Is Full-Duplex possible on the 6000 series?

  • 5
  • Question
  • Updated 4 years ago
  • Answered
Most of my operating is with amateur satellites and the question came up about using my Flex-6300 as a full-duplex interface to transverters. But let me use an example that simplifies the question.

Suppose I am using a Flex-6300 with SmartSDR defining two panadapters. One of the panadapters has slice A on 28 MHz USB connected to the ANT1 port for receive. This panadapter will be used for receive only.

The second panadapter has slice B with TX enabled on 50 MHz USB connected to the XVTR port for transmit. This panadapter will be used for transmit.

Let's assume "Show TX in waterfall" is enabled. When I transmit on slice B, the panadapter with slice B will show the transmit auto on both the spectrum and the waterfall displays. But while I am transmitting on slice B, the panadapter for slice A continues to show the received spectrum on the spectrum display, but the waterfall is black while transmitting and the audio from slice A is muted.

If the spectrum on slice A is displayed, the signal should be available for demodulation. And the demodulated signal should be available for audio output. But the audio output from slice A is muted. If the audio were available, full duplex operation would be possible.

Why not? Is this possible on the 6500 or 6700?

Ron
Photo of Ron Parsons

Ron Parsons

  • 73 Posts
  • 10 Reply Likes

Posted 5 years ago

  • 5
Photo of Jim Gilliam

Jim Gilliam

  • 462 Posts
  • 71 Reply Likes
Perhaps the near field of the transmitter will saturate the semiconductors of the A.D converter and render them inoperable. Unless extreme shieding is employed it could be very difficult to isolate the semiconductors from the intense near RF field of a transmitter located on the same chassis.  
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1057 Posts
  • 1097 Reply Likes
Official Response
What is and is not possible varies by specific radio -- sometimes there is a hardware limitation such as how the relays engage for specific antenna selections.  There may or may not be coupling issues that vary by port on the radios.  Finally there are software considerations -- not everyone will want to run full duplex and so we have to decide how to present the option (if possible) and make it easy to use and understand.  To date, we've not looked at full duplex operation in detail.  I can say that we've done some preliminary investigation on the 6700 and have found some situations where it works well, but we still need to work out the details and decide how/if we would present the operating mode.

The mute of a slice was added because it was very distracting to those that didn't want to run full duplex (most folks), but it is a software switch and it can be changed when we have time to sort things out and decide what does and doesn't work.
Photo of Ron Parsons

Ron Parsons

  • 73 Posts
  • 10 Reply Likes
I agree that full duplex is something that most users will not use, but some subset of users (me included) would find it essential. If it's as "simple" as a software switch, why not implement it with a default of "off" and let us users determine if it (a) works, (b) is useful, and (c) does not cause harm. I'd love to test it.
Photo of Ken - NM9P

Ken - NM9P

  • 4239 Posts
  • 1351 Reply Likes
I would also be interested in this capability if it is available on the 6500 when using two transverters, one for uplink and the other for downlink.  After I get m HF/VHF station complete and towers up, I hope to begin assembling a satellite station based upon my 6500.  Full duplex would be very nice.  Otherwise I will need to enlist my 1500 for the other half of the station.
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3762 Posts
  • 1621 Reply Likes
While I think it would be relatively easy(??...yes I know nothing is ever as easy as it looks) to have full duplex on a 6700 as it has two SCU's

I suspect it would be rather difficult to do the same for a Single SCU 6500
Photo of Ken - NM9P

Ken - NM9P

  • 4239 Posts
  • 1351 Reply Likes
I didn't know whether or not you would need 2 SCU's in order to receive while transmitting since you are only using one receiver.  (I need to look at the block diagram again to see if an SCU is involved in generating a transmitting signal or not.)  It may be a processing power limitation also.  

Perhaps Steve could fill me in on the mechanics?
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1057 Posts
  • 1097 Reply Likes
The 6700 has three major RF parts: two SCUs and an exciter/transmitter.  There's no fundamental reason why you could not do full duplex with a single SCU.  Looking into this in detail is primarily a matter of priority.  In general, there is excellent isolation between antenna ports on the radios and the receive ports (>70dB) which seems promising.  But don't get your hopes up because again, we've only done some testing on the 6700 to date and I just haven't had time to look at this for other radios.  It was not a design goal, but may be possible.  The most likely thing to spur us to look into this further is a compelling use case on HF for having full duplex that is explained well and likely to be used by many people.
Photo of Jim Gilliam

Jim Gilliam

  • 462 Posts
  • 71 Reply Likes
What power level was the transmitter, what kind of transmitting antenna were you using and what kind of receive antenna were you using to get the measurement of >70 dB?
Photo of Ken - NM9P

Ken - NM9P

  • 4239 Posts
  • 1351 Reply Likes
Thanks, Steve.  If I understood your reply, Full duplex may be technically possible on the 6500 & 6300, but not currently in the developmental plans.

I have hopes, after my tower and LP antenna project is completed, of using a pair of transverters to build a satellite station,  It would be very desirable to be able to do full duplex with the transverter output and a RX input with the transverters so I would need only one computer interface to my 6500 which would adjust both frequencies automatically to account for doppler frequency shift.

I don't know about HF full duplex, but as I have read, it is almost essential for a competitive satellite station.

In my opinion, this would be put on the long term development list, for some time after 2.0 is released and perfected.

Ken - NM9P
Photo of Ron Parsons

Ron Parsons

  • 73 Posts
  • 10 Reply Likes
Ken,

I have just completed a conversion of my satellite system, using transverters, from Flex-1500 (TX) and Flex-3000 (RX) to a Flex-6300. The only missing pieces are full duplex on the 6300 and implementation of the I2C protocol on the 6300. I am using my Flex-1500 as a coax relay controller (with the help of W0DHB and his FlexSATPC software). SatPC32 performs the Doppler and rotor control.

I have what I call "semi-full-duplex" since the receive slice spectrum display is active even though the audio and waterfall are missing. It does allow me to see the downlink spectrum for adjusting the transponder offset and see the strength of my downlink.

Consider this another plea for full-duplex on the 6300.

Ron 
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1057 Posts
  • 1097 Reply Likes
Jim, the >70dB number I provided is an in the radio number.  The way I see it, if we were to do full duplex we (FlexRadio) are responsible for the isolation in the radio and you are responsible for the isolation outside the radio.  If you're antennas only have 50dB of isolation, there's nothing I can do to fix that in the radio.  Similarly, if there is an isolation issue in the radio, you can't fix it by increasing your antenna isolation -- it's all about the weakest link.  

So in my tests I ran 100W into a dummy load and measured the power on each receiver on each other antenna port.  But the power level doesn't matter provided you can see the signal in the panadapter --- the result is linear so the isolation is the same regardless of the power used (but you have to use the PA if you want to understand the isolation to the PA, you can't just use the exciter).
Photo of Jim Gilliam

Jim Gilliam

  • 462 Posts
  • 71 Reply Likes

I absolutely agree. My feeling is that the user will expect to use full duplex without regard for isolation. It is quite possible that the transmitted RF can saturate the receiver port unless some kind of pre-filtering is used to provide the necessary isolation. I think duplex operation would be a joy and hope that it proves successful.

Thank you for your reply.


Jim, K6QE

Photo of Jim Gilliam

Jim Gilliam

  • 462 Posts
  • 71 Reply Likes

I am not sure why anyone is disregarding the RF issue. Have you ever listened to your receiver on one band while transmitting on another band at the same QTH? Some times under varying circumstances RF amplifiers experience de-sensing due to the high RF environment. As the engineer said, due to coupling issues, etc., it might not be so simple as it appears "on paper." Also there could be frequencies where it works just fine and others where it falls apart. It will require some testing, and, perhaps, extra shielding. However, I think duplexing is a wonderful idea.


Jim, K6QE

Photo of Ken - NM9P

Ken - NM9P

  • 4239 Posts
  • 1351 Reply Likes
I can imagine that this would be an issue when transmitting on HF, but how about when using the transverter and RX antenna ports?  It seems to me that the levels would be fairly small in this case?  I would think that most of the Full-Duplex users would probably be using transverters for satellite operations.
Photo of Jim Gilliam

Jim Gilliam

  • 462 Posts
  • 71 Reply Likes

In a very high RF environment, frequencies do not become an issue so much as the intensity of the signal coming into the receive antenna from the transmitting antenna. This intensity is a function of the proximity of the receiving and the transmitting antennas. When a signal reaches a certain level, regardless of frequency, it can cause considerable change in the linearity of the receiving circuitry. I have seen this happen time and time again. It might not be an issue, but it certainly would require some rigid testing before Flex would give it a seal of approval.


Jim

Photo of Iain McLachlan

Iain McLachlan

  • 39 Posts
  • 2 Reply Likes
Proper multi-pole filtering on output and input with no overlap of passband will fix this. How do you think companies share antenna masts? 
Photo of Sergey, R5AU

Sergey, R5AU

  • 860 Posts
  • 117 Reply Likes
Iain, i am with you,  but the issue that not everyone read manual completely and understand fully this is in one side on the other way around sometime user expectations are over of the level of the consideration , and i think you know the result  - dissatisfaction. i Guess FRS will develope it and give us also ability to say our wishes.
Photo of Jim Gilliam

Jim Gilliam

  • 462 Posts
  • 71 Reply Likes

How does one filter a receiver whose input is as wide as a barn door and then some? Yes, pre-filtering would probably work, but we weren't talking about putting an array of cavities in the ham shack for casual duplexing.

Photo of Iain McLachlan

Iain McLachlan

  • 39 Posts
  • 2 Reply Likes
That may be the answer though. The question was notably regarding the 6300. The 6700 however has pre-selection built in, but for the cheap option, and that is just what the 6300 is, there are many relatively cheap used 30MHz LPFs available on the open market, mainly because they have been in use for so long. That would cover the 10m side. Remembering that these work both ways so harmonics would be heavily attenuated on the way out, and anything over 30MHz wouldn't make it back in one piece. A cavity at the 6m side, granted more expensive, will provide the same for that side. A couple of decent cavity blocks will still be cheaper than the 6700, although, of course, won't make it a 6700!

The issue with internal RF is another issue however. I have no idea of the quality of electrical and magnetic separation off the tuners inside, though that may prove more of a show-stopper. Although, their may be a maximum recommended RF level that may suit. I would like to see the duplex option also, but I suspect that this was looked at by FRS more than they are letting on, so the limitation may very well be inside the radio.
(Edited)
Photo of Wayne, W5XD

Wayne, W5XD

  • 18 Posts
  • 6 Reply Likes
To reply to an old thread...I recently started using a 6500, and the behavior I see duplicates what Ron reported on his 6300 in his original post many months ago. And I would like to repeat at least part of the original question: is the lack of ability for full duplex transmit while receive a permanent feature?

Similar to Ron's experiment, I configure two slices in SmartSDR, put the TX on slice B at 14MHz CW on ANT1 into my dummy load, and RX on, say, 7MHz CW, on slice A with RXA, which I have cabled to an RX antenna. When I go key down on the 6500's front panel Key jack, the 7MHz panadapter appears to continue to receive. But I can't find any audio output coming from Slice A anywhere. There are lots of places the 6500+SmartSDR might hide the audio--the front panel PHONES jack, the back panel speaker jack, the SmartSDR "REMOTE" button to the PC speaker and Slice A on DAX channel 1. If there is audio in there anywhere, I cannot find it. (I fished around the CW sidetone this way, too--it hides in the flex box itself when using the front panel KEY jack. But if you use CWX, sidetone sneaks out to the REMOTE channel, but never through the DAX--but that is very different question.)

So I repeat Ron's original comment to the effect that there doesn't appear to be any fundamental issue that prevents demodulation of a signal from RXA on one Slice while transmitting on another. The hardware diagram makes it appear that the receive signal flow, at least for the 6500 and 6700, is completely separate when receiving via RX-A and transmitting on ANT1. (The 6700 diagram indicates it should be able to decode concurrently from two antennas, not just one, and with the same caveat that, during a transmission on one of the ANTn jacks, at least one of the RX antennas would have to be on RX-A or RX-B. I haven't found a 6300 hardware description, however, so I can't guess for it.)

There is lots of discussion on this thread about RF isolation. I am sure that is a performance issue to be reckoned with. But getting a stream of zero's from the Slice A demodulator is not a performance issue. Somewhere, a design decision was made to silence the the Slice A demodulator during transmission, and the question remains, is that a permanent limitation? or will there be an upgrade to change that?

Wayne, W5XD
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9188 Posts
  • 3550 Reply Likes
"And I would like to repeat at least part of the original question: is the lack of ability for full duplex transmit while receive a permanent feature?"

There is no lack of ability; it is disabled by design in the software. As Steve mentioned above, it primarily depends on the capabilities of the hardware. Just because it is doable in the hardware doesn't mean it will necessarily become a feature available to users especially if it may not behave in a consistent and reliable manner.  There are additional business concerns too, such as the supportability of said feature just to name one which is an actual expense to the bottom line.

With that said, we are planning on doing testing to access the viability of using a 6000 in full duplex mode.  The activity just hasn't been scheduled yet.
Photo of Gerald - K5SDR

Gerald - K5SDR, Employee

  • 830 Posts
  • 1514 Reply Likes
It is not a permanent limitation on any FLEX-6000 transceiver.  The hardware is capable.  Stay tuned but I will not commit to a release date.
Photo of Lee

Lee, Elmer

  • 680 Posts
  • 287 Reply Likes
heck with all these panadapters how can we not stay tuned
Photo of Sergey, R5AU

Sergey, R5AU

  • 860 Posts
  • 117 Reply Likes
I believe Full Duplex will be a part of the ver. 2.0 or early with 6700
(Edited)
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9188 Posts
  • 3550 Reply Likes
FDX will be earlier, but you didn't hear it from me ;-)
Photo of Ross - K9COX

Ross - K9COX

  • 346 Posts
  • 107 Reply Likes
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1057 Posts
  • 1097 Reply Likes
Official Response
We will be introducing Full Duplex operation in v1.5 which is due out in September.  I'd like to spend a few minutes talking about the specifics.  Full Duplex will allow you to transmit and receive at the same time, but there are limitations.  The FLEX-6700 has the most capability in this area and you will be able to transmit on one antenna and receive on up to two different antennas at the same time.  When you receive while you are transmitting, there are a number of important considerations to ensure good operation.  Here are the considerations:

Isolation and overload: If your transmit energy enters the receiver at sufficient levels with the preamp off (greater than about 5mW in the FLEX-6300 and greater than about 10mW in the 6500 and 6700) or with the preamp on (generally subtract the gain of the preamp from the provided overload point), the radio receiver will overload.  The FLEX-6000 Signature Series radios have a "soft overload" where they will show signs of an overload before going into a complete overload some 5-10dB later.  The key sign of a soft overload are numerous spurs in the panadapter.  There is 40-100dB of isolation inherent in the radio itself.  This will be added to the isolation between your antennas and the total must be below the radio overload point.  More details will be provided, but here's a math example:

Transmitter: 100W = +50dBm
Overload on FLEX-6700 = +9dBm
Isolation from ANT1 to RXA: 90dB
Antenna isolation (at your shack): 40dB

+50 - 90 - 40 = -80dBm ... well below the +9dBm limit (89dB margin)

Transmitter: 100W = +50dBm
Overload on FLEX-6300 with 20dB preamp on: +7dBm - 20 = -13dBm
Isolation from ANT1 to ANT2: 40dB
Antenna isolation (at your shack): 30dB

+50 - 30 - 30 = -10dBm ... we are 3dB above the overload with the preamp on 

We'll show you how to run these calculations, but you can see there is a wide range of possibilities so you will need to do a little math to determine if everything will work for you.  Bandpass filters may also add another 30dB or so of isolation of you are transmitting on one band and receiving on another if your antenna selections are correct.

The radio will also protect itself if you put too much energy in the receiver port and the radio will disconnect the antenna and alert you.

Receiver must be on different frequency from transmitter: The delay that occurs during filtering is guaranteed to drive someone mad if the transmitter and receiver are on the same frequency.  For this reason, we mute the slice that you are transmitting on so you will not hear your own signal.  This may seem counter intuitive at first, but it should fit most use cases.  For example, if you are chasing DX, you can create two slices, a RX slice on the DX and a TX slice where you will TX up (generally) and when you transmit, you can still hear the DX frequency.  You will not hear under your current transmit signal, unless you are in QSK in which case you will hear in-between your transmissions.

Antennas: In all cases, you will need two antennas for full duplex.  We will not be transmitting and receiving on the same antenna.

This is an exciting and new capability for all FLEX-6000 radios, but it will require a little planning and thought to be sure it will work at your location.  We hope this will bring hours of new operating fun!

=================

EDIT:
9/17/15 There are a few errors in the calculations above and rather than just change the text, I'll explain the issue and provide the correct information.  There are really two paths from the transmitter to the receiver.  The first path goes from the transmitter to the transmit antenna, to the receive antenna and into the receiver.  The second path goes from the transmitter directly to the receiver inside the radio.  It is the latter path that prevents most manufacturers from offering full duplex.

For the scenarios above, you should always look at the isolation in the radio AND the isolation outside and take the lower of the two.  

SCENARIO #1
Antenna Path:
Transmitter: 100W = +50dBm
Overload on FLEX-6700 = +9dBm
Antenna isolation (at your shack): 40dB
Receiver Preselector Filters: 20dB (assuming different bands)

+50 - 40 - 20 = -10dBm ... this is acceptable 

Internal path:
Transmitter: 100W = +50dBm
Overload on FLEX-6700 = +9dBm
Isolation from ANT1 to RXA: 90dB

+50 - 90 = -40dBm ... this is also acceptable

So in scenario #1, the configuration will work with full-duplex

SCENARIO #2
Antenna Path:
Transmitter: 100W = +50dBm
Overload on FLEX-6300 with 20dB preamp on: +7dBm - 20 = -13dBm
Antenna isolation (at your shack): 30dB
Receiver Preselector Filters: 0dB (not available in FLEX-6300)

+50 - 30 - 0 = +20dBm ... we are 33dB above the overload with the preamp on.  More antenna isolation is needed

Internal path:
Transmitter: 100W = +50dBm
Overload on FLEX-6300 with 20dB preamp on: +7dBm - 20 = -13dBm
Isolation from ANT1 to ANT2: 40dB

+50 - 40 = +10dBm ... we are 23dB above the overload with the preamp on 

In this case, full duplex may not be run.

More details will be made available in a white paper on Full Duplex and SO2R.  Hopefully I haven't made too many mistakes in the corrected text!
(Edited)
Photo of Sergey, R5AU

Sergey, R5AU

  • 860 Posts
  • 117 Reply Likes
Tim, Steve, I am excited , I will not be limited in any crossbands operations ex.: like SAT
(Edited)
Photo of Stan - VA7NF

Stan - VA7NF

  • 459 Posts
  • 111 Reply Likes

@Steve, now the kimono is slightly opened,

Which scenario does the XVTR port fit.  Can you listen to one transverter while transmitting on another as XVTR Tx levels do not exceed the Rx protection levels?  This may be the satellite environment unless one Rx is put on 10M.  Or even Tx on 8 XVTR frequencies concurrently, subject to how wideband is the XVTR hardware?

Special thanks for the pre-release information.  BTW you didn't handle the duplexer(triplexer) which would support cross band splits on the same antenna because of the two(three) separate radio feeds; each of which would be on separate antenna ports.  SO8R on the same radio!

My next Field Day challenge is using a N1MM script to call CQ on the same (2M/6M) frequency in CW, SSB, then RTTY then listen for the reply mode and log it.  OK, I recognize I'm crazy but for FD operations that may be a virtue.

Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1057 Posts
  • 1097 Reply Likes
Transverters are generally wired as "common IF" or "split IF."  A common IF transverter has an internal switch and transmits and receives on the same IF port.  Since this requires that you transmit and receive on the same antenna of the FLEX, it will not work.  You must have a split IF transverter.  

If you have a split IF transverter, you will need to hook the IF transmit and IF receive antenna ports up to different ports on the radio.  On the FLEX-6700 and FLEX-6500, you can hook the transmit to the low power output XVTR port.  The receive port should be hooked to RX-A or RX-B on a FLEX-6700 or to RX-A on a FLEX-6500.  

I don't see an easy way to do full duplex with a transverter on a FLEX-6300.  You could conceivably use the ANT1 port for receive, but in my microwave station I have a rule that nothing can be hooked up that can physically damage anything if there is operator error.  If you hook a transverter receive IF port to ANT1 on a FLEX-6300 and you forget and send 100W into your IF port, you will damage the transverter.  So I do not recommend this.
Photo of Ron Parsons

Ron Parsons

  • 73 Posts
  • 10 Reply Likes
I have been doing exactly what you describe in your third paragraph with my 6300. Because of the danger you describe, I have automated the switching of the split IFs of my transverters using coax relays controlled by DDUtil and a USB-8 relay board.

Although no automation can guarantee error free operation 100%, using DDUtil (and soon to be replaced with a more flexible control logic), the mode of the satellite, U/V, V/U, etc. can be detected and ensure ANT1 is connected to the RX side of the transverters and the XVTR port is connected to the TX side if the split IFs of the transverters.

I look forward to full-duplex on my 6300 satellite system. Ron W5RKN
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 3894 Posts
  • 949 Reply Likes
Wow, we didn't see this coming!!