Hello Prog,

Btw, while I think the web listing of running SpyServers is a great start, I do think that this will need to move to a self-publish model.  This is where remote Spyserver stations would notify your website that they are up with some sort of keep alive, say every hour.  Something like PSKReporter ssy.  Without this, this email list will slowly fill up with people asking why X, Y, or Z remote Spyservers aren't working.  That's all noise that I don't think any of us want to see.

Anyway, thanks for all the hard work on the hardware and software that you do!  All quality stuff, with usable interfaces, at reasonable price points that few have met.  I do continue to hope to see SDR# for LInux though!  Maybe it now works under Wine but I haven't tried in a very long time.

--David
KI6ZHD

On Sun, Nov 12, 2017 at 10:38 am, <airspy@...> wrote:

Talking about software goodies I'm playing with the tuning arrows. Nice. But I can't get 'sticky tuning' to do anything.

Is it implemented yet?

I'm assuming - possibly wrongly - that the frequency you are tuned to should stay locked even when you drag the spectrum around

Dave

The sticky and center tuning modes only work when you can change the center frequency. If you are connected to a server with no control and no decimation, you cannot change the center frequency => You can only tune within the available bandwidth with the Free tuning.
If you use a lower bandwidth, you can move the center frequency of your DDC in the server.

Talking about software goodies I'm playing with the tuning arrows. Nice. But I can't get 'sticky tuning' to do anything.

Is it implemented yet?

I'm assuming - possibly wrongly - that the frequency you are tuned to should stay locked even when you drag the spectrum around

Dave

On Sun, Nov 12, 2017 at 10:30 am, John wrote:

But Master - unless the lowly worm looks to the sky how will he ever become a dragon?

Chortle.

John.

Sure, that's how it works. The software goodies will be added on the go, as usual.

On Sun, Nov 12, 2017 at 10:01 am, prog wrote:

Let's first walk before fly. Use the thing nominally, then invent new ways of exploiting the capability of the hardware. It's all about software now.

But Master - unless the lowly worm looks to the sky how will he ever become a dragon?

Chortle.

John.

On Sun, Nov 12, 2017 at 08:58 am, John wrote:

Thanks, the GUI doesn't have to be especially nice. The antenna switch showing up in SDRsharp as a menu option when I (or presumably anyone else) logs into my server remotely would be ideal. Will this be possible?
I imagine plenty of folks setting up their own servers would like to be able to switch between antennas in HF.
I suppose another solution might have been to make it possible for the second RF input on the HF+ (normally self assigning to bands 2 and 3) to have been tasked as a second HF input either by dip switch and/or software. Or perhaps that is possible already somehow.
What do you think?
Thanks,
John

Let's first walk before fly. Use the thing nominally, then invent new ways of exploiting the capability of the hardware. It's all about software now.

On Sun, Nov 12, 2017 at 12:28 am, Airspy US wrote:

A question from our support list reference SpyServer:

"I would like to know how many users can utilize this via the server
software and what the limitations are."

Has anyone tested the limits yet or is ongoing development still refining this?

Our interim answer is that it would depend on the bandwidth streamed, the internet speed, the PC speed, and other factors. Would that be an accurate answer?

(and of course would depend on the preset user limits of course)

Airspy.US

There are many variables involved. Here are the most impacting ones:

1. Input sample rate of the device
2. Processing capacity (CPUs, Cores)
3. Memory bandwidth
4. Network bandwidth
5. Output mode: Full IQ or IF + FFT

A If you setup an Airspy HF+ with a 4th generation Intel Core i7 with 1600 MHz memory and 100 mbps Internet, you can host 300+ clients at the same time with the IF + FFT streaming mode.
IMHO, better point the users to this list and we can help them estimating the proper sizing.

Sounds like a robust solution that will better handle more loading from future additional code

Dave GW4GTE

There isn't an accurate answer without knowing every possible variable. I would suggest that your reply would suffice. A lot more depends of how pedantic the user is. Maybe they should order one and find out.

On Sun, Nov 12, 2017 at 08:23 am, prog wrote:

Note that these GPIO's are not supposed to be populated (they won't fit in the enclosure). They are only for experimenters, and yes, some soldering is involved.
Don't bother about the code for now. It can show up as 4 x buttons in a nice GUI, if needed.

Thanks, the GUI doesn't have to be especially nice. The antenna switch showing up in SDRsharp as a menu option when I (or presumably anyone else) logs into my server remotely would be ideal. Will this be possible?
I imagine plenty of folks setting up their own servers would like to be able to switch between antennas in HF.
I suppose another solution might have been to make it possible for the second RF input on the HF+ (normally self assigning to bands 2 and 3) to have been tasked as a second HF input either by dip switch and/or software. Or perhaps that is possible already somehow.
What do you think?
Thanks,
John

On Sun, Nov 12, 2017 at 08:09 am, John wrote:

Thanks, that seems promising, but you assume far more immediate technical knowledge than I have at present.

What does 'The user mode driver already exposes a function for that task' actually mean?
And how is the code at the end on the link you sent above (L135) deployed?
Also - if we take the lid off the hf+ will we find sockets for the GPIOs or is it a matter of soldering something on?

Thanks,
John

Note that these GPIO's are not supposed to be populated (they won't fit in the enclosure). They are only for experimenters, and yes, some soldering is involved.
Don't bother about the code for now. It can show up as 4 x buttons in a nice GUI, if needed.

On Sun, Nov 12, 2017 at 02:25 am, John wrote:

Folks,
I notice that the tech spec of the HF+ includes '4 x programmable GPIOs'.
My understanding of these things is superficial - but does this mean that the HF+ has on-board GPIOs one of which could be set up to drive a transistor that drives a coaxial relay?
If so then presumably remote antenna switching could be an option from the sdrsharp GUI when linked to a remote HF+ via spy Spyserver.
Or have I ventured into La La Land (again).
Please advise,
John

Youssef replied:
'Your intuition is good. You can use these GPIOs to drive any external device.
https://github.com/airspy/airspyhf/blob/master/libairspyhf/src/airspyhf.h#L135
The user mode driver already exposes a function for that task.'

Thanks, that seems promising, but you assume far more immediate technical knowledge than I have at present.

What does 'The user mode driver already exposes a function for that task' actually mean?
And how is the code at the end on the link you sent above (L135) deployed?
Also - if we take the lid off the hf+ will we find sockets for the GPIOs or is it a matter of soldering something on?

Thanks,
John

So, as confirmed by many friends today, the new Spy Server release brings many improvements that will certainly make a lot of future HF+ owners very happy:

• Smoother streaming
• Lower CPU usage
• Faster response to user commands

The downloads are as usual in the official website: http://airspy.com/download

Your feedback is most welcome!

On Sun, Nov 12, 2017 at 07:50 am, Eric Oyen wrote:

ok,
there are a number of solutions to getting max signal to the SDR.
1. mount at the antenna feed point. this will require remote power feed, a weather sealed box and a remote machine. Believe me, USB can't transport signals to the machine much past 12 feet (too much delay) so having a weather sealed machine close by won't hurt.

2. long run of very low coax (rg-6 works up to 1200 Mhz with fairly low loss). btw, there is a low cost LNA out there that can be powered on a USB connection. I ordered one earlier this year to use with one of my USB RTL-SDR dongles. it has good overload protection and adds only .5 Db insertion noise for 30 Db signal gain from 100 Khz to 5 Ghz. the thing is about the size of 4 postage stamps. this would be a good solution for long runs of coax.

so, unless you are doing anything above 500 Mhz, coax loss isn't going to be much of a factor. Thus, using the onboard LNA just seems like adding extra complication and noise.

DE n7zzt Eric

The third solution is to have a Raspberry Pi or Odroid near the antenna and power them with PoE (Power Over Ethernet). No loss, no latency, and even less CPU usage at the client computer.

ok,
there are a number of solutions to getting max signal to the SDR.
1. mount at the antenna feed point. this will require remote power feed, a weather sealed box and a remote machine. Believe me, USB can't transport signals to the machine much past 12 feet (too much delay) so having a weather sealed machine close by won't hurt.

2. long run of very low coax (rg-6 works up to 1200 Mhz with fairly low loss). btw, there is a low cost LNA out there that can be powered on a USB connection. I ordered one earlier this year to use with one of my USB RTL-SDR dongles. it has good overload protection and adds only .5 Db insertion noise for 30 Db signal gain from 100 Khz to 5 Ghz. the thing is about the size of 4 postage stamps. this would be a good solution for long runs of coax.

so, unless you are doing anything above 500 Mhz, coax loss isn't going to be much of a factor. Thus, using the onboard LNA just seems like adding extra complication and noise.

DE n7zzt Eric

On Sun, Nov 12, 2017 at 02:25 am, John wrote:

Folks,
I notice that the tech spec of the HF+ includes '4 x programmable GPIOs'.
My understanding of these things is superficial - but does this mean that the HF+ has on-board GPIOs one of which could be set up to drive a transistor that drives a coaxial relay?
If so then presumably remote antenna switching could be an option from the sdrsharp GUI when linked to a remote HF+ via spy Spyserver.
Or have I ventured into La La Land (again).
Please advise,
John

Your intuition is good. You can use these GPIOs to drive any external device.
The user mode driver already exposes a function for that task:

https://github.com/airspy/airspyhf/blob/master/libairspyhf/src/airspyhf.h#L135

On Sun, Nov 12, 2017 at 04:17 am, <airspy@...> wrote:

Two 1620 servers connected on local LAN to two instances of SDR# 1620 using previous advised config settings.

No obvious packet drops or other issues

Dave GW4GTE

There is a new trick to force Windows to process the buffers on time without increasing the timing resolution. The impact is lower CPU usage and smooth streaming even at low CPU clocks. The only impact on linux is "zero" CPU usage on idle. The streaming works just as before.

In the other hand, the I/O's are now processed by two different threads for the receive and send operations in a "reactor" pattern. This means the sends won't be interrupted while processing new connections or time consuming user commands (eg. those involving device configuration).

Now I'm waiting for more feedback from Internet users. Simon's server and my own Paris server are already running the latest server 2.0.1620.

I hope you will enjoy it!

On Sun, Nov 12, 2017 at 06:54 am, Alessandro La Farina wrote:

Hello, prog.

I can't find the Airspy HF+ at Wimo.

Wimo should update their website on Monday. ie. tomorrow.