AM Operation Using Power SDR with Flex and Other SDR Platforms

[WB4AIO]

[...]
Transmit bandwidths are variable from 2kc to 10kc for AM and 1kc to 5kc for sideband. In addition, there is native support for VST plugins in the Zeus Software.
[...]

Rob W1AEX

Interoperability is a great idea. And it sounds like they are doing some real creative work there.

But only 5 kHz of transmit audio bandwidth? That's disappointing. There's no reason on God's green Earth that we shouldn't get at least 10. Since getting an SDR, I try to get at least 7.5 kHz if at all possible when using high quality modes.

73,

Kevin, WB4AIO.

[n1eu]

Rob said it has 10KHz AM tx bandwidth, ja?

[W1AEX]

I recently started connecting my external hardware audio chain to my ANAN through a UMC202HD digital interface. I was amazed at the simplicity of this approach and also how clean it sounds while running the AM mode. Many others have been doing this for a long time and I was reluctant to deal with the whole latency problem that I heard others describe but after reading the DAW Setup tutorial by Scott - WU2O it seemed like a good time to give it a shot. Scott's PDF tutorial is hosted at the ANAN Wiki site created by Barry - N1EU and can be downloaded here:

http://anan-100d.wdfiles.com/local--files/home%3Ahome/ASIO_DAW_setup.pdf

Following the recommendations in Scott's tutorial I was able to set up my UMC202HD interface in a few minutes. Starting with conservative values for the streaming mode, VAC bufffer size and VAC latency it was a simple process to move to lower and lower latency values in steps. I quickly reached the point where I can no longer hear any latency at all when I monitor my own transmissions through OpenHPSDR. I have not experienced any audio dropouts, glitches, or anomalies of any kind in any voice mode while running through the digital interface. Details of my modest setup can be seen at the link below:

http://www.w1aex.com/anan/anan.html#digital

The advantages I have observed are that you don't need to fuss with the DB25 input on the back panel of the ANAN, no 1:1 coupling transformer is needed, any RF feedback that you might be dealing with will probably be eliminated, and once you have the interface up and running it opens up a whole new world of voice processing with DAW software and a flood of free VST plug-ins. Scott's tutorial goes into those depths if you wish to pursue that path. I'm just going to keep it simple for now!

Rob W1AEX

[n1eu]

Good stuff Rob!  The down side is that you're subject to the vagaries of Windows computers and hunks of software sometimes not playing nice with each other   

I wholeheartedly endorse getting as far away as possible from that DB25 though!!!

Barry N1EU

[W1AEX]

Barry,

That was one of my fears, but even my 4 year-old computer with the original installation of Windows 7 has not flinched once. Not a single "boing", "twang", "pop" or dropout. From what Scott has said I can push things much further with my settings but I am very content at this point because it is running great and I cannot hear any latency at all.

My latency settings for the UMC202HD interface at this time are:

USB streaming mode:   Low Latency (this can be moved one more notch)
VAC and UMC202HD Buffer size:  1024  (some have run this at 512 or even 256 but I have not gone that low)
VAC latency setting:   30ms  (Scott has indicated that 25ms should not be a problem but I have not tried that yet)

I would imagine that using DAW software and VST plugins would introduce a higher potential for more latency but simply using the USB interface is a very light burden for most Windows 7 and later machines. So far so good!

73,

Rob W1AEX

[W1AEX]

Selectable sideband reception for the SAM mode in OpenHPSDR

I often hear people knocking the OpenHPSDR software because of their assumption that it is "the same old PowerSDR from Flex". Nothing could be further from the truth. While the standard GUI looks similar to the old Flex versions of PowerSDR it is ALL NEW UNDERNEATH that familiar looking GUI. One example is the ability to use selectable sidebands with the synchronous detection mode on AM. Notice that there is no degradation of the recovered audio when listening to only one sideband even though you are listening to half the bandwidth. This is very different from grabbing the filter edge and dragging one sideband in. The 30 second video at the link below demonstrates how you can listen to a weak AM station that is adjacent to a much stronger station that is pushing into one sideband.

https://www.youtube.com/watch?v=mBHnJFsgsZk

The all new DSP and digital audio chain along with pre-distortion linearization for TX only scratches the surface of what OpenHPSDR is capable of. Incidentally, if you do not like the busy "old Flex" look of the GUI you can switch to the collapsed mode GUI which looks nothing like the Flex version of PowerSDR!

Rob W1AEX

[WB4AIO]

I am jealous and can hardly wait until people like me can afford used Anans. I have been lobbying Flex for several years to include selectable-sideband sync detection in PowerSDR to no avail. Now it will never happen, of course, because 1) Flex's PowerSDR quit accommodating my SDR-1000 at version 2.5.3, and 2) Flex's PowerSDR is unlikely to be updated beyond 2.7.2 since all their time and effort is now being devoted to SmartSDR.

Selectable-sideband sync detection is even better than it appears to be because you are cancelling the interference from one sideband, but you are still listening to BOTH sidebands of the AM signal. This gives a 6 dB advantage over "dragging the filter edge over" to remove interference. It was also the method designed into the pioneering tube-type sync detectors built by W2CRR and W3DUQ and a few others. It can do this apparent magic because the Q detector recovers NONE of the AM audio, therefore adding or subtracting the output of the Q detector cancels only interference.

That is the magic of DSB -- and, right now, only Anan (and a few homebrewers) are taking advantage of it.

73,


Kevin, WB4AIO.

[W1AEX]

Kevin,

I really like this feature and use it frequently when running AM in areas of the band where a signal starts up in one our sidebands. Your description of what is going on with this reception mode is "spot on". In one of the ANAN discussion threads the developer of this and many other features in OpenHPSDR (Warren NR0V) offered an explanation of how it works:

...it is a phasing technique.  What this means is that you do not have to be centered on the received signal to take advantage of the sideband selection ... it picks out the correct sideband whether you're centered or not.  I've found this useful in a round-table with stations having vintage gear.  They are often a Khz or two apart in frequency (or is that a Kc or two).

Anyway, I have some vintage gear here too; so, I'm sensitive to the issue!

Another interesting feature specifically for AM reception as seen in the attached image is the "fade leveler" which I use quite a bit while doing AM broadcast DX-ing to overcome flutter. It works very nicely but can cause a very long recovery time if you use it in a QSO with a very strong station. You can also control the squelch tail length for AM if you wish to use the squelch. I do use that when I leave a receiver on 50.4 MHz or even 29.0 MHz to catch an opening now and then.

73,

Rob W1AEX

[WB4AIO]

Very nice, Rob.

I am not sure what exactly the Fade Leveler does -- how is it different from normal AGC? Does it somehow try to compensate for the phase shifting inherent in the cacophony of mixed carriers on a typical AM broadcast channel? Or is it just a gain thing?

Is there a url where I can read more about the development of this detection system -- such as the one you quoted from?

One thing they ought to change, though: the sideband select for sync mode ought to be on the front panel, not in a menu. It's the sort of thing one might need to change rapidly if an interference situation suddenly develops.

Thanks and 73,

Kevin, WB4AIO.

[W1AEX]

Yeah Ralph, I got lucky and heard the squelch break on my 10 meter bandwatch receiver. It didn't last long though! You've definitely got good eyes young man!

Kevin, the attached PDF file written by Warren NR0V is somewhat dated in some areas but the section on page 4 regarding selectable sideband synchronous AM and the fade leveler are current. Warren's explanation from page 4 of how the Fade Leveler functions is pasted in below:

"For AM Demodulation & SAM Demodulation, the “Fade Leveler” function will often reduce
fading and improve the audio for AM. It operates by removing the fading carrier and replacing it
with a much more stable carrier. Note that if this function is used with the Automatic Notch
Filter, you may wish to select the “Post‐AGC” option for the ANF. Otherwise, the ANF will notch
out the carrier before the AGC and therefore the carrier will not have the intended quieting and
stabilizing effect."

Ignore the obsolete sections on ALC and also the section on noise blankers. There is now a look-ahead algorithm that makes it impossible to drive the ALC past 0 dB so you can drive the snot out of the audio and it will not distort or degrade the IMD. Also, there are now 3 noise blankers (2 different wideband noise blankers and a spectral noise blanker that was just introduced this week).

There's a fair amount of used ANAN hardware around Kevin as guys work their way up the hardware chain. If you are not already a member I would suggest joining the Apache Labs user group at the Yahoo forums as people list stuff there frequently.

https://groups.yahoo.com/neo/groups/apache-labs/info

If you are interested, you can also download the user manuals for all the ANAN hardware at the Apache Labs download link below:

https://apache-labs.com/al-downloads/1007/ANAN-User-Guides.html

73,

Rob W1AEX

[WD5JKO]

   Some big reverberations in the Flex Power SDR 2.7.2 world are emerging as Darrin, KE9NS is wading through the source code and has made several interesting and useful enhancements for Flex 1500,3000,5000 owners. There seem to be several updates a week where the Powersdr.exe file is already compiled, and easily downloaded, and then switched out with the original file.

One caution is to backup or rename the original Powersdr.exe file. :-)

http://ke9ns.com/flexpage.html

Revision History:
http://ke9ns.com/flexpage.html#GPLREV

Attached the Water Fall Call sign. It is a little blurred, but that is my call as I print it...


Jim
Wd5JKO

[W1AEX]

That's great news Jim and it will definitely enhance the fun factor for the 1000/1500/3000/5000 owners beyond anything anyone might have expected! I love the analog meter!

73,

Rob W1AEX

[WD5JKO]

Look at the last release that came in last night:

"02/10/16 N6: Fix: Increase TX Bandwidth on Flex1500 and Flex3000 to 10khz wide.
Fix: Improved Waterfall ID 24bit bitmap transmit. Automatically reduced the TX High filter down to 2.6khz during transmit, then back to its prior value. Still learning Bitmaps, and PCM and the conversion between them."

The bitmap waterfall of an owl worked perfect on my F3K. On AM the individual side band images are of course mirror images of each other...

Jim
Wd5JKO

[W1AEX]

Jim,

The bandwidth limitation for the 1500/3000 series has always been a sore point for a lot of Flex users. Those who use them on AM are going to be celebrating with that news!

Rob

[WD5JKO]

Jim,

The bandwidth limitation for the 1500/3000 series has always been a sore point for a lot of Flex users. Those who use them on AM are going to be celebrating with that news!

Rob

   Rob, I tried things out last night set to 6 Khz max audio. Tuning around the Ghetto on 75m, I would have to decide on whether I would cover up 4 QSO's or 2 with my bandwidth selection.    Actually the original Flex 4.5 Khz upper end is not too restrictive, especially since my ears don't go up like they used to.

   Looks like Darrin KE9NS has another update since I downloaded last night....

   I wonder what the Flex folks are thinking about this?

Jim
Wd5JKO

[Pete, WA2CWA]

   Looks like Darrin KE9NS has another update since I downloaded last night....

   I wonder what the Flex folks are thinking about this?

Jim
Wd5JKO

Maybe one of the reasons they just instituted a buy-back/trade-in program for all the old Flex SDR rigs to apply to a purchase of a Flex 6000 Series radio.

[K5IIA]

I would guess this will not work with the anan, but does anyone know of a way to print out a call sign on the waterfall with the anan, or will this actualy work. I just figure if I replace the file it would wipe out all the features of the openhpsdr that I like so much, and the simple little print out a callsign thing that I would probably never use on the air would not be worth it.  I think it would be neat to do a waterfall call sign atleast once though.

73
Brandon


hope to hear you on the air jim, I have not heard you on much like I used to.

[WD5JKO]

  I was on 15m AM today trying out the Flex 3K with 6 khz +/- audio. Also sent a waterfall image on AM that was just "JKO | OKJ" to Robert W0VMC, and it came through!

Big weekend for Flex SW updates from KE9NS:

02/13/16 O6: ADD: Flex3000 Now has 192k Sample Rate (you can view around 170khz on the screen at 1 time), in addition to the 10khz wide TX (edit from JKO: 20Khz AM!)
02/13/16 O5: Fix: TX WaterID Now runs on its own background Thread (see video).Creates file when you change the Callsign text field.
02/12/16 O4: Fix: TX WaterID much faster to create image if running 96k or 192k Sample rates. Fix: Flex Audioplayer has a automatic resampler, but the code to make it work had some missing pieces, fixed the missing pieces, so now 48k waterfall resamples for 96k and 192k SR.
02/12/16 O3: Fix: TX WaterID labels: Creating->Transmit-Tx WaterID. Add Grayscale to database. Database name now: ke9ns1.dat
02/11/16 O2: ADD: Database to save mod settings: Callsign ID, Wide waterfall, etc.
02/11/16 O1: Dttsp compiled to SDK 8.1. and PowerSDR compiled under NET 4.5.2. PowerSDR.exe file slightly larger because Cdecl requires each function to have its own cleanup code. The Downside is that you must also copy over a "PowerSDR.exe.config" file becuase of the old FLEX1500USB.dll. NET 4.5 is supposed to be an improvement over the NET 3.5 PowerSDR was using.

Jim
Wd5JKO

[WD5JKO]

Steady progress in PowerSDR. We can now monitor ourselves when on AM...

Release notes from Darrin Ke9NS:

http://ke9ns.com/flexpage.html


02/23/16 P5: Add: MON toggle between MONpr (Pre-Processed monitor) and MONps (Post-Processed monitor) for 3000 and 5000.

02/22/16 P4: Add: MON (pre-processed audio) to AM/FM for Flex1500 only in VAC mode though (not through the PHONES jack on the 1500).

02/20/16: P3: Add: MON (but pre-processed audio instead of post-prosessed) to AM/FM modes for Flex3000 and Flex5000.
Add: turn off DoScope() if your not actually using it to save cpu cycles

The MON is normally turned OFF for AM/FM modes. I enabled it for AM, but the MON output in the speakers, headphones, VAC is modulated AM (to the edge of the passband), this is because MON wants to use post processed audio. I will need to use pre-processed audio for AM/FM mode, so its not the audio you are transmitting, but at least the MON function works now.

02/18/16 P2: Add: Auto Waterfall Transmit Low level adjustment. Key the radio and Click "Auto Wtr Level" to set Waterfall for transmit input level.
Fix: Waterfall only, not displaying correctly due to addition of avgP, avgB previously. Auto Wtr Level turns off if you just in panadapter mode.

02/18/16 P1: Add: Auto Waterfall Low Level Adjustment. Just click the "Auto Wtr Level" button below the display area.

02/17/16 O9: Fix: Found mistake in Source code when compiled under VS2015 caused the check for atu.dll to fail, so Flex3000 units were using OLD ANT Tuning algorithm. Also Dttsp.dll now is updated to current 2015 v140 toolset, hopefully no issues. I embeded vc_redist_x86 into the DttSP.dll.

02/16/16: Figured out how to compile DttSP.dll under v140 platform toolset. Dumpbin showed v140 requires vc_redist.x86 since DttSP is a 32bit C application. Recompile DttSP.dll under C code generate runtime /MT (instead of /MD) will include all required DLL's

02/15/16 O8: Fix: Add toolTips to TXWaterID and Callsign box. Save the AvgP or AvgB settings in database.
Right Click on Callsign Box to open up Folder where you put your own bitmap files to send to the waterfall.
Right Click on the REC button to open up Folder where you will find SDRQuickAudio.wav files from your REC / PLAY use
Right Click on the your current skin (Setup->Appearance->General->Skins->current skin) to open the Folder to the SKINS
.
02/14/16 O7: Fix: Prevent (if you accidently hit the TX Waterfall ID with the Radio OFF) keying the Transmitter when the Radio is OFF (sitting idle)

[n1eu]

Is there zero latency with the P5 pre-processed audio monitor?  In other words, you hear no delay between talking in the mic and hearing yourself in the headphones?

Thanks & 73,
Barry N1EU

[WD5JKO]

Barry,

   I was playing with that last night using P5 release. With my F3K, for AM MON, it is pre-processed, and there is no delay that I can tell. For SSB, the MON button has the option of pre or post processing. I can hear some delay on post processing, but that depends on the sample rate and buffer size.

   I was experimenting with my AM saved profile settings on SSB Mon where I can tell a lot from pre/post processing. Then I can save it for AM.

I just noticed P6 is out..

Jim
Wd5JKO

[WD5JKO]

Many more enhancements to PowerSdr 2.7.2 since my last post.

All the SW broadcast bands are added under the 'GEN" button. Also many improvements to the audio recording and playback. This has always been a problem for me with AM playback since the carrier level would come on too strong. Lowering the Wave Gain would fix the carrier pumping, but would also drop the modulation percentage. With the Q3 release, these issues seem much better.

http://ke9ns.com/flexpage.html
http://ke9ns.com/flexpage.html#GPLREV

Jim
Wd5JKO

[WD5JKO]

Power SDR has seen steady progress with about 5 updates a week thanks to Darrin, KE9NS. The latest stuff has to do with SWL and DX spotters indicated on the pan adapter. Do a cntrl right click on the DX spot, and it takes you to the fellows QRZ page. Do the same with SWL spots and it takes you to Google search with the search field filled in. Then as of today we have a world map background along with the options of adding gray line and sun positions that track over time.

I attach some pictures of both DX and SWL spots. The center picture shows a big Ham AM signal on 7290...

Some DX Spotter sources:

spider.ham-radio-deluxe.com   port 8000 (must have an account)
k1rfi.com    port 7300
ve7cc.net    port 23

Video highlighting S5 by Darrin, KE9NS:
https://www.youtube.com/embed/3InG8QgMTIY

Jim
Wd5JKO

[WD5JKO]

There have been many updates since my last post. Darrin KE9NS has been very busy over the summer. One area where there have been significant advancements has been towards the short wave broadcast SWLer.

I include a screen shot of the T9 release. Notice the Desktop Map with active grayline, and position of the Sun with the Solar Flux index.

The screen shot is from my Flex 3000.

http://www.ke9ns.com/flexpage.html

Jim
Wd5JKO

[Pete, WA2CWA]

Jim:
From your attached picture, what's the lower right meter used for in your Flex-3000?