AM Operation Using Power SDR with Flex and Other SDR Platforms

(1/18) > >>

W1AEX:
There are lots of Flex users and SDR advocates in the AM forum who know how to get under the hood of Power SDR and make it do all kinds of good stuff. Many of these users sound exceptional on AM because they have worked at it and figured out what works and what doesn't. Although I have studied the manual, lurked in the reflectors, accessed the official Knowledge Base, and skimmed the Flex Wiki, much of what I have learned regarding Power SDR and AM has come from discussions with others on the air. While the documentation available to Flex/SDR owners is extensive, people still seem to struggle with some areas of setup and operation. If you have conquered some aspect of Power SDR that might be a problem for a fellow AM operator, consider adding it to this thread to help someone who is starting out with one of the platforms that use Power SDR.

W1AEX:
Audio Gain Distribution Inside PowerSDR:

I have encountered lots of Flex/SDR users on the air who sound great in every voice mode. However, it’s not uncommon to hear stations who seem to have a lot of grit, distortion, and harshness in their audio in all the voice modes. As it turns out, many of these stations are completely unaware of the multi-function TX meter and how to use it to correctly set the audio gain distribution throughout the Power SDR audio chain. It’s not unusual to find that one of the stages is inadvertently running at maximum gain while another stage is set near the minimum to compensate. This can invoke strange events such as ringing in the audio, distortion, and aggressive ALC action with pumping and audio artifacts. I certainly don’t know everything there is to know about squeezing the most out of the audio functions in Power SDR, but I did find that balancing the gain from start to finish has brought very satisfactory results for me and a number of others who were becoming very frustrated. Pages 77 and 78 of the 2.x.x Power SDR manual "sort of" infer the information given below, but I have found that many users are completely unaware of any audio level settings beyond the front panel Mic slider. Hopefully the steps below will help someone to avoid a disappointing result when they start to transmit!

1. Connect the Flex to a 50 ohm dummy load and select one of the voice modes in Power SDR. Make sure that you have a TX profile selected that has your bandwidth set as you deem appropriate for the phone mode you are using. For AM I typically create TX profiles that use 6000 or 4999 as the High setting to produce bandwidths of 12kc or 10kc on AM. The 4999 setting allows you to operate on 7.295 without encountering the “Out-of-Band” error.

2. If you have either the “DX” or “Compander” button selected on the front panel of Power SDR, unselect it now.

3. The “Mic” slider on the Power SDR front panel GUI is scaled from 0 to 70. Place it at the mid-point value of 35.

4. In the upper right corner of the Power SDR interface use the drop-down menu in the TX meter to select "Mic”. This allows you to view the level of the first audio stage of your Flex. Look at the attached picture at the bottom of this post if you are unfamiliar with the TX meter drop-down menu.

5. From the dropdown menus along the top left edge of the Power SDR interface, select “Mixer” and position it on your desktop so you can access it easily. With the Flex connected to a dummy load, key the transceiver and speak into the microphone at the voice level you typically use on the air. Observe the TX meter Mic level and adjust your level up or down with the “Mic” input level slider (or the slider for whichever input you are using) in the Flex Audio Mixer so that the maximum peaks reach no more than –2 dB. I'd like to mention that this is an absolute maximum level that I never reach while operating. I have found that I get the best results if I set this stage up much more conservatively. In fact, my levels at this stage average around -15 dB and rarely ever exceed -10 dB. As someone wisely pointed out to me recently in an email, this not "a cookie cutter" kind of thing and your setup will certainly differ from what someone else comes up with. A little experimentation and adjustment of gain levels at each stage will bring you to a very satisfactory result as long as you don't overdrive one stage and then underdrive another stage to compensate.

6. Now set the TX meter to “EQ”. This allows you to view the level of your second audio stage. From the drop-down menus along the top left of the Power SDR interface, select “Equalizer” and position it on your desktop so that you can access the Transmit Equalizer easily. I would suggest that you enable the 10 band equalizer function at this time, if you have not already done so.

7. Observe the TX meter EQ level as you speak into the microphone and adjust the “Preamp” slider in the EQ interface so that your maximum peaks reach no more than –2 dB. I'd like to mention again that the -2 dB level is an absolute maximum level that you should never exceed. In fact, with my transmit profiles I have found that I get the best results if I set this level so that my average audio rises to around -15 dB with the maximum peaks reaching to around -10 dB as shown in the picture here.

8. Set the TX meter to Leveler. Open up the PSDR tab for DSP and go to the AGC/ALC settings page. As you transmit and speak into the mic, adjust the “Leveler” setting upward or downward from the default setting of 5 but make sure that peaks on the TX meter do not exceed 0 dB. You can further adjust the attack/decay/hang times if the Leveler is not responding quickly enough, or if it seems slow to release. This is largely trial and error, but I found the default decay and hang times to be too long, and reducing them smoothed things so that there was no audible pumping or excessive periods of gain reduction. (Pages 156 - 157 of the PowerSDR 2.x manual provide more details)

9.  Set the TX meter to ALC. Open up the PSDR tab for DSP and go to the AGC/ALC settings page. Apply audio and observe that the maximum level of peaks does not exceed 0 dB. If everything in the preceding stages has been set optimally, you should see that your maximum peaks will end up between -5 dB and -1 dB. You can also adjust the attack/decay/hang times of the ALC if you see a peak sneaking through now and then. Increase the default ALC attack time of 2 ms to a higher value and you will find that the ALC will be restrained from intervening too early. I have adjusted my profiles to the maximum allowed (10 ms) without any negative effects. Again, I found that by reducing the decay and hang times in small steps I could find a setting that resulted in smooth audio without pumping or extended periods of gain reduction. If your ALC levels are still exceeding 0 dB on peaks, use the Leveler to reduce the overall gain of the audio chain, or reduce the “Mic” level by using the slider on the Power SDR front panel GUI. A combination of both settings should allow you to keep the ALC below the 0 dB threshold. The goal here is to find settings that achieve fairly equal gain distribution throughout the software audio chain. The worst possible scenario occurs when one stage is overdriven and a following stage is underdriven in an attempt to compensate.

10.  When you have finished, your audio should sound smooth and clean. If your transmitted frequency response is not to your liking, the 10 band TX equalizer built into Power SDR works very well and is more than capable of making up for deficiencies in whatever microphone you have selected. Always remember that after you EQ your audio, you should go back through the steps to compensate for any changes in gain.

11. When you get all done with your adjustments, don’t forget to save them in a TX profile! Each TX profile saves its own bandwidth filter settings, front panel Mic setting, DX-Compander setting, EQ preamp level with either the 3 or 10 band adjustments, Leveler setting, AGC/ALC attack and decay, and RF transmit level. Additionally, starting with PowerSDR version 2.3.5 the mixer input selection and level settings, DSP TX buffer settings, VOX, DEXP, and VAC1/VAC2 settings are saved within each TX profile.  If you don’t save your settings, they will be instantly lost when you switch TX profiles! To save a profile, from the drop-down menus along the top left edge of the Power SDR interface, select “Setup”. From the available tabs that are displayed, select "Transmit". On the "Transmit" tab, in the upper left side, find the box named "Profiles". Select the "Save" button and this will bring up a dialog box that will allow you to save your new profile with a unique name. I usually select a name that identifies the mode and bandwidth for which the profile was created, but you can use whatever system works best for you.

Note regarding PowerSDR version 2.3.5:  In discussions with several others using this version it has been noted that the MIC and EQ stages seem to behave a bit "hotter" than previous versions. There's no hard data for this but just some general observations that lower settings seem to be working better. I found evidence of tearing on sibilant peaks if I allowed excursions to reach -6 dBm at either stage and it was also more difficult to keep the ALC from exceeding 0 dBm with levels that were not problematic with earlier versions. With my Flex 5000A everything behaves perfectly as long as my peak levels at both stages stay between -15 dBm to -10 dBm. You can make up the gain difference by using the "Leveler" if needed. Obviously, each installation will behave differently, so just keep this in mind if you notice either of the behaviors mentioned.

Summary:  When your levels are set correctly throughout the audio chain, engaging reasonable levels of the Compander or the DX function should result in a very significant loudness difference in your signal, without invoking the ALC or producing any signs of ringing or distortion. It's possible to produce a very decent AM signal using a cheap microphone without any external processing. My "naked" Flex 5000A runs this cheap Radio Shack electret element connected directly to the front panel mic jack:

http://www.w1aex.com/mic/mic.html

At my station, the preamplification, EQ, and compression are handled completely by Power SDR. The MP3 file below of my "naked" Flex on 75 meters was recorded by Jeff W2NBC. It's a very good sample of what a straight "out of the box" Flex sounds like with no external processing. As long as the gain is set correctly at each stage, it should sound smooth and natural. In my opinion, I don't think any other amateur multi-mode rig can match it on AM.

W1AEX:
The AM Carrier Level Setting:

Note:  If you are using PowerSDR versions 2.4.4, 2.5.3, 2.6.4 or 2.7.2 it might be helpful to skim down the page to see the information regarding changes that were made to the AM carrier levels in those versions.

With the Flex/SDR platform, the AM mode is derived completely within the code of PowerSDR. Other than providing RF power at the end of the process, the hardware is not involved in the generation of the AM signal. In a brief description of what goes on, one of the Flex engineers explained to me that they generate a mathematically perfect double sideband signal, run it through the software DSP, and then add a carrier at the end. I don’t know much beyond that, because I’m not a software engineer, but the end result is an AM signal with the potential to produce some interesting asymmetry.

As the software evolved, the authors of PowerSDR saw an opportunity to enhance AM operation by adding a setting called the “AM – Carrier Level” which is located in the upper right corner of the “Transmit” tab in Power SDR. This setting gives you complete control over the ratio of carrier to audio as your signal is generated. If you leave this at the default value of “100” you’ll most likely find when you check your scope that you can almost reach 100% positive peaks fairly consistently and very cleanly. The interesting thing about this setting is that as you reduce its value, your Flex/SDR will begin to develop some impressive positive peaks. With a setting of “50” you will find that your carrier power drops, but your positive peaks will have grown significantly. Unlike plate modulation, you will not spray artifacts up and down the band as you adjust this setting to modulate way beyond 100% positive.

A short video of how the Flex looks on a monitor scope and how it sounds with the AM carrier level set at 50% can be viewed at the link below. The audio that you hear in the video is the Flex as it was received on a Kenwood 590S receiver with a bandwidth setting of 10 kc:  

http://www.youtube.com/watch?v=h2MXgrDIqrA

The downside of this setting is that if you overdo it some receivers that don't have synchronous detection will not take kindly to your signal. I've found that with the way my audio levels are set, AM carrier settings lower than 35% tend to cause my rig to cross the threshold to the point where it starts sounding like double sideband when detected with a receiver employing diode detection. I experimented with levels of 40% and 45% but a few voice peaks now and then would make excursions that didn't sound right. After a lot of critical listening, I found that an AM Carrier Level of 50% sounded smooth, undistorted, and loud with my settings but I suspect that it's best for each user to experiment to find the most effective level for their AM operation.

You may find that it’s really worth playing around with this feature to cleanly achieve a significant loudness factor while operating AM. Stations on the other end will no doubt inform you when you’ve gone too far.

Note regarding PowerSDR version 2.3.5:  This release of PowerSDR corrects an issue that was present in later 2.x.x versions that allowed users to generate a 100 watt carrier while running AM. Obviously, this left no head room for positive peaks so to prevent unknowing operators from generating horrendous signals on AM the engineers at Flex made some necessary changes. The short story is that an AM carrier level setting of "13" seems to correspond fairly closely to a value of "50" with previous 2.x.x versions. I've settled upon a setting of "15" and find that this results in clean and undistorted audio while producing +125% positive peaks without even getting close to the baseline. Comments from other users indicate that they have ended up with similar settings. With values below "10" my signal would occasionally begin to exhibit traces of "sideband sound" on a standard AM receiver.  

W1AEX:
Custom Receive Filters for AM in Power SDR:

As the "Stock Filters" part of the picture below illustrates, many of the the default receive filters for AM and SAM are not really optimal. For example, why would you want a 2.4kc filter for receiving AM or SAM? It’s really easy to configure your own custom receive filters by simply right-clicking on any filter you want to change, selecting Configure, and then entering the desired data. As can be seen in the right side of the picture below, I ended up installing receive filters for 20kc, 18kc, 16kc, 12kc, 10kc, 9kc, 8kc, 7kc, 6kc, and 5kc to satisfy my uses for BCB, SWL, and AM operating. Each mode has its own filter set, so these changes will not have any effect on other modes. Of course, that means you’ll need to create filters for both the AM and SAM modes assuming you use both modes for receiving AM signals.

One other useful thing you might consider doing is to rename the VAR1 and VAR2 filters for the AM and SAM modes to something like AM-L and AM-U. What I do for these is to pre-set them to receive either the lower or upper sideband to allow me to quickly eliminate interference that shows up in only one sideband. Of course you can always do this manually in the panadapter, but I find it convenient to have this function as a single button press pre-set. With the ability to define all your own filters you can pretty much do whatever you like to suit yourself.

For a brief tutorial on how easy it is to make custom filters, take a look at the first minute of W9OY's excellent video at the following link:    

http://www.youtube.com/watch?v=E0tLtxAt24E&feature=mfu_in_order&list=UL

Remember to save your database when you finish so that you can import your custom filters back into any new PSDR versions you upgrade to.

Mike/W8BAC:
Thanks for going to the trouble of pointing out the specific AM adjustments for PowerSDR. I haven't upgraded the software in a few years. I'm still running 1.10.4 and an older M-Audio Delta 44 sound card. The rig is used regularly but time to play is hard to find.

I recently picked up an FA-66 and I'm making some other changes like adding a Fire Wire card and the newest Power SDR software. Your tips will help during setup.

I wonder if your rack mount Power SDR GUI is an easy upgrade and if you'll share that? ;D You need to to add some rack mount washers under those screws.  ::)

Navigation

[0] Message Index

[#] Next page