Using SDR# to drive the SDRplay RSP

[TyTanYGraig]

Hi Folks,
The SDRplay team have put together a very good technical that describes the gain control and AGC concepts very well:

Reading through this, it became clear to me that the SDR# plugin allows the user to programme gain settings that are not valid settings for the tuner.

The tuner LNA has discrete gain steps, which vary depending upon the band that is selected. For example, in VHF Band II, the step is 24 dB. It appears that the mixer gain can only be reduced AFTER the LNA has already been switched to its low gain state. This means that the maximum 'legal' LNA GR Threshold that can be set is 59 dB, which is the maximum IF gain reduction. Attempting to set the LNA GR Threshold higher than this, say 61 dB for example, means that if you try to set the overall gain reduction to 60 dB, it can only do this by attempting to set the IF gain reduction to 60 dB, which is greater than the maximum possible! Similarly, there is a minimum LNA GR Threshold that you can set. If the LNA gain step is 24 dB, then you cannot have an LNA GR Threshold of less than 24 dB. If you were to try to set the threshold to say 20 dB and then set the total gain reduction to 22 dB, then the S/W will try to set the IF gain reduction to -2dB which is also an 'illegal' value.

I have emailed the SDRplay support team and they have confirmed that what I am saying is correct and that they will implement checks in the next version of the SDR# plugin to ensure that you cannot programme 'illegal' values. It seems to me that this might also explain some of the problems that people have had with S/W instability. I have not seen this myself, but I have not tried experimenting with different LNA GR Thresholds. I simply left it at 59 dB and everything seems to work well

[DaveB]

if you try to set the overall gain reduction to 60 dB, it can only do this by attempting to set the IF gain reduction to 60 dB

What about the Mixer gain reduction of 19dB that will kick in? After the LNA has turned off (-24dB) when the IF GR reaches a certain level the mixer gain will drop by 19dB - so 59+19 = 78 dB.
My experiments indicate it is possible to confuse the AGC loop by setting parameters that take it outside the maximumum range that it can handle - but unplugging the aerial - which effectively sets everything back to zero GR - enables it to recover.

My big question for the SDRPlay team is - in the VHF bands is the pre-amplifier (U3) controlled by the AGC loop - or is it always on above 60 MHz? i.e do the internal LNA's n the tuner chip and the pre-amp get turned of when the Gain Reduction threshold f the LNA is reached.

Rgds

David

[TyTanYGraig]

Ooops, here is the ling to the technical note on AGC: http://www.sdrplay.com/docs/SDRplay_AGC ... e_r2p0.pdf

[CaptainNemo]

Hi TyTanYGraig

If you are in Zero IF mode, then 1.5 MHz bandwidth is 750 KHz in I and Q and so a SR of 1.5 MHz does NOT breach the Nyquist criterion.

"750k in I"+ "750k in Q" is not clear to me.
Would you like to say something more? Many thanks in advance!!!
Maybe you can add some comments with reference to rsp schematics and components (pdf in sdrplay site) rather than to "I/Q data model".

bye

[TyTanYGraig]

Hi CaptainNemo,

"750k in I"+ "750k in Q" is not clear to me.
Would you like to say something more? Many thanks in advance!!!
Maybe you can add some comments with reference to rsp schematics and components (pdf in sdrplay site) rather than to "I/Q data model".

bye

In Zero IF mode, the RF signal is mixed down to a 'zero IF', i.e. the LO is positioned to be at the centre of the wanted signal. There are two signal paths I and Q, with a 90 degree phase relationship between one and the other. These I and Q paths represent the complex spectrum where the complete signal comprises I + j.Q which simply means that the signals in I and Q are the baseband cartesian version of the wanted RF signal. Here the filters operate as low pass filters and so a 750 KHz low pass filter in each of the I and Q paths is equivalent to a 1.5 MHz wide bandpass response at RF. As long as the sample rate in both I and Q is at least 1.5 MHz, then the Nyquist criterion is not breached.

The tuner in the RSP can also operate as a heterodyne tuner where the signal RF is mixed to a real IF and the filters are 're-organised' to operate as true bandpass filters. In this case, you don't technically need the signal in both the I and the Q channels, as only one will suffice.

If you look on Page 2 of the schematics, U5 is the tuner and you can see two differential signal ouptuts (I and Q). I you look carefully, you can see that the filters which follow the mixers have both low pass symbols and bandpass symbols, which I believe means that the filters can be configured to operate as either low pass filters (for zero IF mode) or bandpass filters (for IF mode). In IF mode, you must set the SR to be at least as great as the highest corner frequency of the bandpass filter.

I hope that this makes sense

[TyTanYGraig]

Hi Dave B,

My big question for the SDRPlay team is - in the VHF bands is the pre-amplifier (U3) controlled by the AGC loop - or is it always on above 60 MHz? i.e do the internal LNA's n the tuner chip and the pre-amp get turned of when the Gain Reduction threshold f the LNA is reached.

I asked SDR Support the same question and the reply that I got was the the pre-amplifier (U3) is NOT controlled by the AGC loop and is always on for bands above 60 MHz. They say this pre-amp was put in to overcome filter and switch losses and improve sensitivity. Below 60 MHz, atmospheric noise tends to dominate and so it gives no real benefit

All the best

TyTanYGraig

[CaptainNemo]

Hi TyTanYGraig.
Thank you for your comments about zero IF and I/Q etc etc. related to SDRplay implementation. Very useful to me.
Bye

[CaptainNemo]

To the "sdrplay ltd" team. I try to explain (one of) my problem.

In order to visually understand signals on the sdr# fft display, one need to use zoom.
Just to give an idea if you set the zoom so that an ssb ham phone tx take 3 to 5 mm (more or less) in this way you can visually recognize a lot of HF modes. When you have set the zoom in this way you start changing the VFO looking here and there.
In this way sooner or later it may happen to fall downhill I mean you reach the corner of the "filter" and you may tune a signal outside the "flat" bandpass.
I don't know if this is a problem of SDR#, of SDR# plugin or what else...
I will be happy if, while tuning up and down, I'm always in the middle of the bw (like in any receiver with a tuning knob ).
tnx
73
bye

[DaveB]

I mean you reach the corner of the "filter" and you may tune a signal outside the "flat" bandpass.
I don't know if this is a problem of SDR#, of SDR# plugin or what else...

It is a function of the software and it is very irritating . Try using HDSDR instead. It has a LO (Local oscillator) frequency and a Tune frequency.

You can tune using the LO and the display moves with the frequency change. Or you tune with the Tune frequency and it moves up and down the display window until it "falls off the edge". There is more to it than that, as you have options to lock the LO and Tune - with or without an offset.

Of course the ability to shift the cursor across to an interesting frequency because you are displaying a spectrum beats a tuning knob hands down - but I must say that on the ham bands it's like playing "whack-a-mole" ("hey there's a strong signal - lets jump to it - oh it's gone!").
On the FM broadcast band displaying 4.5MHz means you can see weak signals appearing and punch in the exact frequency with direct frequency entry.

As I've posted elsewhere - another reason for trying HDSDR is it allows the Zero IF image that appears on strong signals to be suppressed to the noise level manually. Without this option you won't get far with weal signal dx-ing on FM.

DaveB

[CaptainNemo]

Thank you very much DaveB.
I've been using sdr# from a couple of year with rtl-sdr and funcubePP.
I engaged a "battle" with sdrplay team in order to have sdrplay working like other dongles (I'm talking of "user" operations, look and feel and so on).
I will wait the new plugin (announced for march) then I will raise white flag (means start using HDSDR).

Very amusing (and centered) your joke about "whack-a-mole" (since english is not my language I did some googling...) .

bye