The MGF1302 is a low noise GaAs FET designed for S band to X band amplifiers, and was very popular in ham equipment until the arrival of pHEMT devices.

An important characteristic of the MGF1302 is that matching the input circuit for maximum gain (maximum power transfer) does not achieve the best Noise Figure… and since low noise is the objective, then we must design for that.

The datasheet contains a set of Γ_{opt} for the source impedance seen by the device gate, and interpolating for 1296MHz Γ_{opt}=0.73∠-10.5°.

Lets convert Γ_{opt} to some other useful values.

The equivalent source Z, Y and rectangular form of Γ_{opt}= will be convenient during the circuit design phase.

The other important characteristic is Zin of the gate in the final circuit configuration, and that is derived from a model of the device in circuit. The value is 7.8-j164Ω. At low frequencies, the FET gate looks like an extremely high impedance, at higher frequencies more like tiny capacitance with very low equivalent series resistance (ESR), and still higher frequencies the capacitive reactance is lower and ESR higher, and still higher frequencies lead inductance comes into play and the gate looks inductive with even higher ESR.

A model of the antenna to gate circuit was built in SimSmith v16.9.

Above, the SimSmith model which includes a swept length of LDF4-50A transmission line.

The standard display was supplemented with impedance, admittance and Γ looking from the gate back towards the antenna. These make it easier to adjust the L and C components for the desired outcome. The G component of Y is most sensitive to adjustment of L and the B component to adjustment of C. So, they are both adjusted to approximately obtain Γ_{opt}=0.7178-j0.1330 from the conversions done earlier.

A sweep is shown for power from the source (antenna) and power into the FET gate for a range for transmission line lengths. Whilst loss between antenna and gate may seem high, the LNA delivers around NF<1dB and 13dB of gain from its input terminal (ie looking into L1) to output despite the lack of conjugate matching.

The behavior demonstrates the complex interaction of source, transmission line and load… worth studying.

The discussion is about a low noise receiving system where optimal results come from an input circuit that is not designed for maximum power transfer.

]]>## Effective Bandwidth

The contribution to the available output noise by the Johnson-noise sources in the signal generator is readily calculated for and ideal or square-top band-pass characteristic and it is GKTB where B is the bandwidth in cycles per second. In practice, however, the band is not flat; ie, the gain over the band is not constant but varies with frequency. In this case the total contribution is ∫G

_{f}KTdf where G_{f}is the gain at frequency f. The effective bandwidth B of the network is defined as the bandwidth of an ideal band-pass network with gain G that gives this contribution to the noise output.

Above is the response of the ‘factory’ 2400Hz soft filter in an IC-7300 (SDR) transceiver. It is not an ideal rectangular response.

To perform the calculation described by Friis, ∫G_{f}KTdf, we firstly need a G(f) dataset. The above plot is of the log of G(f) and to perform geometric operations to find the area under the curve is quite misguided.

Above is a plot of G(f) (measured with Gaussian noise integrated over a period), and we can find ∫G_{f}KTdf wrt G at some reference frequency (1kHz in the above example as that is what is used for sensitivity measurement).

In this case the filter -6dB response is 377-2616Hz=2239Hz, and Effective Bandwidth wrt gain at 1kHz is 2077Hz.

In my own articles and software I usually refer to this as the Effective Noise Bandwidth (ENB) to be clearer.

ARRL to be different refer to Equivalent Rectangular Bandwidth but they do not expose how they calculate it (it is a hammy thing).

The term Equivalent Noise Bandwidth is sometimes used.

- Friis, HT. Noise figures of radio receivers. Proceedings of the IRE, Jul 1944 p420.

(Friis 1944) suggested that temperature as reference temperature and it has been widely used since. One may also see 293K (eg in certain ITU-R recommendations), but in my experience, 290K is most commonly used and is for instance the basis for calibration of Keysight noise sources in Excess Noise Ratio (ENR).

The assumption in measurement of Noise Figure or of sensitivity is that the ‘cold’ source has a known source resistance with Johnson noise equivalent to 290K (16.85° C). That noise producing resistance is commonly achieved using a large attenuator at the generator output.

- Friis, HT. Noise figures of radio receivers. Proceedings of the IRE, Jul 1944 p420.
- Keysight. Jul 2018. Keysight 346A/B/C noise source operating and service manual.

The update corrects an error in conversion between ENR and temperature where Tcold<>290K.

- Duffy, O. 2007. Noise Figure Meter software (NFM). https://owenduffy.net/software/nfm/index.htm (accessed 01/04/2014).

The analysis assumes linear systems (eg no signficant intermodulation distortion).

Above is Fig 2 from ITU-R P.372-13 which shows some key components of total ambient noise. The solid line is entitled “minimum noise level expected”, and it is a combination of curves B, C and D. Above 0.7MHz, only curves C and D are at play.

Curve C is the Quiet Rural curve from Figure 10 (curve D), and D is the Galactic curve (curve E) from the same figure. It is important to note that Galactic noise is diminished below foF2 (due to ionospheric shielding), and the extrapolated curve D crosses curve C at 1.931MHz. So, the minimum noise curve is in fact dependent on foF2 at the time of interest.

Note that ambient noise levels experienced by hams in suburban residential areas are quite likely to be 20dB or more higher than the Quiet Rural curve above.

For a conservative analysis let’s make an assumption that foF2 is fairly high, say 15MHz (it very rarely reaches this value).

Above is a calculation of the maximum system noise figure on that basis for a S/N degradation of 1dB in a receive system.

If for example, a receive system had a NF of 18dB with preamp off and including an allowance for antenna system loss, that is an adequate performance for the 7MHz band and below.

Low foF2 results in ionospheric shielding above the lower foF2.

The plot above shows the effect foF2 of 5MHz, and it can be seen that the figures in the region 5-15MHz are higher than in the previous plot.

In this case, the 18dB NF receive system is quite adequate for the 14MHz band and below.

So, what is the right condition for ambient noise as it varies from place to place, time to time etc.

A simple approach is that if the received noise power increases by 6dB when switching from a matched load to the antenna, S/N degradation will be less than 1dB, and very close to 1dB when the receive system NF is greater than 10dB.

Traditional ham wisdom describes comparisons between antenna disconnected (implying the receiver input socket has nothing attached) and antenna connected. Like most traditional ham wisdom this is bad advice, the equivalent noise power input of a receiver with matched load is defined by its NF, not so for the ‘disconnected’ condition.

An even better option is to observe the S/N of a steady signal (ie not subject to fading) and calculate the S/N ratio for different configurations.

Here is an example using Spectrum Lab to calculate the SINAD (similar to S/N) of a signal at 1kHz in the passband.

It does not matter whether the noise and the signal changes, the key statistic is the calculated SINAD. So, you can turn preamps off and observe the impact on SINAD. The beauty of this method is that it also captures the actual noise at the time, including contribution due to intermodulation distortion.

- ITU-R. Sep 2016. Recommendation ITU-R P.372-13 (9/2016) Radio noise.
- ITU-R. Aug 2019. Recommendation ITU-R P.372-14 (8/2019) Radio noise.

Carol gives the following table of measurements and calculated results.

Table 1. Transverter Measurements | |||||||
---|---|---|---|---|---|---|---|

Freq MHz |
Noise Source ENR (dB) |
Noise/10 kHz |
Conversion Gain (dB) |
Noise Figure (db) |
|||

50 Ω expected | Noise On | Noise Off | On-Off (Y) | ||||

144 | 15.2 | -134 dBm | -118.8 dBm | -132.1 dBm | 13.3 dB | 26.5 | 2.1 |

432 | 15.3 | -134 dBm | -118.7 dBm | -131.7 dBm | 13 dB | 24.1 | 2.5 |

Lets focus on the 144MHz measurements.

These are a measurement of the system (ie Flex 1500 transceiver + Electraft XV144 transverter) and Y factor based calculation of Noise Figure (NF), and note that the Noise ON and Noise Off figures are from the Flex 1500 which has 26.5dB subtracted (a calibration adjustment for expected transverter gain).

This article presents a simulation of a two stage measurement which establishes the NF and Gain of the transverter. The two stage technique is described at Noise Figure Y factor method calculator.

To achieve that, we must determine the NF of the Flex 1500, the ‘instrument’. Since the necessary measurements were not made, for the purposes of this article I will assume that the NF of the instrument is 10dB, that the noise source has an ENR=16dB at the transverter output frequency, and calculate the Noise Off and Noise On powers in the same measurement bandwidth for that scenario. Instrument NoiseOn=-117.0dBm and NoiseOff=-124.0dBm.

For the system measurements, I will add the gain adjustment back in to the Table 1 figures, so DUT NoiseOn=-92.3dBm and NoiseOff=-105.6dBm, and the ENR was given as 15.2dB.

Entering them into Noise Figure Y factor method calculator we have:

and the calculated results:

The process calculates the Gain and NF of the transverter itself to be 26.3dB and 2.05dB respectively, and the system NF=2.11dB.

These figures depend on my assumption of the NF of the ‘instrument’, and different figures will flow into some differences in the calculated results.

- Milazzo, C. Aug 2015. Signal level measurement with PowerSDR and external transverters.

]]>

Farson gives a table of MDS in 500Hz bandwidth figures for the 6700 on certain bandws, including MDS for 4 RF Gain configurations, 0, 10, 20, and 30dB.

Above is Farson’s data with my chosen RF Gain option (selected for SND<3dB) and calculated values in yellow and orange for:

- system noise figure calculated from MDS and assuming that the effective noise bandwidth is 500Hz;
- ambient noise figure for Quiet Rural precinct from P.372-14;
- total noise power (ie internal plus external); and
- calculated signal / noise degradation due to receiver internal noise.

Note that Quiet Rural may be seen as a lower bound for man made noise from 0.7 to 30MHz, but it should be remembered that at frequencies above foF2 as varies from time to time, Galactic noise is greater… so in a sense this is a very conservative analysis towards 30MHz.

A lossless antenna system is assumed.

A further issue is that the effective noise bandwidth (ENB) of a nominal 500Hz filter may be significantly different, and we have no knowledge of the actual ENB.

Notwithstanding these issues, the calculated values in the SND column give indicative expectations.

The same table reworked for P.372 Rural precinct gives better SND.

A reality check for your own scenario is to compare the Pt column with receiver power measurement in 500Hz. Higher than Quiet Rural will increase the figure, antenna system loss will decrease the figure. It will be rare for most of us to observe Quiet Rural ambient noise.

- Allison, B; Tracy, M; Gruber, M. 2011. Test Procedures Manual Rev L. ARRL Newington.
- Ellison,T. Jun 2019 How to determine the amount of RF Preamp gain to apply for band conditions (accessed 22/09/2019).
- Farson, A. Oct 2014. Test Report for Flex-6700.
- ITU-R. Sep 2016. Recommendation ITU-R P.372-13 (9/2016) Radio noise.
- ITU-R. Aug 2019. Recommendation ITU-R P.372-14 (8/2019) Radio noise.
- Youngblood, G. Oct 2018. 6600 noise levels +10dB higher on my new 6600 vs. my older 6500 (accessed 21/09/2019).

optimal receiver noise figure relationship to antenna noisein a blog posting about SDR receivers.

This article discusses that posting in the context of linear receivers, ie effects of intermodulation distortion are not included.

His gives the following advice:

For optimal weak signal performance near the atmospheric (antenna) noise floor you want your receiver noise floor (sensitivity/MDS) to be 8 to 10 dB below the noise coming from the antenna. For strong signal reception, less sensitivity is almost always better.

The terminology is not industry standard, but that is quite usual for hams who have a need to redefine well known terms, and this is really loose with implied equivalence (eg sensitivity/MDS).

ITU-R P.372-14 speaks of natural noise as including atmospheric noise due to lightning

, and also speaks of man made noise.

It is likely Youngblood is actually talking about man made noise since he uses man made noise figures from an earlier revision of P.372.

Optimal

is a compromise between weak signal performance (ie S/N degradation due to internal receiver noise) and handling of strong signals that might clip in the ADC of an SDR receiver.

He gives a table of measured MDS (minimum discernable signal, which actually is synonymous with Noisefloor) for recommended configurations of a Flex 6600 radio on several bands.

Above is Youngblood’s data with my calculated values in yellow and orange for:

- system noise figure calculated from MDS and assuming that the effective noise bandwidth is 500Hz;
- ambient noise figure for Quiet Rural precinct from P.372-14;
- total noise power (ie internal plus external); and
- calculated signal / noise degradation due to receiver internal noise.

Note that Quiet Rural may be seen as a lower bound for man made noise from 0.7 to 30MHz, but it should be remembered that at frequencies above foF2 as varies from time to time, Galactic noise is greater… so in a sense this is a very conservative analysis towards 30MHz.

Note also that Youngblood states the table below shows the gain setting and the expected MDS with no antenna attached…

. This suggests a misunderstanding of the meaning of MDS. MDS or Noisefloor is the total internal and external noise with a nominal dummy load on the input as demonstrated by (Allison et al 2011).

A lossless antenna system is assumed.

A further issue is that the effective noise bandwidth (ENB) of a nominal 500Hz filter may be significantly different, and we have no knowledge of the actual ENB.

Notwithstanding these issues, the calculated values in the SND column give indicative expectations.

The estimated SND is perhaps higher than desirable on some or even most bands if the user was actually experiencing typical Quiet Rural ambient noise with a reasonably efficient antenna.

Readers will note that where the total noise power Pt is barely above MDS (ie most of the noise is internal), SND is relatively high.

If the configurations did achieve Youngblood’s suggested internal noise of at least 8dB below external noise, the SND would be 0.64dB, a fairly satisfactory statistic for many.

The same table reworked for P.372 Rural precinct gives better SND.

A reality check for your own scenario is to compare the Pt column with receiver power measurement in 500Hz. Higher than Quiet Rural will increase the figure, antenna system loss will decrease the figure. It will be rare for most of us to observe Quiet Rural ambient noise.

- Allison, B; Tracy, M; Gruber, M. 2011. Test Procedures Manual Rev L. ARRL Newington.
- Ellison,T. Jun 2019 How to determine the amount of RF Preamp gain to apply for band conditions (accessed 22/09/2019).
- ITU-R. Sep 2016. Recommendation ITU-R P.372-13 (9/2016) Radio noise.
- ITU-R. Aug 2019. Recommendation ITU-R P.372-14 (8/2019) Radio noise.
- Youngblood, G. Oct 2018. 6600 noise levels +10dB higher on my new 6600 vs. my older 6500 (accessed 21/09/2019).

Essentially, two questions were asked:

- what is the minimum HF ambient noise level; and
- explain observation of lower HF ambient noise level.

Above is Fig 2 from ITU-R P.372-13 which shows some key components of total ambient noise. The solid line is entitled “minimum noise level expected”, and it is a combination of curves B, C and D. Above 0.7MHz, only curves C and D are at play.

Curve C is the Quiet Rural curve from Figure 10 (curve D), and D is the Galactic curve (curve E) from the same figure. It is important to note that Galactic noise is diminished below foF2 (due to ionospheric shielding), and the extrapolated curve D crosses curve C at 1.931MHz. So, the minimum noise curve is in fact dependent on foF2 at the time of interest.

Note that ambient noise levels experienced by hams in suburban residential areas are quite likely to be 20dB or more higher than the Quiet Rural curve above.

For a conservative analysis let’s make an assumption that foF2 is fairly high, say 15MHz (it very rarely reaches this value).

Above is a calculation of the minimum ambient noise figure on that basis, and the calculated S/N degradation in a receive system with system noise figure 20dB. The degradation at the lower end is quite small, but increase at the higher frequencies, though still only modest degradation.

It will be rare, but not impossible to observe lower ambient noise.

The figures from P.372 are medians, so one expects some variation high and low, and some locations might be consistently a little lower

The survey on which P.372 is based used a short vertical monopole, and slightly lower Fam may be observed with a horizontal antenna.

Low foF2 results in ionospheric shielding above the lower foF2.

The plot above shows the effect foF2 of 5MHz, and it can be seen that S/N degradation in the region 5-15MHz is lower than in the previous plot.

In some solar disturbances, the D layer thickens and there is greater attenuation of Galactic noise, again reducing ambient noise where Galactic noise made a significant contribution. This reduction might not have great benefit as the D layer thickening will also result in ionospheric propagation fade outs.

It must be kept in mind that the receive system NF shown on the plots is the NF at the air interface, and it includes the increase in NF caused by loss in the antenna system. For example, a receiver with NF=10dB and a small loop antenna with efficiency of say 1% (-20dB) has a system NF of 10+20=30dB which would result in a S/N degradation of 6dB at 14MHz under the minimum noise scenario discussed in this article. So, whilst observed S/N degradation might be relatively high, the cause it that receive system NF is very high, a result of antenna system loss.

- ITU-R. Sep 2016. Recommendation ITU-R P.372-13 (9/2016) Radio noise.
- ITU-R. Aug 2019. Recommendation ITU-R P.372-14 (8/2019) Radio noise.

This discussion considers the question applied to linear receivers, ie receivers with zero intermodulation distortion (IMD) and other non ideal characteristics, other than their internal noise which can be described by their Noise Figure (NF).

By definition, NF is the amount by which the component or system degrades the NF, so in dB it is the difference in the S/N in to S/N out. Implicit in that definition is that it is based on source internal noise of 290K equivalent.

So for example lets say a receiver with effective noise bandwidth 2000Hz measures sensitivity of -125dBm for 10dB S/N out. We can calculate the noise in 2000Hz bandwidth from a 290K source to be -141dBm, and therefore the input S/N is -125 – -141 = 16dB. The ratio of the input S/N to output S/N is the difference in those in dB, 16-10=6dB. The NF is 6dB. We can also calculate an equivalent internal noise temperature of (10^(6/10)-1)*290=865K.

By convention, ambient noise (or external noise) is expressed in Kelvins, or dB wrt 290K. That does not imply that an antenna contributes exactly 290K.

So if Fam=13dB, Tam=10^(13/10)*290=5786K.

We can then say that if this receiver adds 865K of noise to the ambient 5786K, the total noise is now 6651K, an increase of 10*log(6651/5786)=0.6dB, so S/N has been degraded 0.6dB by the receiver in this scenario.

So, there are two factors that influence the extent of degradation, the ambient noise, and the receiver internal noise.

The chart above shows the expected median ambient noise figure in Rural precincts from ITU-R P.372-13, and the calculated degradation in S/N ratio caused by internal noise of a receiver with NF 20dB (which is about what might be expected with the PREAMP OFF).

Let’s consider a receive system with NF=1.0dB (Teq=75.1K) at the antenna connector, and quiet radio sky of say 40K (we are pointing to the sky, this is not terrestrial noise). Though this might look like a pretty low noise receive system, you can see that total noise temperature is 75.1+40=115.1, so the degradation is 10*log(115.1/40)=4.6dB.

The situation would be very different for an antenna pointed horizontally as Tam would be a lot higher, and site dependent.

We can calculate the degradation in S/N caused by receive system internal noise, and that should drive considerations about how good the receiver needs to be.

The discussion has assumed a linear receiver (zero IMD), and presence of IMD will add noise to the system, further degrading S/N. Measures to reduces susceptibility to IMD, even where they seem to reduce sensitivity, may result in improved S/N recovered.

]]>