AM stations all over the band

[Ramakrishnan Muthukrishnan]

Hello,

I have been using my Radioberry intermittently for the past several weeks. I took it out again after a gap of a month or so and noticed that at zero hertz there is a strong chinese AM station being heard. On all the bands, I see these AM stations, every 10 khz or so. I wonder what is going on. Has anyone seen this issue? I checked with my regular radio (an HPSDR based radio) and don't see similar behaviour there. Any ideas? I tried both dev and stable releases of radioberry. Both of them have this issue. Any hints or help in debugging the issue appreciated.

73
--
Ram VU3RDD

[rent...@comcast.net]

Hi Ram,

I'd say this is due to the often encountered lack of adequate front end filtering on SDR's.  The Hermes Lite 2 (which the Radioberry s based on) uses a HPF on it's LPF board to help mitigate this issue.

73,

Robert, WA2T

[Ramakrishnan Muthukrishnan]

Hello Robert,

Agreed, that was my first thought too, but interestingly I never encountered this issue so far. And the AM station I am hearing strongly do not seem to be the local AM station, it is a Chinese station which is probably many kilometers away.  That said, it seem like worth trying, I will homebrew something quickly and try it out. Thanks!

73

Ram VU3RDD

--

You received this message because you are subscribed to the Google Groups "Radioberry" group.

To unsubscribe from this group and stop receiving emails from it, send an email to radioberry+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/14990a80-e3dc-42cd-be21-5c1d82b7d868n%40googlegroups.com.

--

Ramakrishnan

[softerh...@gmail.com]

Hi Ram,

What is your LNA set to? Between +15 and and +20? Do you see ADC clipping? Setting the LNA gain too high can result in problems like this.

73,

Steve

kf7o

[Ramakrishnan Muthukrishnan]

Hello Steve,

My LNA is set to +14. I modified the radioberry "firmware" (the code
that runs in Pi which listens to network connections from the SDR
software and passes the packets to/from the hardware) to print ADC
overload bit in the protocol (24th bit of the 32-bit value when address
is 0x00).

When I move the VFO to zero hz, I get ADC overload continuously. But not
when I am on other bands. I keep seeing the local AM station
(broadcasting on 612khz) across the band (say in 40m band) at every
5khz. I also see some chinese stations. At 0Hz, it is the local AM
station which is the strongest.

I tried moving RF gain to even lower values, but didn't get much
results. I guess by the time samples entered into the ADC, damage is
already done to the samples!

73
Ram VU3RDD

On 22/12/2021 10:17, softerh...@gmail.com wrote:
> Hi Ram,
>
> What is your LNA set to? Between +15 and and +20? Do you see ADC
> clipping? Setting the LNA gain too high can result in problems like this.
>
> 73,
>
> Steve
> kf7o
>
> On Monday, December 20, 2021 at 7:28:11 AM UTC-8 r...@rkrishnan.org wrote:
>

> __

>> <https://groups.google.com/d/msgid/radioberry/14990a80-e3dc-42cd-be21-5c1d82b7d868n%40googlegroups.com?utm_medium=email&utm_source=footer>.
>
> --
> Ramakrishnan

>
> --
> You received this message because you are subscribed to the Google
> Groups "Radioberry" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to radioberry+...@googlegroups.com

> <mailto:radioberry+...@googlegroups.com>.

> To view this discussion on the web visit

> https://groups.google.com/d/msgid/radioberry/9eae37d4-d49b-4d95-be9a-f5eb37af6055n%40googlegroups.com
> <https://groups.google.com/d/msgid/radioberry/9eae37d4-d49b-4d95-be9a-f5eb37af6055n%40googlegroups.com?utm_medium=email&utm_source=footer>.


--
Ramakrishnan

[pa3gsb]

Hi Ram

Iam not getting completely clear what you are doing/seeing?

Does the problem occur only after getting down to 0Hz? and when returning to 40m the AM broadcast interference is there.

After a cold start and starting directly on 40m the problem is not there?

please let us known

73 Johan

PA3GSB

Op donderdag 23 december 2021 om 11:18:55 UTC+1 schreef r...@rkrishnan.org:

[Ramakrishnan Muthukrishnan]

Sorry Johan for not being clear.

On a cold start, on say 40m, I see AM carrier spikes every 5 khz. I
first thought they are RFI (I already have RFI from a neighbouring house
at every 10khz). But then I noticed a fairly strong AM station within
the 7MHz HAM band (7.125MHz I think) which is unusual.

And then I tuned around to lower bands and I see these AM stations all
over the place, particularly in lower bands but also seen even in 20m.

At 0Hz, I see a very strong AM station (the same station, either the
local BC one or a Chinese station, or both mixed up). It is very very
strong. If I print the ADC overload bit, as I tune down to 0 hz, I see
it turned on. If I tune off 0Hz, ADC overload coming in as 0.

I tried both dev release and stable release. Both have the problem.. May
be it has something to do with my hardware? It is a Beta3 RadioBerry
board with the fpga pin mod installed and was working just fine until I
tested it last month. I don't have filters or a PA installed yet. I have
been using it for listening while I do the work on my computer and often
connect to it from pihpsdr running on my PC or via Quisk.

Please let me know your thoughts and any suspicions on what could
possibly be wrong. Thanks!

73
Ram VU3RDD

> <https://groups.google.com/d/msgid/radioberry/14990a80-e3dc-42cd-be21-5c1d82b7d868n%40googlegroups.com?utm_medium=email&utm_source=footer

> <https://groups.google.com/d/msgid/radioberry/14990a80-e3dc-42cd-be21-5c1d82b7d868n%40googlegroups.com?utm_medium=email&utm_source=footer>>.
>
> >
> > --
> > Ramakrishnan
> >
> > --
> > You received this message because you are subscribed to the Google
> > Groups "Radioberry" group.
> > To unsubscribe from this group and stop receiving emails from it,
> send
> > an email to radioberry+...@googlegroups.com
> > <mailto:radioberry+...@googlegroups.com>.
> > To view this discussion on the web visit
> >
> https://groups.google.com/d/msgid/radioberry/9eae37d4-d49b-4d95-be9a-f5eb37af6055n%40googlegroups.com
> <https://groups.google.com/d/msgid/radioberry/9eae37d4-d49b-4d95-be9a-f5eb37af6055n%40googlegroups.com>
>
> >

> <https://groups.google.com/d/msgid/radioberry/9eae37d4-d49b-4d95-be9a-f5eb37af6055n%40googlegroups.com?utm_medium=email&utm_source=footer

> <https://groups.google.com/d/msgid/radioberry/9eae37d4-d49b-4d95-be9a-f5eb37af6055n%40googlegroups.com?utm_medium=email&utm_source=footer>>.
>
>
>
> --
> Ramakrishnan
>

> --
> You received this message because you are subscribed to the Google
> Groups "Radioberry" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to radioberry+...@googlegroups.com
> <mailto:radioberry+...@googlegroups.com>.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/radioberry/ead1ec7d-54a2-4d30-909d-7cf3ed7fbf5fn%40googlegroups.com
> <https://groups.google.com/d/msgid/radioberry/ead1ec7d-54a2-4d30-909d-7cf3ed7fbf5fn%40googlegroups.com?utm_medium=email&utm_source=footer>.


--
Ramakrishnan

[pa3gsb]

Hi Ram,

You mentioned that the RF gain had no influence.; this does not seem right.

The RF gain is controlled via a SPI interface between the FPGA and the AD9866 frontend.

This SPI interface is also used for controlling the clock for FPGA based on the 38.4 clock connected to the AD9866.

Did you check that you are actually tuned to 40m??? and is the audio of the received stations ok?

When you go to transmit mode... is the timing of sending the samples within 1 second (start the firmware in a command window)... giving us an idea that the

clock is initialized OK; by using the internal (as above described) SPI bus.

If the clock signals are ok we need to see what filtering is doing; a simple resistor att or a more freq dependent filter...

Iam curious what the problem is because you mentioned it was working before...

Hope this helps to get the cause of the problem.

73 Johan

PA3GSB

Op donderdag 23 december 2021 om 15:14:14 UTC+1 schreef r...@rkrishnan.org:

[Ramakrishnan Muthukrishnan]

Reviving this old thread.

I added a Broadcast band high pass filter in the RX path today. The
filter was designed with the software called Elsie and homebrewed on a
piece of copper clad board and tested with nanoVNA. The nanovna
responses agree totally with the designed values. The -3db cutoff is at
around 1.4MHz.

After inserting the filter in the RX path, the situation is way better
than it was in the past. I can see the difference with and without the
filter. The local broadcast station is -45dB down from the cutoff point.

I will have more to say only after more testing on the air.

73
Ram VU3RDD

On 20/12/2021 17:33, John Smith wrote:
> Hi Ram,
>
> I'd say this is due to the often encountered lack of adequate front end
> filtering on SDR's.  The Hermes Lite 2 (which the Radioberry s based on)
> uses a HPF on it's LPF board to help mitigate this issue.
>
> 73,
>
> Robert, WA2T
>
> On Sunday, December 19, 2021 at 10:34:30 AM UTC-5 r...@rkrishnan.org wrote:
>

[Ramakrishnan Muthukrishnan]

Did more testing after sending this email. The local AM station no longer appeared during my testing this morning on 40m band. Instead I hear chinese stations, perhaps from somewhere above 7.300 MHz, the intermod products are not too strong, but enough to be annoying when heard along with another weak CW or SSB station and they appear every 10khz or so.

73
Ram VU3RDD

> --
> You received this message because you are subscribed to the Google
> Groups "Radioberry" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to radioberry+...@googlegroups.com.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/radioberry/7330b595-12ee-01f3-1fb2-3a7a0b2acb71%40rkrishnan.org.

--
Ramakrishnan

[Ramakrishnan Muthukrishnan]

I tried another radioberry board and the problem doesn't exist there.

Two observations:

1. The RF gain on the problematic board is higher than the new board I
tried. With same settings on pihpsdr, when I put in the new board, the
board was very quiet.

2. On the problematic board, the chinese AM station is heard on all
frequencies regardless of the VFO freq on 40m and below. In other words,
nothing changes when I change the VFO. The pihpsdr mode is AM. On 20m it
was feeble.

Are there anything in hardware that I can measure and verify? Any help
appreciated.

73
Ram VU3RDD

[Radioberry]

Ram,

There seems to be an imbalance in the frontend. Feeding the signal to both input pins of the AD9866 is important... is the signal path to both pins ok?

73 Johan

PA3GSB

Op woensdag 9 februari 2022 om 11:27:58 UTC+1 schreef r...@rkrishnan.org:

[Paulh002]

Hi Ram,

Do you still have this issue?  I also experience something like this, but i my case I think it is a software issue.

I am still checking this, but when a very strong AM station is in range I experience interference over the whole bandwidth (half sample freq).

When change the frequency of the radioberry the issue is gone. 

73 Paul PA0PHH

Op zondag 19 december 2021 om 16:34:30 UTC+1 schreef r...@rkrishnan.org:

[Gopakumar K]

I also had the same issue, I just put on 10dB attenuator on Rx line and calibrated the S meter in the software . After that there is no AM stations all over the band.

Thanks

Gopan VU2XTO

[Paulh002]

Good idea, I think I will build an I2C attenuator for this purpose.  But I also made a mistake assuming radioberry iq data is 12 bits, that is not the case it is 24 bit. So I have to change that.

Op wo 16 mrt. 2022 om 03:35 schreef Gopakumar K <vu2...@gmail.com>:

--
You received this message because you are subscribed to a topic in the Google Groups "Radioberry" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/radioberry/VttUKDYnMQI/unsubscribe.
To unsubscribe from this group and all its topics, send an email to radioberry+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/76151214-26a7-40cc-bf04-f169784e0697n%40googlegroups.com.

[Radioberry]

Paul

last week i saw that also ( wanted to add support soapy for the radioberry juice board) .... did not find the time to react 

getting I and Q containing a total of 6 bytes.

for (i = 0; i < nr_samples; i += 6) {

left_sample = (int)((signed char) rx_buffer[i]) << 16;

left_sample |= (int)((((unsigned char)rx_buffer[i + 1]) << 8) & 0xFF00);

left_sample |= (int)((unsigned char)rx_buffer[i + 2] & 0xFF);

right_sample = (int)((signed char)rx_buffer[i + 3]) << 16;

right_sample |= (int)((((unsigned char)rx_buffer[i + 4]) << 8) & 0xFF00);

right_sample |= (int)((unsigned char)rx_buffer[i + 5] & 0xFF);

right_sample = right_sample * -1;



target_buffer[iq++] = (float)left_sample / 2048.0; // 12 bit sample

target_buffer[iq++] = (float)right_sample / 2048.0; // 12 bit sample

//printf("nr_samples %d sample: %d %d \n", nr_samples, left_sample, right_sample);

}

Please provide a pull request; i will do the merge.

73 Johan

PA3GSB

Op woensdag 16 maart 2022 om 10:29:12 UTC+1 schreef paul.hol...@gmail.com:

[Paulh002]

Yes, will issue pull request when fixed.

Thanks,

Paul

Op wo 16 mrt. 2022 om 11:18 schreef Radioberry <radio...@googlegroups.com>:

To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/67eacda6-48f8-42f1-9fed-a2a9f0a89730n%40googlegroups.com.