Static vs Adapative - Pure Signal vs Adaptive Pre-Distortion

Mike VE3CKO

I know more information is forthcoming and I haven't seen any discussions here yet about the elephant in the room, I've came to some conclusions so please someone correct me if I'm wrong.

The long awaited APD feature has finally been officially announced at Dayton Hamvention 2024 and will come later after the release of the new 8000 series radios. Adaptive Pre-Distortion or APD is the term use by FlexRadio and from what I gather from the video Michael Walker did with Kyle AA0Z EXLUSIVE FlexRadio Announcement Hamvention 2024  I've jumped to the conclusion, "ADP" will be superior to "Pure Signal" (used in software for Apache Labs radios). Pure Signal have had bragging rights for quite a while. It appears to me the coming APD will give FlexRadio the competitive edge and because of that FlexRadio's 8000 series will be be a home run for FlexRadio. This screenshot taken from the video:

They go in detail as you can see the difference between static and adaptive pre-distortion. Without actually saying it, I think STATIC pre-distortion is what FlexRadio describes what Pure Signal does with about 8 - 15db benefit based on the manual calibration. I have used Pure Signal in Thetis for a time when I had an Anan 8000DLE, to get it going proper I had guidance from the knowledgeable group of Anan gurus as there is zero support from Apache Labs.

ADAPTIVE pre-distortion as described examines the spectrum is real time, adapts to changes in frequency, voltage, temperature and mode on every transmission, and typically sees 15 - 25db improvement. In the video another screen shot says FlexRadio is implementing ADAPTIVE which by the numbers and being automatic is clearly superior to STATIC.

Adding this all up, it appears FlexRadio's ADP is not only the answer Flex users have been waiting for but they have upped their game by improving upon their adaptation thus making it superior to Pure Signal. Past complaints why some users switched to Apache Labs though the SSDR gui vastly better, were because they had Pure Signal and better noise reduction, well FlexRadio has taken aim at their competitor by addressing those two major issues in the announcement. Looking forward to see what is in SSDR v3.8

I should note I have drank the KoolAid as I operate a 6700, PGXL, TGXL and AG.

Duane N9DG

Actually it is the Icom IC-7610 implementation that looks to be static. The Apache approach uses a "sampling" RX like Flex will be doing.

Gary NC3Z

Flex(s) and Anan(s) owner here. I am pretty sure Thetis PureSignal is adaptive. In all the PureSignal documentation the approach used in PureSignal, is called “Adaptive Predistortion.”. And you can see the process during a transmission with the Anan the constant updates. There is a way to do a manual single calibration but not sure of anyone that does it manually.

The only other radio using ADP is the IC-7610 and that certainly seems to be Static.

One thing I have not been able to see on the back panel of the 8xxx is the input for ADP when using the "full loop" with an amplifier. Has anyone seen a dedicated input?

Gary NC3Z

Found this little excerpt in the Pure Signal manual and it is reference the "manual" mode vs the desired automatic mode used in Thetis.

Mike VE3CKO

I haven't experienced the 7610 implementation but have Pure Signal with the PGXL. Glad to get more clarification on this, thanks Gary. Sure hope Flex's APD is more stable than PureSignal from my experience. I wish I had taken better notes as what exact settings needed tweaking and what values have worked for others but do remember needing calibration every coldstart, then during operation, changing bands etc, then other times without touching anything it would be ok. Though I typically set my audio and leave it alone, any audio chain alterations would require values to be tweaked again.

We'll have to wait and see what and when gets actually released and the flavour of it. As per APD input, the talk was using the aux. connector on the 6000 series (if implemented), you would think there will be dedicated connector for the upcoming 8000 series. Perhaps the final back panel was not ready in time for the announcement. Thanks Gary and Duane for your input.

Gary NC3Z

Yes, the #1 step in getting PS to work well is the audio drive level must be set correctly, it does impact tremendously but once set is very reliable. Earlier PSDR and Thetis did has some PS instability but the last couple of versions of Thetis with the new WDSP has nailed all that down.

I am holding out hope that ADP would be an option for 6xxx but I think the statements made were no due to processing power that the new 8xxx wil have over the 6xxx. But time will tell, guess we will see the new SSDR with ADP in Aug??

WX7Y

You mean Aug 2025 or 2026 for ADP right Gary?

The Radio may be available Aug 2024 but from Mikes release video he hinted and said next year for the ADP SmartSDR software and it wont be available at the Radio's Release date.

And yes I am with you in hope that a ADP adapter would be available to keep there word on the 6000 series back in 2017 or earlier even if we have to pay extra to get, it would be worth it to many people.

Maybe someone should start a thread or questionnaire on the Message board asking people how much would THEY pay for Flex to build and sell them an ADP adapter to run on the 6000 series so FLEX could see how many people even care about ADP on the 6000 series radios.

IF Flex decides to not build it then how would someone go about it and what interface should they be using to handle the Throughput? Maybe a 3rd party would build it.

EtherNet, API, or FlexWire to interface to the radio.

Russ Ravella

Regarding the connector(s) to close the loop for APD in the new 8XXX, a chart set I saw described by Flex on YouTube over the weekend actually covered the back panel item by item. And there were no dedicated connectors for APD (plural for SO2R as supported by the PGXL). It almost seems like rather than actually design a new radio, Flex just took their ML-9600 and stripped it down for ham use. Much less non-recurring cost that way. And in and of itself that's not necessarily a bad thing of course.

But going so far as to push APD as the 8XXX's key new distinguishing feature without even providing dedicated connectors for it - if that's what ends up happening in the end - is very disappointing. After all they included APD output ports on the PGXL FIVE YEARS ago in anticipation of this. What they've said is they're likely to use either the transverter or RX ports instead. And if that's the case users are going to have to choose between functionalities which seems pretty lame for a very expensive brand new radio. Perhaps there will be a way to use the AUX connector as Mike mentions above which would be far better though it doesn't seem clear that it's suitable for two of this kind of signal ? At least Flex apparently hasn't decided that it is.

It seems to me based on past experience that it could be quite a while before most of the advertised benefits like APD and advanced noise mitigation features are going to actually be available to users anyway. In a recent post exchange I had with Steve Hicks on the Flex groups.io reflector, he made it extremely clear that hams should be happy to get what they get in terms of Flex resources and that could be zero at any time and for any length of time as it has for so much of the past.

73 Russ KR6W

Al_NN4ZZ

Some questions about APD….I don’t have any experience with APD so these questions are just to learn a little more about what to expect. It may be too early for Flex to have decided on the details and interface yet and figure I have plenty of time to learn a little more.

Sounds like APD for the 8xxx hardware might be ready by the end of 2024 and if it can be applied to the 6xxx models it could be much later depending on the solutions. Since this new feature has been announced already, maybe Flex will be able to show some mockups of the user interface early and we can ask questions and give feedback.

QUESTIONS

**********

- is it typical to have options to enable / disable the APD functions. (i.e. ON/OFF for the internal radio amplifier and ON/OFF for the external amp ). It seems like these options would allow you to view the output and see how the APD is working in all 4 of the cases.

- are there common settings / controls to adjust the APD function? What are they? What do other vendors offer?

- if your SSDR is setup to monitor your xmit signal on the waterfall, would you expect to be able see the impact of the APD settings (assuming you can control them as noted above). Or is the signal displayed on the waterfall for your transmitted signal too strong to see the effects locally?

- most of the examples I have seen are showing the APD effects on SSB transmissions. Is it safe to assume the effects of APD on CW are also the same and as significant?

- do other vendors offer the option to select either STATIC or ADAPTIVE APD? I asked Steve whether STATIC might be an option for the 6xxx hardware since it would be less intensive from a processing standpoint. He said they have discussed it but generally felt it would not be as acceptable since STATIC APD is not as effective so probably not. But they are going to look at options to provide external processing for the 6xxx hardware first and see if ADAPTIVE APD would be feasible.

Thanks, Al / NN4ZZ

Gary NC3Z

Al, just to answer your questions as it relates to Anan/Thetis.

You can turn Pure Signal (ADP) from the GUI

You can set ADP to run full loop (e.g. and Amp with a ADP output or coupler) or you can set it to be just the 100W radio from within settings.

You have the option to run Static or Adaptive. Static is more used to set IMD from a 2-Tone for measurements and Automatic is for use with ever changing SSB and Auto is the preferred setting.

You have a slate of settings you can experiment with if you choose.

You have a graphical display showing the raw linearity vs the correction so you can see how well or hard it is working.

Al_NN4ZZ

Gary,

Thanks for the info and details. Very helpful.

Do you have any feedback on the application of APD on CW signals? I only operate CW so trying to understand if it is applicable to that mode. I have some feedback that it's not.

Thanks again, Al / NN4ZZ

KD0RC

Hi Al, CW does not require a linear amplifier so APD doesn't help.

Martin AA6E

To add to KD0RC's comment:

APD is a way to minimize intermodulation distortion (IMD), which is a problem only for "multitone" modes like analog voice and some digital modes like psk31, MFSK, DMR etc.

CW, FT8 etc. transmit one tone at a time, so there is no IMD and no need for APD.

73 Martin AA6E

Al_NN4ZZ

Len and Martin,

I thought APD addressed CW also after reading an article on the ARRL web page about the Clean Signal Initiative ( CSI ), excerpt below.   After re-reading it I can see that while CSI does address CW signals, it is not tied to the APD efforts.  

So it sounds like for CW only operators, the APD features in the 8xxx series won't be applicable to my use.  I'll have to wait and see what other new features the 8xxx radios implement that would be of interest and whether those software updates are also supported on the 6xxx hardware.

Thanks for calling this to my attention.

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com
6600 SSDR V3.6.8
TGXL 1.1.20
PGXL  3.7.32
Win10 Pro V2004

Mike-VA3MW

GM all

I'm back and my feet aren't so sore and my **** is less sore from the 8 hour drive. (look at that, you can't type the synonym for Donkey).

• CW is solved not by APD, but by rise time. Essentially, rise time should be 6-8ms or more. If I remember correctly, some radios have it variable to about 2ms (ouch)
• for the 'Mask' no one vendor yet meets those standards, but we have a couple of radios that are vary close (no, I won't say which until we can 100% meet that metric)
• APD requires an RF tap from the last amplifier, either barefoot radio or Amp. Many Amps today have a -60db tap built in. (PGXL, RF-Kit, SPE, etc). This RF is fed back to the radio.
• The RF is fed back to the radio and into an RX or XVTR port and then processed internally on the 8000 series. If running without an amp on an 8000, it is all internally wired. See the diagram below.
• No external sampler is required for those with a -60db tap
• If an external sampler is required, we will build one that will have a T/R switch in it so you don't lose an RX antenna if that is important. It is not built yet, so it is subject to change
• 6000 series: It was determined that there wasn't enough horsepower to run ADP on these units reliably (I don't have the details as there are a lot of things to consider). The thought is to have a small external processor complete with a sampler that will feed the correction data back to the radio. There is more work to be done on this and we need to make sure this is going to be a cost-effective solution.

Thanks everyone for your interest in ADP and CSI. We are really looking forward to making it happen.

Here are some of the key slides from Hamvention.

73, Mike va3mw

rdwing

What signal level is expected on the APD return line to the radio? Will options be given to enable the feature in static mode vs. adaptive mode? Will there be a switch to indicate which sampler the radio should use? What about for an SO2R scenario, will it support multiple sampler inputs? Has there been any work to address memory effect in SSPA's?

Mike-VA3MW

• -60db of a 2000 watt signal :) There is a fair bit of range there.
• At this time, only Adaptive will be available
• Yes, you select the RX antenna port you wish to use
• SO2R — good question and it will have to be investigated and tested
• You'll have to explain memory effect more so I can research it. It is a term I am not familiar with

Mike VE3CKO

As per the APD input to an 8000 series radio, will the aux port (pre-specified pin) also be an option? I would hate to give up either RXA RXB or either XVTR input. Saying that I suppose it will be easy enough to connect up another relay for my node red to control. My flow would engage the relay when the PGXL goes in operate mode that would connect the APD from the PGXL to the RXB antenna, otherwise the normal receive antenna would be connected to the RXB. Or a simple A/B switch, lol.

Also there must be a method to select use APD with internal PA, or external AMP ? May I suggest perhaps along the bottom status bar beside FDX ?

Gord-VA7GP

Hmmm… I am not sure any of us would be happy with the Sample Signal coming in on an Auz pin. The reason: the Sample is an RF signal, and not super-strong either (few milliwatts). It's a real-time, RF sample of what is being transmitted. Personally, I feel it should be handled with an RF connector (BNC, SMA, perhaps even an RCA (hey - it was good enough for Collins and Heath 😀). But I wouldn't feel comfortable feeding this back onto one pin in a multi-pin connector.

YMMV…

Mike VE3CKO

Good point Gord.

Russ Ravella

Hi Mike,

Mike Walker addressed this question in the updating my station to 8600M with Mercury LUX or other amps" thread further down. What he says totally escaped me over the weekend. Sure seems like dedicated ports are called for but apparently this is what they're going to do instead:

"No, you don't have to give up an RX port. 

We will be building a sampler box that is a T/R relay that will allow you to use the port for an RX antenna while in RX and an RF sampler port while in TX. N5AC showed a picture of what it would look like during the banquet presentation. (I don't have a copy of that slide deck, so I can't grab the picture). 

However, my guess is that it is a very small list of people who may use all 4 possible options on an 8600. I come close with my home station with 2 transverter ports being used for Satellite work and 1 RX port being used by a DX Engineering active antenna. 

It is good news that for those that might be fully populated, there will be an option."

73 Russ KR6W

Mike-VA3MW

The APD sampling is an RF signal is it comes in an RX or XVTR port.

TomWA1MBA

I believe (but might be corrected) that FT8 all by itself does not benefit from predistortion and can be run on a non linear amplifier. Fox & Hound (F/H) FT8 howerver is a different animal. Often employed by DXpeditions, F/H runs multiple FT8 signals simultaneously and therefore is subject to degredation (ghosts) and (therefore) should improve performance with predistortion.

Tom Porada

A few days ago another ham friend less than 500' from me, an owner/user of an ICOM 7610 installed Icom's latest firmware with DPD (Digital Pre Distortion). A non adaptive fixed proprietary implementation and asked that I look at my 6600 panadapter, take some screenshots and tell him what I saw. To say the least, I was impressed! In this screen shot, the bottom portion shows his signal w/o DPD and the top with it turned on. Quite a difference in the unwanted products outside the signal. Remember, he is 500' from me and no matter what we do, our signals are 'large' to each other.
Pre-distortion works. I sure would like to have it for my new/reserved 8600 and my existing PGXL.
Nothing better than a clean signal!!

WX7Y

WOW, now that's amazingly clean signal for such close proximity to your Station.

Mike-VA3MW

It will be interesting if those contestors who need to hold a run frequency actually use it.

It is a well known and discussed contesting tactic to splatter so that others do not encroach on your run frequency.

Wayne VK4ACN

Very impressive