Multichannel HF / VHF Radiometer

This project materialised from the ongoing frustrations of unattended 20MHz recording (with  RJ phased dipoles) . Sporadic radio transmissions occur on or near a selected 20MHz frequency at various times of the day and evening. Unless that activity was very strong, weaker transmissions of some duration may or may not represent solar (of Jovian) radio bursts. 

For solar radio burst observations a much wider frequency range is needed than can currently be provided by available SDR devices.
Scanning type radio spectrometers ( "frequency agile" ) are normally implemented , typically at amateur level  the CALLISTO system.
There is a penalty for this , which is a reduction in sensitivity .  So I opted to build a system which operates on 7 different frequencies across the spectrum from 18 - 73MHz.  An additional "spare" ADC channel is used to record the signal level from the NWC  VLF station on 19.8kHz .  I could have made the frequency much wider , but I had no specific need for this , because it is covered by my Callisto station from ~50MHz upwards anyhow. 
Another point to consider was the antenna system  .  The Radio Jove dual dipoles were really only useful around 20MHz , but I wanted to cover 18 - 75MHz with one antenna system without having to resort to a huge (steerable)  LPDA . 
I decided to build a couple of T2FD type dipoles , which cover typ. a 5:1 frequency range if cut for the lowest frequency .
They are installed in a slightly sloping fashion , spaced ~ 7.m apart and are phased together with a splitter into a simple pre-amplifier
before the ~ 70m RG-6 coax. run back to the house.

The multichannel receiver had to be inexpensive , so I designed it around the MC 13135 dual conversion FM I.C.
These can be had from Futurlec for about $1.20 each plus a handful of other components. The same PCB was implemented for each channel , the output signals are log.scale dc level .  Not as impeccably accurate as an AD 8307, but ok for the purpose ,especially when considering the cost factor. 
In my case here I also included the VLF signal , however this one is not log scale , but happens to co-exist with the others on the SkyPipe chart .
Below is a 6 hr. section of a SkyPipe chart recorded .  During the daytime and early evening the 18 and 20MHz frequencies are quite riddled with the signals from ionospheric sounding stations .
And here is a complete 24 hr. recording
As can be seen , it can get very busy especially at the lowest two frequencies .  Also the 20MHz background noise increases every evening from about about 7pm until 11pm , for reasons not completely clear to me at this time.

Nov. - Dec.2010 has seen a reasonable amount of solar activity , mainly evident from some spectacular flare videos recorded by the SDO (solar dynamic observatory) /HMI (helioseismic + magnetic imager) . Perhaps surprisingly this rarely resulted in HF/VHF radio flare activity !  It is notable that even quite substantial X-ray flare-ups seldom now occur very rapidly , instead they build up over a few minutes .  Theory has it , radio flares can only be generated from sunspots with mag. fields of ~ 1500 Gauss the least.  Most current spots seem to have ~2000 Gauss average field strength , in years past this figure was ~ 2700 Gauss.  This makes for very nice and regular shaped spots, however only rather contorted and irregular spots with rapid mag.flux swings give us radio bursts .

Below is a typical recording , it is of a C5 magnitude X-ray flare , the uppermost recording trace shows the signal level from the NWC
(North West Cape) VLF station operating on 19.8kHz . The increased X-ray flux enhanced the signal markedly, but nothing was noted from 25 - 73MHz , this was checked with data from IPS Culgoora (18-1800MHz) and Learmonth (25-180MHz) who also failed to record the event .
2010-Dec-15                                                    Corresponding X-ray sat. data