference, then the source of the noise is probably one of the above components.
 
I've read elsewhere that the mixer is the largest contributor to noise within the tuner.

Well I am rusty but I don't think that is correct. The mixer is a biggest source of IM distortion from signal overload and the first active device that becomes overloaded with too strong RF signals even if it's not the station you are tuned to. For a long time the best peforming mixer was the passive dual balance diode bridge, it has very good overload specs but requires a local oscillator with much higher power then an active transistor or fet mixer stage. I've lost track in details but the current state of the art mixers are, I think are MOSFET balance mixer ICs and give balanced diode performance without the high LO drive requirments.

I was thinking that there might be a modern, low noise FET for the front end (currently a pair of 2SK19Y and a 2SC535B local oscillator) but I haven't found one.

I'm sure their are modern FET and GasFet transistors that have wonderful noise and gain specifications. The problem is that one has to design the circuit for proper impedence matching and bias for that specific device. It's not as simple as just replaceing the transistor and doing the alignment adjustments.

I've posted extensively on the possibility of adding a GaAs pre-amp up antenna - that's certainly lower noise gain than the 30 year old stuff inside my box.

However, once the signal gets to full quieting, signal strength doesn't seem to matter - the noise is still there, albeit near the limit of discernment.

Even with low noise devices the large IF bandwidths of FM tuners (even in 'narrow mode') means that there will always be a certain amount of noise, it's a mathamatical thing. The best signal to noise ratio communications mode used in ham radio was CW (morse code) because one could run the IF filtering down to say 500hz and the 'noise window' drops dramtically. Obviously one can not get a 20khz hi-fi signal through a 500hz filter so their is always a trade off of noise Vs bandwidth.


Lefty
__________________
 
Back
Top Bottom