AM Operation Using Power SDR with Flex and Other SDR Platforms

[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? You need to to add some rack mount washers under those screws. 

[W1AEX]

Hi Mike,

Yah, I forgot to put the washers in there and have paid the price with scratches all over the virtual panel. Thankfully, Photoshop makes it easy to fix!

Lots of SDR-1000 users have tried out the 2.0.22 final release package and many like it. The new software takes some getting used to, and several settings, such as the AGC-T behave differently from the 1.x versions, but there are many enhancements. The biggest change is that they moved the ALC module to the output side of the DSP module. This eliminated issues that many users had with overshoot when driving solid-state amplifiers. If you like to play with Photoshop, you can modify the GUI interface with the 2.x versions and that adds to the fun.

I believe the guys with the SDR-1000 series can float back and forth between the 2.x and 1.x versions without any changes in firmware, so it might be worth looking at the new software to see if it suits you.

Good luck with the FA-66!

Rob

[W1AEX]

Creating Your Own "Skins" for PowerSDR:

While this is not exclusively related to AM operation, a number of smug AM Flex users have asked me about how to create custom skins like the one shown in the image above. If you are interested in messing around with this, a brief tutorial along with some custom starter graphics packs are available at this link:

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

In addition to being able to design your own skins with the custom templates you can download pre-made custom skins. A few vintage examples, such as the Black Wrinkle, National NC-155, and the Collins Retro 32V skins are shown below. Mike - W8BAC collaborated on the Collins skin with some excellent photography and color setting advice. You can take a YouTube video tour of about a dozen custom skins that are available by visiting the link below. What do you make of the station that the Flex is tuned to? Is this an updated 2012 version of a shortwave numbers station with overly processed audio?

http://www.youtube.com/watch?v=U0bv_WYMck4

Have fun!

Rob W1AEX

[WA1GFZ]

Rob,
I just bought a second receiver module for HPSDR and doing diversity/ beam steering in software. it is very cool. I lost all my audio settings for TX so need to reload at some point. I'm actually considering a third module once the software guys proves his software works with three.

[W1AEX]

Frank,

The second receiver module coupled with the diversity capabilities makes things very interesting. I have no doubt that function will amuse you for a very long time! While you are moving forward and making things more complicated, I'm continuing down the path toward making things simpler at this place. I only have to throw about 15 switches to get on the air instead of 30 or 40. Yah... that's much better!

Flex just released the beta version of PowerSDR 2.2.2 Tuesday night. It appears to be rock solid, and they have a new tunable notch filter to go along with the auto-notch filter. The new TNF is awesome for use on AM as you can lock it to a carrier that's any distance from your operating frequency, adjust the width of the notch to whatever you want, and adjust the depth to the point where an S-9 carrier simply disappears. It's perfect for when propagation causes our QSO's here in the northeast to collide with the mid-west guys. It also will completely notch out SSB monkey chatter that shows up in either sideband, without having to remove the whole upper or lower sideband. You can make a permanent notch that stays active on any assigned frequency if you wish. That takes care of the annoying spurious crud from someone's stupid baby monitor that shows up here close to the beacon frequencies on 6 meters.

So many toys, so little time!

Good luck with the diversity/beam steering setup!

Rob

[W1AEX]

Notes regarding AM Operation with PowerSDR version "2.2.3 Final Release"

This release of PowerSDR includes the long awaited fix for the AGC problem that plagued AM users with all versions of 2.x that were released previously. At some point during the end of the 1.x versions, Flex altered the way that the AGC responded to AM signals. Instead of responding appropriately to the carrier of a signal, the AGC responded instead to the energy in the sidebands. This caused the receiver to detect AM signals in such a way that they sounded horribly over-compressed. By riding the AGC-T control, and using a very slow AGC repsonse time, users could sort of compensate when strong signals were present, but when a weaker station showed up it became necessary to constantly fuss with the AGC-T levels. Fortunately, with the 2.2.3 release Flex has fixed this so that the AGC now responds appropriately to the carrier of an AM signal and ignores the sidebands. The AGC-T level now behaves very much like an RF gain control, strong and weak AM signals sound fine together, and you can pretty much set it and forget it.

This is an important note regarding the potentially disastrous behavior of the Drive control in versions 2.2.2 beta and 2.2.3 final release while in the AM mode. Previously, the maximum carrier power with the drive control set to 100 for both AM and FM was roughly 35 - 40 watts. An AM carrier setting of 50% pretty much guaranteed that you could not inadvertently exceed a carrier power of 20 watts even with the drive control set to 100. This left plenty of head room for modulation, and made it possible to drive an amplifier without too much worry about over-driving things. When Flex completely re-wrote the FM module to add new features and to correct issues with that mode in the 2.x versions of PowerSDR, the behavior of the drive control changed radically for both AM and FM. At this time, if the AM carrier control is set to 100% and the drive control is set to 100, the Flex will produce a carrier at full output (+100 watts). This would result in a horrendous AM signal for the obvious reason that there would be zero headroom for modulation, but more importantly, this could be disastrous for those with amplifiers that only require a small amount of drive power. Keep this in mind when upgrading to either of these versions when creating TX profiles for AM and FM! Once my amplifier (2 x 3-500's) is tuned and loaded, I typically run with my AM carrier setting at 40% and my drive level set between 15 and 30 to run anywhere from 150 watts to 250 watts of carrier power depending on conditions. When modulating heavily, I see peak power reaching between 1000 - 1500 watts regularly and the scope reports a healthy looking signal.

Rob W1AEX

[W1AEX]

Regarding the "Limit Slew" setting in PowerSDR 2.2.3

In previous versions of PowerSDR some users have encountered issues when driving an amplifier with their Flex 5000/3000 hardware. Certain amplifiers have randomly tripped when keyed, and sometimes when unkeyed. The Acom and Tokyo High Power amplifiers seem to be particularly sensitive to this although users with grid-driven amplifiers have also reported the issue. Flex studied the problem and even re-wrote the code in PowerSDR to alter the position of the ALC module in the TX chain to fix this. They did seem to resolve the issue when the amplifier was being keyed, but the unkey problem still remained. After much experimentation they appear to have corrected the issue. Although I haven't seen any sign of the "Limit Slew" setting in the release notes, it appeared without a lot of fanfare in the 2.2.3 version of PowerSDR. On the "Transmit" tab underneath the "AM Carrier Setting" you will find the toggle box for this setting. If you hover your mouse cursor over the toggle box a brief explanation of the function of this setting will appear, as seen in the image below.

A friend with an Acom amplifier was on the verge of selling his Flex 5000A due to the "tripping" issue. He employed every trick he could think of to overcome the problem but was unable to resolve it. The amplifier manufacturer acknowledged the problem but offered no thoughts on what might be done to overcome it. The folks at Flex listened and came up with the "Limit Slew" resolution that cured the problem completely even with stubborn cases such as his. So, by using the "Antenna" form to select the best keying delay, and toggling the "Limit Slew" setting to softly unkey the amplifier, you should be able to find a combination that will work even with the most sensitive of amplifiers.

Of course this whole thing is a non-issue for my pair of 3-500's which don't flinch no matter what I do, but the "Limit Slew" option makes operating AM or any other mode possible with even very sensitive amplifiers.

Rob W1AEX

[IN3IEX]

I am not an owner of a Flex radio, nevertheless I see that AM configuration is very poor. With a SDR system it should be possible to select between constant carrier and controlled carrier with controls for: residual carrier, carrier decay time constant, single side band AM (carrier with one sideband), minimum power of the envelope, maximum power of the envelope, a selection of RF envelope clipping algorithm, etc.  Analog AM has solved most of these issues, now just put them in software....
If these options are already available, please let me know, maybe I will buy a Flex radio.
Tnx

[W1AEX]

IN3IEX,

I don't work for Flex so I can't really comment on the "features" you are looking for or the "issues" you are focused on. You might want to join the Flex Reflector and present your thoughts to the Flex hardware and software engineers.

73,

Rob W1AEX

Join Flex Reflector here -   http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz

[W1AEX]

Notes regarding AM Operation with PowerSDR version "2.3.5 Final Release"

There were a number of changes made to this version that require some attention when running AM. When things are set correctly, this version behaves beautifully, so don't be afraid to give it a try. If you decide you don't like it, the 2.2.3 version works just fine with the new firmware that is installed with the 2.3.5 release. You can simply start up the earlier version and it will behave like it always did. The list below briefly addresses changes that have been noticed:

1. As noted in the "Audio Gain Setup" article above, a number of users have mentioned that the first two audio settings at the MIC and EQ stages seem to behave a bit hotter. You may find that you have to back the Mixer and EQ gain sliders down a few notches to smooth things out. More details are noted in the Audio Gain Setup article.

2. As noted in the "AM Carrier Level" article above, the engineers at Flex corrected the issue with excessive carrier output power while running AM. The maximum carrier output power in early versions of 2.x.x was limited to around 40 watts of carrier, however, at some point this function was broken and it was possible to run 100 watts of carrier power in AM. Obviously, this left no headroom for modulation. To correct this, the maximum setting that can be entered in the AM Carrier Level field is now 50, and the default is 25. More details are available in the "AM Carrier Level" article above.

3. The most important change in my opinion is that many more settings have been moved into the "TX profile" region of the database. Most notably, the "Mixer" settings and the "Buffer" settings are now saved into each individual TX profile. This it a huge step forward and adds a lot of customization to each profile. With AM, the ability to swap buffer settings on the fly gives you the ability to change the behavior of the DSP transmit filter by simply switching TX profiles. The first picture below illustrates a 10KC profile using a TX buffer size of 4096 that creates a vertical brickwall effect. This is perfect for working in a crowded area of the band, but it can produce some harshness in the audio due to the abrupt cut-off. The second picture shows an identical 10KC profile using a TX buffer size of 512 that creates a much softer filter skirt. The gentle filter completely eliminates the harsh sound of the abrupt cut-off seen in the previous filter. With my setup, I did find that a buffer size of 512 introduced a noticeable garbling effect into the audio, so after some experimentation I settled on 2048 for my AM profiles. This is really worth playing around with as it allows a lot of flexibility of operation.

If you give the 2.3.5 version a try I don't think you'll be interested in going back to the earlier versions.

Rob W1AEX  

[W1AEX]

Using the PowerSDR I/Q Recorder to Listen to Your Transmitted AM Signal

Unfortunately, the way the “Monitor” function is implemented in PowerSDR, you can’t listen to your transmitted AM signal in real-time as you can with SSB, however, it is possible to hear what your AM signal sounds like by using the “Recorder” function of PowerSDR. By setting the recorder to make a pre-processed I/Q recording, you can view your transmitted signal in the pandadpater and listen to your actual transmitted audio. This is very useful for checking your audio for distortion, clarity, and tonal balance. The setup for accomplishing this is very simple:

1.  From the menu choices along the top left of PowerSDR select Wave.
2.  This produces a settings box named Wave File Controls where you should select Options.
3.  In the Wave Record Options box select Pre-Processed Audio for the Receive setting.
4.  We won’t be re-transmitting what we record so ignore the setting for Transmit.
5.  Click the “X” in the upper right corner of the Wave Record Options box to close it

Note: The pictures at the bottom of the article may help with the setup instructions above.

When those settings have been made, in the PowerSDR GUI select the AM transmit profile you want to sample and select SAM for the receive mode, then press the Record button in the Wave File Controls box. As the recorder is running, key the transmitter to make a test transmission. I would suggest that you wait about 5 seconds before you start speaking. The reason will become clear when you read the playback instructions below. At the conclusion of your test transmission, press the Record button again to stop the recorder.

To play back the recording you made, press the Add button in the Wave File Controls box and browse for the recording you just made. It will be time stamped so it should be very easy to locate. After selecting the correct file, check to make sure that you have PowerSDR set to receive with a bandwidth that matches your transmitted bandwidth. When everything is set, press Play in the Wave File Controls box. Because there is an 11 KHz offset when you transmit, your recorded signal will not be on the frequency you are listening on. Simply press the “0 Beat” button repeatedly until it locks in. Note that the reason for waiting a few seconds before speaking when you made the recording was to give you time to get the signal centered in the receive passband for playback. Adjust your AGC-T and Volume settings and take a listen to how you sound. What you hear is an excellent recording of what your signal sounds like when you are transmitting. What you see in the panadapter will give you an indication of the peaks and valleys in your transmitted frequency response as well as an indication of your transmitted bandwidth.

The link below is a video recording of my Flex playing back a recorded I/Q file of an AM test transmission. You’ll see at the beginning of the recording that it took 6 presses of the “0 Beat” button to get my transmitted signal into the passband and centered.

http://youtu.be/WtOGzfnu-Ho

In the second video recording, I used the I/Q recording feature to help set up the compression levels for a Marshall MXL-770 studio condenser microphone that I plugged into the balanced input on the back of the Flex 5000A. It was very easy to see and hear when the software compressor reached the point where the audio began to fall apart.

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

Although you can’t monitor your AM transmissions in real-time, the I/Q recording function built into PowerSDR can be used to give you an excellent “air check” of your AM signal. For real-time monitoring, a simple RF sampling monitor can be built with a handful of parts, but to be honest, the I/Q recordings reveal much more about your signal.

[Mike/W8BAC]

Nice job on this Rob. I am in the process of upgrading to a 5K and will be using your new technique for my new installation. I also like your article regarding the new gaming computer. Have you used the new box for Power SDR yet and if so, how dose it work?

Thanks for everything, Keep up the good work.

Mike

[W1AEX]

Hi Mike,

I think you'll enjoy the simplicity and reliability of the Flex 5K hardware platform. It's pretty much plug and play at this point. I did run the Flex with the new computer for about a week as I did some cleanup work on the retired gaming computer and it ran PowerSDR very well. The retired gaming computer that's handling Flex Radio duties now is based upon a four year-old Intel dual-core Pentium (Wolfdale e8400) running @ 3.0g that handles PowerSDR very smoothly. Hopefully, it will keep running for a few more years!

Are you planning any options for your 5K? Mine is stripped, no second RX, no VHF/UHF, no ATU, and I have to admit that I have no interest in adding any options in the near future. My big project now is to finish up restoring a 10 year-old Hex Beam. If I can get a couple of decent days in a row I'll be able to finish and move it to the top of the tower without too much effort.

Good luck with the new Flex!

Rob

[W1AEX]

Notes regarding AM Operation with PowerSDR versions "2.4.4 through 2.6.4 Final Releases"

The move to version 2.4.4 or any later version represents a major release for PowerSDR with a huge change in performance for the Flex 3000 and Flex 5000 that utilize a firewire interface. With these versions, Flex is using a completely new firewire driver with all new dependency drivers and a new firmware version to accommodate the new driver architecture. One very important thing to understand about upgrading to either of these versions is that once you upgrade, you will not be able to load earlier versions of 2.x.x unless you completely uninstall version 2.4.4 (or later) and follow a fairly convoluted procedure (involving older firewire drivers and firmware) as detailed in the release notes. Fortunately, if you export a copy of your 2.3.5 database before you upgrade, you can use the SDRTranfer utility to import your settings into version 2.4.4 (or later) and it should work perfectly. This makes the upgrade completely painless as all your old settings (TX profiles, band stack registers, custom bandwidth filters, etc) will be there without any fuss. Note that you need to uncheck the "Strict DB Import Compliance box" (Look in Setup>Options at the bottom of the tab) to allow importing from a previous version. Another good thing is that once you move to version 2.4.4 or later you probably won't want to go back!

The new firewire driver eliminates issues with DPC (Delayed Procedure Calls) that plagued some users with computers that exhibit high latency as a result of hardware and driver variables from assorted combinations of peripherals and motherboards. This eliminates audio glitches and subsequent distortion that some users experienced after a DPC event with their computer. The new streaming driver has the ability to function with a latency factor that approaches 0 ms making it very close to real-time. If you have experienced the disconcerting sound of your voice arriving late while you monitor your transmitted audio you know what this is about.

The next change is relevant to all Flex hardware that is used on AM. Flex has corrected the scaling error that has been evident with the AM Carrier Level setting for most of the 2.x.x releases. They also added the ability to fine tune this setting by increments of .1 if desired. You will almost certainly find that the AM Carrier Level setting that you used in version 2.3.5 or any other earlier 2.x.x version will require adjustment. Your results may vary from mine, but using my scope to adjust the AM Carrier Level it was very easy to arrive at the following settings for my Flex 5K:

100% modulation with 10 watts of carrier:   Drive = 20 --- AM Carrier Level = 45
125% modulation with 10 watts of carrier:   Drive = 25 --- AM Carrier Level = 40
150% modulation with 10 watts of carrier:   Drive = 30 --- AM Carrier Level = 35

I generally run my AM profiles to generate a signal that has the potential to hit a maximum of 125% positive peaks and find that this produces a smooth sounding signal that is compatible with most receivers that employ a standard diode detector. If you get reports that your Flex sounds distorted or gritty on modulation peaks it's possible that you might have the AM carrier level set too aggressively which translates as a number that is too low. Note that while you are setting this, you should keep an eye on the PowerSDR TX Meter (set to ALC) to assure that your audio levels remain below -2dB on peaks. Assuming that each audio stage in PowerSDR is adjusted correctly (see Audio Gain Distribution Inside PowerSDR in an earlier post above if you are not sure) the transmitted audio should be absolutely clean. When everything is adjusted optimally, whatever audio you put into your Flex is exactly what will come out of it!

These two versions look like excellent releases of PowerSDR for AM, but as always, it would be very wise to read the release notes before upgrading to this version. The links below will download each version's release notes in a PDF file with everything you need to know. Note that it's not necessary to install version 2.4.4 if you want to go directly to the 2.5.3 version.

Version 2.4.4  http://support.flexradio.com/Downloads.aspx?id=402

Version 2.5.3  http://support.flexradio.com/Downloads.aspx?id=411

Just remember that it's not an easy process to go back to an earlier version once you move to PowerSDR 2.4.4 or 2.5.3 but to be honest, I can't imagine why you would want to take a pass on the huge performance increase in these new versions.

Rob W1AEX

NOTE ADDED 11-10-2012:  The computer that I have been using with my Flex was loaded with Windows XP/SP2 when I updated to version 2.4.4 of PowerSDR. Several times over the past month I had a DRIVER_IRQL_NOT_LESS_THAN_EQUAL issue with the legacy Windows XP "ohci1394.sys" driver when PowerSDR 2.4.4 was running concurrently with programs that accessed the sound card heavily, such as Camtasia or Adobe Audition. I emailed Flex and they replied to me in less than 3 minutes. As it turns out, if you are retro like me and running Windows XP it is necessary to move to SP3 to avoid issues with the new Flex streaming firewire driver for the 2.4.4 version of PowerSDR. It is also necessary to have NET Framework 3.51 (version 3.5 + NET framework 3.5 SP1).

[W1AEX]

Free Band Text Editor for PowerSDR from Ray - K9DUR

This is a very slick program that allows you to insert the established AM operating frequencies or AM windows on any band. There's nothing worse than seeing 75M SSB in the "Band Text Field" as you're having a nice AM conversation with your buddies. Now you can set ranges of frequencies or individual frequencies to be labeled however you want them. It took me less than 10 minutes to label every band with my AM operating areas. Note that the "BandText Editor" only works with versions of PowerSDR that use the .xml format file for the database. This means it is compatible with PowerSDR version 1.18.1 and all the 2.x.x versions that followed.

Over the years Ray has written some very nice utilities for managing the PowerSDR database files and this is another fine piece of work. The software is available for free from his site at the following link:

http://k9dur.rnaconsultingservices.com/Downloads.aspx

Ray generously offers his software utilities for free but you might want to consider sending him a few bucks if you enjoy what he has created.

Rob W1AEX

[KK4YY]

Hi Rob,

Trying out my new Flex 1500 into the dummy load while monitoring on a 'scope was producing terrible AM. Carrier dropping with modulation just as you noted in a post on the FlexRadio Forum 2 years ago.
http://forums.flexradio.com/Topic3515-56-1.aspx?
After much trial and error I got my 1500 to work by setting the "Mic" gain at "0". Yes, ZERO! Then brought the Audio Mixer up to near Max, the Drive about 50 and get 1 watt of what appears be be actual AM and not some weird inverse controlled carrier AM. I don't know if others have found this trick or not but I could find no reference to it anywhere. I'm posting it here in hopes that it may save some other unfortunate soul the hand wringing, teeth gnashing, and garment rending that I went through to figure it out.

I still don't know if I have it where it should be but it looks good on the scope and my cheap PEP watt meter now swings up with modulation instead of down. The Flex article on how their ALC works explains the mechanics of it but not the particulars for this radio. Your posts in this thread are much better in relating to actual settings but pertain mostly to the 5000. So this is for the 1500 guys. Hope it helps.

Don
KK4YY

[kx9dk]

Don,

Thanks for posting the info for the 1500. I have a 1500 I'd like to try on AM. You saved me from posting a question regarding some ballpark settings to start out with. I'll give yours a try and let you know what happens.

Thanks,
Dave
KX9DK

[W1AEX]

Don,

Thanks for posting this information on the 1500 anomaly with AM. This will definitely help the guys who encounter the issue where the carrier and modulation interact inappropriately. I know this drove Pete W1VZR crazy a couple of years ago. It seems that a setting of "0" with the MIC slider actually sets the gain for that stage to "0" rather than dropping the audio level to "none". The first guy I ran into who used this to advantage was Craig W1MSG who wrote an article about it to explain how he overcame an audio anomaly he encountered when using an external rack with his 5000A:

http://w1msg.net/media/blogs/a//Flex5000.pdf

I'm guessing that your conclusion about ALC interaction is spot-on as PowerSDR apparently monitors audio levels at the MIC stage and based upon levels it sees there the ALC will intervene, sometimes inappropriately when it comes to AM with the 1500. That's great news that you have found a way around this! Thanks again for sharing your findings.

73,

Rob W1AEX

[K1JJ]

Rob,

I just "discovered" this thread.   Great job putting it all down on paper. It will help us SDR Newbies a lot.


I'm getting back to my HPSDR project maybe next week, so will read it over and go from there.


It's a lonely whirl, writing these help articles,  and we need a slap on the back once in a while...    [SLAP!]

T

[W1AEX]

Hey, thanks Tom! Make sure you add your findings as you discover them when you roll up your sleeves and get busy again with the HPSDR project. It's my understanding that there are some notable differences between the Flex version of PSDR and the version that has been adapted for use with the HPSDR platform. Items that are specific to setting things up with the HPSDR hardware would be helpful to those who use the HPSDR hardware on AM. Nag Frank GFZ to document some of his HPSDR kung-fu as well!

Rob W1AEX

[n1eu]

(Updated)
I've been using an Apache Labs ANAN-10/Hermes SDR with PowerSDR OpenHPSDR mRX v3.1.5 and ANAN-100D/Angelia SDR with PowerSDR OpenHPSDR mRX PS v3.2.8 on AM for a few months now, in conjunction with an REA Modulation Monitor.  I operate mainly on 10M and 15M.

You must initially set proper audio levels watching the meter in MIC and EQ positions.  My goal is -5dB level on voice peaks.

To assess your tx audio equalization, use the built-in wave recorder/player with PSDR "Dupe" engaged, so that you are recording what is being transmitted and received.  Without "Dupe", you are just recording the audio input going to the transmitter.  Recheck the EQ meter level after any changes.

You then must set the carrier level control to optimize the modulation level.  You want to be just hitting 100% negative - needle shouldn't be slamming into and staying at 100% negative.  I'm finding that this setting varies greatly from band to band.  I therefore save different transmit profiles for each band.  Every person you ask will give you a different answer for proper carrier level setting.  With the REA Modulation Monitor, I find that optimal carrier level settings go from about 70 on 10M down to about 40 on 160M.  These settings provide 100% negative modulation and 100-130% positive modulation.  [update Jan2015: this step no longer necessary with PSDR v3.2.19 and later - just leave carrier level at 100]
I  typically transmit with 4KHz wide audio in AM mode and consistently receive rave reviews on the tx audio/modulation quality.

73, Barry N1EU