techinfo

1400 / 121 MHz dual converter description:

Click on converter link to display the schematic for this module.

The circuit is etched on standard 1.6mm double sided F/G pcb material. One side is left

unetched, being the ground plane.

The tuned lines between the RF input stages and the mixer stages are 2mm wide strip-

lines, 10mm long, spaced 5mm. Their ends are offset from their adjacent lines 3mm.

Tuning capacitors " Ct" are 1...5pf ceramic pcb mount types.

All tracks which carry RF signals are 2mm strip lines. All coupling capacitors in the RF

paths (around the MAR-7's ) are porc. chip types. After the DBMs normal ceramic plate

caps are used (121MHz).

The RF stages are MAR-7 MMICs, they have better stability than MAR-6's.

The local oscillator is a POS-2000 free running VCO. As the exact frequency of opera-

tion is not important for this purpose, the VCO offers several important advantages :

No spurious responses due to lower frequency oscillator/multiplier stages and very

little harmonics without extra oscillator filtering (unlike Xtal oscillator/multiplier type

local oscillator ). However, it will drift with temperature up to 4...5 MHz over our typ.

temperature variations. It is therefore not suitable for H-line work.

Have operated for several years with a 96MHz oscillator/ x16 multiplier line-up

and it worked fine. But the VCO type is more compact and requires very few instruments

to align. And the frequency can be easily changed as well.

The whole converter is housed in a diecast enclosure and "lives" out in the yard ,

inside another metal enclosure between both dishes.

>>>>>> Photo of converter <<<<<<<

Dual channel 121MHz I.F. preamp. and phase switch :

All the circuitry from hereon is in the receiver( in the house.) like a 121MHz phase switched

interferometer. Click on IF preamp to display the schematic.

Two surplus DBM modules are configured as variable attenuators, potentiometer adjustable

from the front panel. This enables exact matching between the levels of each channel.

I used some old NEC types, but the more usual SBL-1 would also be fine.

The " T1" transformers are homebrew, very simple. Just 4 turns of enamelled wire

trifilar wound through small UHF TV antenna twin hole balun ferrite cores.

D1 to D3 are low capacitance ,.low power pin diodes. All RFCs are 47uH moulded

miniature chokes.

The circuit is built on 1.6mm double clad F/G pcb material, the unetched side serving

as ground plane and shielding.

Phase switching is only done in one channel (left), but to preserve the matched levels

the other channel has the same line-up. It also allows for individual channel selection

for testing or full power operation, as well as a form of comparison switching, similar

to Dicke switching, only far removed fro the antenna terminals !

>>>>>> Photo of I.F. Preamp. <<<<<<<<

The main 121MHz I.F. strip:

Four MMICs provide the bulk of the required gain (>70dB). After the first stage there is

a multistage tuned band pass filter. Accurate alignment is necessary to obtain the wider

bandwidth ,~ 14MHz. The BPF is homebrew, made from readily obtainable TOKO coils,

slugs and cans. The detector output transformer is wound on a small VHF toroid,

primary 5 turns, secondary 15 turns enamelled 0.3mm dia. copper wire.

Click on I.F.main strip for the schematic.

The circuit is built on two 1.6mm double sided F/G pcb's , one side unetched, provi-

ding all ground connections and shielding. The filter circuit and gain block/detector stage

are built into seperate shielded boxes.

Post-detection circuits ( the " back-end ")

From the I.F.detector the signal is first buffered, after which it is split into 3 circuits.

One goes to the audio monitor ,not used very often, but essential, nevertheless!

Another goes to a simple IF level analog indicator.

The most important one goes to an a.c. amplifier, because only the switched compo-

nent of the detector output level is of interest here, this being around 800Hz in this case.

The coupling transformer is a small transistor radio 2KOhm to 2x 1.5kOhm type.

The 4066 IC performs the full wave rectifying action in synchronism with the I.F.phase

switching diodes, as both circuits are fed from the same square wave oscillator.

The 4013 IC ensures equal mark-space ratio of the switching waveform.

In absence of a (point source) radio signal , the net output from the synch .detector

is essentially zero. A point source signal will have a constantly changing phase delay

between the two antennas, and when the two channels get combined at the I.F.phase

switch the combiner is being rapidly switched between in phase and opposite phase.

causing a slowly varying d.c. component to appear at the synch.det. output. This varia-

tion occurs around the zero Volt mark and will go positive as well as negative during

the drift scan of a radio point source. Extended sources generally get suppressed

that way, depending on the spacing between antennas in wavelength and the angular

extent of the radio source.

The synch.det. output is then buffered and zero fine adjusted, and also displayed on

another analog panel meter.

A large amount of d.c. amplification will be required in order to make those tiny varia-

tions in level visible at the output. Extensive temperature stabilization of all the gear

starting with the LNAs can go a long way to keep things steady. I have chosen to

avoid such extremes and pass the varying d.c. level through a differentiator of ~ 10mins

time constant. This is long enough to prevent the suppression of "source fringing", the

phase changes caused by radio sources , in my case typ. ~ 2....5mins, and short

enough to catch up with the slower environmental variations,which get discriminated

out.

After this the integrator smoothes out the more rapid fluctuations caused by system

random noise. I normally use 30 secs time constant.

The 15uF capacitors in the differentiator and integrator circuits are very low leakage

types (electrolytics and tantalums not suitable!), they are metallised polyester / 63V.

Finally all gets d.c. amplified, for the weaker sources I use typ. x400. After this the

final signal is buffered , level shifted and presented to the output from where it is

connected to the MAX186 A/D converter and processed with RADIO SKYPIPE.

Click on Synch.det. for schematic.

Click on d.c. output for schematic.

Although there is nothing terribly critical about the pcb layout of the back-end circuits,

all signal handling linear ICs have their + / - 12V supplies individually decoupled via

100 Ohm resistors and 2u2 tantalum caps. This is not shown on the schematics.

>>>>>>> Photo of synch.det. and dc amp. pcb <<<<<<<<

Module alignment :

For the back-end : only a DMM , also any cheap ,old 'scope

I'll soon add some CRO screen shots for synch.det. waveform

illustration and troubleshooting.

For the I.F.main amp.- if a commercial bandpass filter is used, it will already be

fixed, pre-aligned, no alignment possible. This may limit the

maximum I.F.frequency to 70MHz. For higher I.F's the filter

needs to be sweep aligned. I use the HP sweeper and mix

it with the output from the Marconi sig.gen to virtually give me

any I.F.frequ. I may need up to nearly 300MHz.

For the I.F.pre amp.- only signal gen. required to verify that both channels work

properly in conjunction with the DBM type attenuators.

For the converter - the homemade 1400MHz signal source (VCO) , also the

3 GHz frequency counter. The module's local oscillator

frequency is picked up via a short piece of coax with a small

"snoop loop" a its end, holding it near the oscillator striplines

which go to each DBM . In my case I trim the module's VCO

tuning voltage to +7.7V, resulting in 1525MHz injection frequ.

Next I adjust the L-band signal source to 1400MHz and dis-

connect its antenna. Only a leakage signal escapes from its

diecast box, enough to drive the module input stages. Then the

3 strip lines are carefully peaked with a non metallic trim tool.

The 3 trim cap adjustments are a bit interacting, so the pro-

cess needs to be repeated a couple of times. The input band-

width will be ~ 45 - 50 MHz. Both channels must be adjusted

seperately.

For the LNAs - If any of the commercial broad band LNAs are used, no align-

ment will be needed. For homebrew and/or narrowband

LNAs (which I prefer), some optimisation will be in order.

Not having a calibrated NF meter , I can only make relative NF

measurements with the homebrew diode meter. In any case,

at very low noise figures the commercial automatic NF meters

often have error margins comparable to the NF they are sup-

posed to measure.