Softrock Centre Phase/Amplitude

Tests done using on-board VIA sound at 48Khz. Note the Rocky charts show all 48KHz, the HDSDR charts examine the centre only. The Phase/Amplitude figures provided by Rocky and HDSDR do not match, I have no information as to how they are presented.

Note that the centre of my charts is not likely to be exactly 0Hz, my receivers are fairly close but I did not check calibration before starting.

This one shows a 30/20/17m Softrock Ensemble as original.

No further checks have been done on this. From results below we can assume considerable variations in the centre, not shown by Rocky.

This one an unmodified RX3. Note this has automatic BPF selection, input transformer is wideband.

Plot of unmodified RX3 using HDSDR's manual adjustments. Enables view a little closer to the centre. Readings nearer the centre would have been difficult due to rising noise levels. Of course at the very centre there will be no signal.

Original

Plot of RX3 with 100uF extra decoupling at centre of input secondaries.

Here, a Softrock 40, a USB powered version. Rocky shows a "good" response but HDSDR shows what happens in the centre.

The QRP Labs receiver showed a perfect spectrum in HDSDR. The centre was perfectly clean. This has the op-amp stabilisation, no capacitor at all.

The transformers were removed, I've never been happy about transformers, careful attention to the prevention of ground loops makes them redundant.

But phase curves were "normal".

Original. no transformers.

A close-in at the centre.

Adding a 100uF capacitor did not improve things.

Reference this thread https://groups.io/g/softrock40/topic/55843839#87844 Message 87813 onwards.

It is strange that the first example did not show the effect. I think all Softrocks have the input transformer's secondary centre tap decoupled by at least a 0.01uF and a 4.7uF. Is the circuit sensitive to the quality of the decoupling capacitor? The one I added I selected for lowest ESR using a component tester. (It was like most components I use old, recycled.)

Note that the balance is affected by the antenna, changes here, even maybe an antenna swaying in the wind might affect it? A preamp or buffer might reduce this effect.

I consider this more as simply interesting than a defect. The centre is not usually used, a fixed offset (~10KHz) is advisable.

Eric's suggestion of balancing the values of the integrating capacitors to flatten out the phase response and opamp feedback capacitors to flatten the amplitude might be of interest although that would probably make only a small difference over a signal's bandwidth.