SDR of the future

[AF6WL]

Well I actually have an all HF bands digital cartesian loop transmitter on my test bench!

The trick is to keep the total group delay around the decimate/interpolate->DA->Filter->Power amp->Filter->AD loop small compared to the period of the highest pole in your loop filter, this requires that the loop be closed at a sample rate of a few Mhz for a 50Khz loop bandwidth, sufficient to attenuate the IMD out to the 11th order for an SSB transmission. Not really  a problem in a decent FPGA.
In fact most of the RX path can be reused to close the loop, it needs a separate NCO so the phase can be adjusted between tx and rx paths, and the loop gets closed a few stages up the CIC chains to keep the total group delay acceptable.
...
At some point I must finish this project and do a proper writeup, but you know how projects are.... Always another cool idea to try out!

73 M0HCN

That is an impressive project.
I would very much like to be reading about it one day - not just the hardware but also the VHDL/Verilog.

[KE5JPP]

Well I actually have an all HF bands digital cartesian loop transmitter on my test bench!

The trick is to keep the total group delay around the decimate/interpolate->DA->Filter->Power amp->Filter->AD loop small compared to the period of the highest pole in your loop filter, this requires that the loop be closed at a sample rate of a few Mhz for a 50Khz loop bandwidth, sufficient to attenuate the IMD out to the 11th order for an SSB transmission. Not really  a problem in a decent FPGA.
In fact most of the RX path can be reused to close the loop, it needs a separate NCO so the phase can be adjusted between tx and rx paths, and the loop gets closed a few stages up the CIC chains to keep the total group delay acceptable.

One of the many nice things about this approach is that the same AD hardware is needed anyway if you are doing a digital down conversion rx, a small signal relay suffices to switch the AD input to the power sampler on transmit.

For most amateur service cartesian loop is better then polar as polar loops have issues with signals where the envelope crosses zero as the phase becomes indeterminate at that point, and many amateur mods do this.

The PA is somewhat interesting as both the bias and the drain voltage are under DAC control, so the thing usually starts off in class AB, at maybe 12 - 15V on the drains, as the envelope comes up the drain voltage rises to track, and the bias gradually reduces moving towards class B, and even C, finally as the power rises further and the driver starts to run out of puff the bias is automatically increased to get a bit more gain. The gain changes would normally be a disaster, but 30db of envelope feedback cures most ills, and the variable drain voltage means that efficiency stays much higher then would normally be expected as you turn the power knob down. 

For CW and PSK the system can run the amp in AB during the rise and fall of the carrier then transition to full class C for the constant amplitude bits, very nice in a too hot shack.

prototype performance numbers:
@ 450W transmitter output from pair of SD2933 (Gives just under 400W at the feedpoint, UK legal limit) into 50ohms, 2 tone IMD3 ~-60db ref one tone (There are a few spurs from the switchmode drain regulator, I must rebuild that board to sort that out).

I am working on the logic to monitor load Z so that the drain voltage can be increased if needed to support a higher then nominal drain impedance due to a load mismatch.

At some point I must finish this project and do a proper writeup, but you know how projects are.... Always another cool idea to try out!

73 M0HCN

Yep, and I have a rocket ship in my garage that runs on water and takes me to the moon and back in 10 minutes.  Until you document your system so that others can duplicate your results, I will remain skeptical.  I have seen many other claims like this that never see the light of day (or withstand the light of independent inspection).

Gene

[N9VV]

Did anyone mention that for 4 years there has been a Ham band
DDC/DUC Transceiver called the ADT-200A available from Switzerland
http://www.adat.ch/index_e.html
http://www.eham.net/reviews/detail/7037
offering PRE-DISTORTION as a feature in it's digital design by
Hans HB9CBU and his EE daughter Doris HB9DRV :-)

We are just at the entry door to exploring DDC/DUC Transceivers.
TAPR has announced open ordering for their affordable "Hermes" and
Apache Labs has announced a full 10-15W kit with full
warranty/service/support/ in the U.S. and factory repair.

Hang in there, there are many amazing dimensions of DDC/DUC to
explore with FLEX, TAPR, Apache, SunSDR2, ZS1SDR, HiQSDR, and many others :-)

73 de Ken N9VV

[ZENKI]

HB9DRV, is that not Simon's call from Ham Radio Deluxe?

ADT200A has been extensively debated in here and in other forums. The excellent transmitter and PC driven bandscope are both excellent. The onscreen dynamic range and speed cannot be matched
by any other bandscope/SDR solution. You can clearly see that this radio was designed  by someone who knows how a spectrum analyzer should work, and by someone who has used a quality spectrum analyzer.
The rubbish software that comes with most SDR radios are just token gestures whose potential accuracy is squandered by poorly thought out bandscope software that is useful for little else.

Its a real shame that the ADAT transceiver is not marketed more aggressively. In my book the ADAT will always be the better radios because so much attention was payed to the spectral purity of the transmitter.
All the other SDR solutions  and projects like the Hermes have transmitters not  much better than CB grade power amplifiers with poor inherent IMD. What good is it having such fine receivers and yet the very transmitter will never allow
most other hams on the air from realizing their full dynamic range potential because of bad TX IMD. When viewed from this perspective you might as well stick with a IC706 despite all the hype. Hams and the way they apply technology
to their hobby should do so within  the parameters of the laws of physics. Dirty transmitters when combined with state of the art SDR receiver cant bend the laws of physics, we still dont get it when it comes to the subject of transmitter
design... A splatter blanker would be more useful than a new SDR technology transceiver!  Heck even my KWM2a puts out a cleaner signal than 99% of transceivers of today.

The reality of SDR transceivers is making the ham band even worst when we combine sub-standard transmitter with wideband ESSB audio. The snap, crackles, pops, ALC buckshotting and general crud from station using wide bandwidth audio is really unacceptable abuse of valuable resource. Its really unforgivable that hams have to put up with  this deliberate interference equipment that meets no standards and just causes QRM. It seems that most of the adopters of SDR radios adopt them for this very reason, SDR radios can transmit wideband ESSB. Its unfortunate that this technology  cant improve ineligibility  and all that they are good at is causing a lot of misery on the ham bands.
Its ludicrous that this technology can be used without limits regardless of  the harm that they cause. All I know as we get supposedly better SDR transmitter technology the average TX audio quality and ineligibility is getting worst and likewise the splatter. The trend in military HF radios is the opposite, cleaner transmitters and quality TX audio that gets the message through. Hams have truly gone off on tangent  that is nothing more than turning  the ham radio hobby into a glorified version of CB amplified microphone practice.

Did anyone mention that for 4 years there has been a Ham band
DDC/DUC Transceiver called the ADT-200A available from Switzerland
http://www.adat.ch/index_e.html
http://www.eham.net/reviews/detail/7037
offering PRE-DISTORTION as a feature in it's digital design by
Hans HB9CBU and his EE daughter Doris HB9DRV :-)

We are just at the entry door to exploring DDC/DUC Transceivers.
TAPR has announced open ordering for their affordable "Hermes" and
Apache Labs has announced a full 10-15W kit with full
warranty/service/support/ in the U.S. and factory repair.

Hang in there, there are many amazing dimensions of DDC/DUC to
explore with FLEX, TAPR, Apache, SunSDR2, ZS1SDR, HiQSDR, and many others :-)

73 de Ken N9VV

[NI0Z]

The reality of SDR transceivers is making the ham band even worst when we combine sub-standard transmitter with wideband ESSB audio. The snap, crackles, pops, ALC buckshotting and general crud from station using wide bandwidth audio is really unacceptable abuse of valuable resource. Its really unforgivable that hams have to put up with  this deliberate interference equipment that meets no standards and just causes QRM. It seems that most of the adopters of SDR radios adopt them for this very reason, SDR radios can transmit wideband ESSB. Its unfortunate that this technology  cant improve ineligibility  and all that they are good at is causing a lot of misery on the ham bands.
Its ludicrous that this technology can be used without limits regardless of  the harm that they cause. All I know as we get supposedly better SDR transmitter technology the average TX audio quality and ineligibility is getting worst

About ESSB, I do believe the vast majority of ESSB transmitters are non SDR users.  If you are referring the splatter that goes outside the 3.2K bandwidth, as a flex radio user I see that with all kinds of radios, including other flex users.  It's easy to see when you have a Panadaptor regardless of which radio your using the Panadaptor with.  I certainly did not buy my radio for ESSB and I am confident most others users don't buy them for that either.  The sad main two reasons I hear of others buying SDRs are because they wanted to try something different and it looks really cool!

This all in my mind brings us back to why we need a real SDR in the future that allows for modular software defined functions to be added on a as needed/desired basis.  Most people, including myself, don't see very many real benefits being realized with today's radios by their virtue of being an SDR.  Meaning, most non SDRs do all the same things as SDRs and in many cases surpass the current crop of SDRs in performance.  SDR radios are not yet really leading.

Just my two cents!

[NI0Z]

After some soak time in thought, the future of SDR will never really be about the Average hams.  It will be for those with deep pockets ( I am talking about a full 1st class SDR transceiver when I say this).

The future of SDR is likely to play out more in HTs and the digital realm.

We won't see a mass produced radio that has anything more to do with a PC than typical rigss we see today.  The SDR will come in the black box and be more geared towards making life easier and cheaper for the manufacturer than necessarily the end user.

That's my humble prediction at this point.

[K0IZ]

55Mhz direct conversion, 7 receivers, 1/2W xmt output thru 10M, 3/8 watt on 6M, great specs.
Digital Up and Digital Down.  Single board.
Orders for the ONLY production run must be placed by July 25.
$895/$940.  Future is here!