An easy to build HF/VHF/UHF Noise Generator and
its Calibration. Updated on Dec. 13 2021

Introduction.
Many years ago I bought a NoiseCom NC302LBL (1) noise diode in order to build a noise generator. Unfortunately, after moving to a new home I was unable to remember where I stored this small  but expensive component!
Recently, I was looking for a Varicap diode in my workshop when I came across the NC302LBL among a batch of variable capacitor diodes.
So it was the time to build the noise generator.

1. The schematic diagram.
I use the design presented by W. E. Sabin in May 1994 QST (2).
I just added a LM317 regulator supplying the noise generator with a stable 20 V voltage : so, I don't depend on an external regulated power supply.
The schematic diagram is shown below :

NG_Schem


2. Construction.
I designed a small PCB and etched it using the Laser Toner Transfer method (3). An aluminum box houses the board and some other components.
The photos below show the front panel and the inside of the box.

Front

NG

3. Calibration of the Noise Generator : December 2021 update.

When I built this generator, I didn't have access to an accurate Noise Generator and I couldn't calibrate it.

Recently I tried the following calibration method to compute the ENR of my noise generator. It requires  a signal generator and a receiver.

When the noise generator is OFF, the noise power available at its output is the thermal noise    Noff = kTB     where T is the temperature and B the bandwidth. Thus Noff = -174 dBm/Hz at room temperature.
When it is ON, the noise level is Non =  Noff+Ne   where Ne is the noise created by the generator.

The ENR is defined as the ratio of excess noise to thermal noise   ENR = Ne/Noff
 If we introduce the Y Factor as Y = Non/Noff  we have  Y = 1+ENR

If the noise generator is connected to a receiver having a Noise Factor NF, Sabin shows that   NF = ENR/(Y-1)                                                                       

To determine the NF we measure the Minimum Discernable Signal (MDS) of the receiver having a Bandwidth B.

We know that (4)                                                MDS(dBm) = -174(dBm)+ 10Log B(Hz)+ NF(dB)

Thus                                                                    NF(dB) = 174-10Log B(Hz)+MDS(dBm) 

And                                                                     ENR = NF (Y-1)

I measured the MDS of my K2 receiver (bandwidth B # 2.15 kHz) by connecting it to the output of my HP8657B signal generator (F = 14.1 MHz) and found :

                                                                          MDS # -126.2 dBm  thus NF # 14.5 dB

I got the Y factor using 2 methods : one with the RX audio output connected to an RMS voltmeter and the second from the spectrum of the same  audio signal, when the noise generator was switched ON and OFF. I used the Spectrum Lab software (5) to get the spectrum. These 2 methods give an average Y # 16.7 dB.

Thus                                                                   ENR # 22.1 dB at 14.1 MHz

The error can be estimated at +/- 0.2 dB.

This ENR is thus in good agreement with the 22.4 dB at 15 MHz published by Sabin.
       


Notes

(1) https://www.noisecom.com/products/components/nc100-200-300-400-series-chips-and-diodes

(2)  W. E. Sabin W0IYH,"A Calibrated Noise Source for Amateur Radio," QST, May 1994, pp. 37-40

(3) https://hackaday.com/2016/09/12/take-your-pcbs-from-good-to-great-toner-transfer/

(4) W. Hayward, "Introduction to Radio Frequency Design", ARRL, Newington, CT, 1994, p. 351

(5) https://www.qsl.net/dl4yhf/spectra1.html


 Home