Spurious signals at low gain

Post information or questions regarding SDRplay products here
Post Reply
Fox
Posts: 1
Joined: Wed Oct 07, 2015 3:52 pm

Spurious signals at low gain

Post by Fox » Wed Oct 07, 2015 5:08 pm

Hello,

with a low gain my RSP1 displays spurious signals. They start to appear when gain reduction is more than ~50 dB. I am not sure if I have a faulty RSP1 or this is a normal behaviour. These Signals do not change when Frequency is changed, but chage a lot if gain is being changed. In Zero-IF Mode there is always an Image of the spurious signal visible, sometimes even eight signals at the same time.

Maybe someone can explain what I see or reply if you have the same behaviour.
The Input of the RSP1 was terminated with a 50R SMA-Dummyload for the following screenshots.

Zero-IF Mode:
Image
Zero-IF Mode, gain changed:
Image

Low-IF Mode:
Image

Regards
Fox
Last edited by Fox on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

g1hbe
Posts: 494
Joined: Sat Jan 17, 2015 3:28 pm
Location: Cheshire, UK

Re: Spurious signals at low gain

Post by g1hbe » Wed Oct 07, 2015 9:32 pm

Yes, I see these as well and only at low gain settings. The 'humps' are slightly unstable in frequency as if they might be spurious oscillations or maybee very high harmonics of a clock frequency or something. It would be great to find out what they are!
Last edited by g1hbe on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason
Andy

13dka
Posts: 136
Joined: Sun Feb 14, 2016 8:40 am

Re: Spurious signals at low gain

Post by 13dka » Sun Feb 14, 2016 10:27 am

I'm seeing this too. Kind of reassuring that others are getting those (and the pronounced DC spike) too, on the other hand I'm not sure if that just means "my RSP1 is broken too", or at least suffering from being at the edge of parts tolerances a.k.a. "got a bad one". The lack of feedback on this thread reinforces my doubts, got my RSP1 yesterday and I would like to know if I should have it replaced.

The spurs are wandering in and out when increasing gain reduction, when doing that pretty suddenly I can see them coming from outside the baseband and crossing in the middle over the course of a few seconds. The DC spike generally turns up above 600-700 MHz, below that I can bury it in noise by increasing gain but above 1GHz there"s no cure anymore except switching to low IF mode. If I turn off the post-tuner correction the spike has so much power that the AGC bricks reception entirely, which makes me think something is not quite in order.

The past 24 hours with the RSP1 have been quite a rollercoaster in "user experience" so far, it's not easy to find a middle ground for my expectations between all the enthusiastic reviews, the price and my experience with a standard RTL/R820T2 stick. The latter had absolutely no DC spike or gap and gave me surprisingly little problems with spurious signals and IM products in my (rural) neck of the wood. The RSP1 has much more sensitivity of course but it also has filters and so I did not expect to see images from the FM BC band on 350 MHz. The RSP1 has a variety of means to mitigate all that, but it always comes at a cost, I have to keep gain high to get rid of the DC spike but then I'm getting images, I can reduce baseband width or switch to low IF mode but then I can't have an overview of a large spectrum chunk, at least not without constantly switching stuff in the ExtIO menu.

I don't want to come off like an angry customer with a fit of buyer's remorse though, of course there's also a bright side, none of these problems seem to show up on HF and I did notice the generally improved (over an RTL stick) SNR across the entire reception range, the much lower CPU load even at wide baseband settings and so on, but the DC spike and the image spurs running across the waterfall in opposite direction in some frequency bands are bothering me a little.
Last edited by 13dka on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

13dka
Posts: 136
Joined: Sun Feb 14, 2016 8:40 am

Re: Spurious signals at low gain

Post by 13dka » Sun Feb 14, 2016 11:17 am

To follow up on my post, here's another thing that bothers me - it seems the filter switching may or may not be a bit bugged, or defective in my unit:

In this screencap

https://www.dropbox.com/s/8hh6tpdspbsfz ... t.gif?dl=0

you can see an image of some TETRA station or similar at 251.300 MHz. Then I switch LO frequency from 251.000 to 249.000 and back - the image is gone. Next I switch to 1251.000 MHz to pick it up again. Works every time.
Last edited by 13dka on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

Tech_Support
Posts: 499
Joined: Mon Jun 01, 2015 7:00 pm

Re: Spurious signals at low gain

Post by Tech_Support » Sun Feb 14, 2016 12:00 pm

Dear 13dka,
The noise bumps you see at low gains are not actually a fault. They simply relate to the use of sigma-delta ADCs. These converters use a highly oversampled sigma-delta modulator to 'shape' quantisation noise from lower to higher frequencies. Decimation filtering is then used to increase the resolution at the expense of reduced bandwidth. These converters rely upon having either signals or noise to be large enough to 'toggle' the converter or the sigma-delta modulator will not shape the noise in the desired way. At very low gains if the AC signal (including noise) is not sufficient to overcome the DC offset at the input of the ADCs, then the noise shaping will not work in the correct way. There is no reason for you to have the gain set this low as you do not have strong signals present within the receiver bandwidth.
Regarding the DC centre spike, this is entirely due to the use of Zero IF mode in the tuner. The R820T2 based RTL stick does not use zero IF, it only supports the use of a low IF and as a consequence you will not see a DC offset spike. If you had an E4000 based RTL stick, you would see the same thing as the Elonics E4000 also used zero IF. You can select Low IF mode and this will not give a DC offset, but the maximum bandwidth that can be used in Low IF mode is 1.536 MHz.
You don't say what SDR software you are using, but if you are using HDSDR or SDR#(v 1361 or earlier) in the in the advanced setting within the ExtIO window, you can enable DC offset compensation and this will substantially suppress the DC spike and virtually all cases. The latest versions of SDR console also use the latest features of the API of which, DC offset compensation is just one.
Finally, the observation that you have made in that switching between 249 MHz to 251 MHz causes a spurious TETRA signal to appear is an artefact of the way architecture works. From 60 MHz to 250 MHz and from 420 MHz to 2 GHz, the RSP uses a direct conversion (zero IF) architecture. Form DC to 60 MHz and from 250 MHz to 420 MHz, the RSP uses a block converter to shift the input spectrum to an IF before it is then converted to either a zero IF or a low IF. The use of the block converter can give rise to spurious mixing products, where harmonics of the first LO create higher intermediate frequencies which can in turn mix with harmonics of the synthesizer to produce 'false' in-band signals. In HDSDR or SDR# (version 1361 or earlier), open up the advanced tab and under 'LO frequency plan' deselect 'Auto' and manually pick one of the available IFs. This will generally help move spurious mixing products outside of the receiver bandwidth. If you specifically want to receive a signal at 251 MHz, the easiest thing may simply be to set the LO to 249 MHz and make sure the IF bandwidth is 5 MHz or more and 251 MHz will fall within the bandwidth of the receiver.
The RSP is a highly configurable and flexible receiver with many more modes of operation than a RTL based dongle. it does take a bit of time to get to understand how to get the best out of the device for a particular use case.

As well as this particular community forum, if you are a Facebook user, you might want to consider joining the independent Facebook group for the RSP at:

https://www.facebook.com/groups/sdrplay/?fref=ts

Here you will find a community of over 1800 RSP users who are very knowledgeable and helpful for how to get the best out of the RSP for a particular application. Alternatively, as this forum is a community forum and not specifically a support forum, if you have any particular problems with your RSP, please feel free to contact support@sdrplay.com

Please remember to quote your RSP serial number, OS and application S/W when contacting support, so we are best equipped to help you.

Sincerely

SDRplay Tech_Support
Last edited by Tech_Support on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

13dka
Posts: 136
Joined: Sun Feb 14, 2016 8:40 am

Re: Spurious signals at low gain

Post by 13dka » Sun Feb 14, 2016 2:43 pm

Hello Tech_Support,

Thank you for getting back to me so quick, on a Sunday! Much appreciated!

Thank you also for the explanation of the wandering bumps, for the most part I wanted to make sure my RSP is behaving like any other RSP and not defective. I actually did mention that I'm using HDSDR and SDR# build 1361 - in the first version of my post, which I didn't copy into the clipboard before I tried drag and drop of images onto the edit window () as if I was posting for the first time in a forum. :lol:

As for the TETRA image, it was interesting to read how you solved the gap problem and I already noticed the options to shift the gap around if need be. I'm not quite sure if I pointed out clearly enough that I can switch between 249.000 and 251.000 all day without getting the image and that I have to "pick it up" by tuning to a much higher frequency first (I conveniently used the GHz digit to do that), which appeared odd to me, also given the fact that the offending station is probably around 390 MHz.

However, setting the LO frequency to 168 MHz does help with that (120 and 144 MHz return even more images), and of course low IF mode, which also removes the otherwise "incurable" image of the (probably) same station at 131 MHz.

I think I understood/experienced that the RSP will not have any of these issues and will be as well-behaved and non-fiddly as I hoped for in low IF mode, and that it provides a zero IF option to use a wide spectrum on a more experimental basis, that needs some more consideration by the user. I"d like to mention that a part of my not so great initial encounter with the RSP could have been avoided if I had known about that (and I do try to read a lot about things I'm going to buy), and/or if low IF would be the default setting after install (OK I probably would have changed that anyway :)).

Thank you again, all the best and have a nice Sunday!

Ollie
(weasels off to play more with the RSP :mrgreen:)
Last edited by 13dka on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

F1BJB
Posts: 118
Joined: Wed Jul 15, 2015 4:24 pm
Location: Beauvais France

Re: Spurious signals at low gain

Post by F1BJB » Sun Feb 14, 2016 6:44 pm

Hi thanks for the explanation.
I think it would be very helpful to have a more detailed schematic in the interface.
Knowing what RF filter and local oscillator is used in real time would help understanding the results.
Actually some parameters are hidden and we need to look in the documentation to understand.
A simple box with + or - 120/144/168MHz depending of the centre frequency chosen could be very helpful.
Last edited by F1BJB on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

DaveB
Posts: 142
Joined: Sun Feb 01, 2015 6:04 pm

Re: Spurious signals at low gain

Post by DaveB » Thu Feb 18, 2016 12:29 am

Just to add more to the spurious signal debate - in low IF mode you can get IF images. I can find weak signals from BBC Radio 2 on 85.0 and 85.9 MHz on the SDR Play - and I can also find IF images on the RTL820T2 at twice the sampling rate.

The SDRPlay ones work out as follows. Sampling rate 8.2 MHz - visible bandwidth 1.536 MHZ - the sample has a divide by four decimation so the sampling rate is 2.05. Twice 2.05 = 4.1. So BBC R2 on 90.0 MHz gives an image on 85.9 and BBC R2 on 89.1 gives an image on 85.0 MHz. The low IF image suppression on SDRPlay is around 45dB - and for the same sample rate is around 40dB on the RTL820T2 dongle

Low IF images are for more of a problem to me for FM DXing than the residual zero IF images left after auto correction as apart from being stronger they are at a fixed place in the spectrum.

But I'm still able to achive far, far, more than I ever can with a single frequency analoque tuner!
Last edited by DaveB on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

Post Reply