Print

[W9AC]

The IC-7610 does have a predistortion input.  The ALC jack is dual-purposed and is directed by an internal relay (see IC-7610 Service Manual and follow signal line "PW2FB").  Presumably so named for "PW2-Feedback." 

The default position is as an ALC input from any amplifier.  In the other relay position, it blocks DC from the jack and passes RF through a 0.1 uF capacitor, untimely leading to the receiver's ADC input for linearizing during transmit.  New firmware will open up a sub menu, allowing control of the internal relay.

What's not clear is the amount of preconditioning accomplished within the PW-2 and inside the IC-7610.  And to Glenn's point, if a new jack will be added to the back of the final version of the PW-2.  Or, will one of the amp's ALC jacks also be dual-purposed?

Paul, W9AC

[W9IQ]

That is an interesting piece of the puzzle, Paul. A very nice find! Thanks.

- Glenn W9IQ

[W9AC]

I can't take credit for the find.  It was posted on the IC-7610 forum. 

Paul, W9AC

[VR2AX]

Does it represent a move forward in technology?

[AC2RY]

The IC-7610 does have a predistortion input.  The ALC jack is dual-purposed and is directed by an internal relay (see IC-7610 Service Manual and follow signal line "PW2FB").  Presumably so named for "PW2-Feedback." 

The default position is as an ALC input from any amplifier.  In the other relay position, it blocks DC from the jack and passes RF through a 0.1 uF capacitor, untimely leading to the receiver's ADC input for linearizing during transmit.  New firmware will open up a sub menu, allowing control of the internal relay.

What's not clear is the amount of preconditioning accomplished within the PW-2 and inside the IC-7610.  And to Glenn's point, if a new jack will be added to the back of the final version of the PW-2.  Or, will one of the amp's ALC jacks also be dual-purposed?

Paul, W9AC

If this is the case, then solution may be extended to use with other amplifiers.

You made me look at IC-7610 manual. What I found there is even more interesting: the same circuit that processes PW2FB signal, is processing sampled RF signal from embedded amplifier called RFMON. Thus if software can implement pre-distortion for external amplifier, it should be able to do the same for power amplifier inside transceiver. Looking at measured IMD numbers I do not think we have that feature now, but if developers will do it for PW2, there is nothing stops them from offering it for standalone transceiver.

But I am still concerned if ADC in 7610 is fast enough (130Msps) to be used in digital feedback loop for 50MHz amplifier.

[VR2AX]

How should we compare with:

- make AB exciter more linear;
- make amplifier no less linear than AB exciter

?

[W9AC]

Just to clarify my obvious typo:

"...RF through a 0.1 uF capacitor, untimely ultimately leading to the receiver's ADC input for linearizing during transmit."

Paul, W9AC

[AC2RY]

How should we compare with:

- make AB exciter more linear;
- make amplifier no less linear than AB exciter

?

If they use digital feedback loop, then both exciter and amplifier will be inside the loop. Thus pre-distortion will be applied for the whole chain. If pre-distortion does not use feedback, but LUT for forward correction, then only amplifier non-linearity will be compensated. They also may have a compromise with self calibration cycle to create LUTs rather than pre-defineds ones. This also will work with the whole chain of exciter plus amplifier.

[VR2AX]

Accepted, WP, but doesn't the same principle apply to the basic exciter, or our normal transceiver with usually 3 internal amplifiers, to produce ~ 100 watts output

[AC2RY]

Accepted, WP, but doesn't the same principle apply to the basic exciter, or our normal transceiver with usually 3 internal amplifiers, to produce ~ 100 watts output

It does. But the technology used here is only useful with direct sampling transmitter. Any direct sampling transceiver can be easily equipped with DPD. Cost of additional circuitry is very low. Looks like IC-7610 already has this circuitry in place.

[W9IQ]

From what I understand, the 7610 in a standalone situation is not equipped to predistort its own signal. This is because it lacks RF sampling of its own output. Not that this should matter as the 100 watt output is not likely to wake the IMD purists

When mated with the PW2, the 7610 will have an output sample from the PW2 - likely via the PW2 ALC jack that will serve a simular dual purpose as the 7610 ALC input. Then the 7610 can adjust its output such that the resulting output from the PW2 has the lowest possible spectrum footprint.

Hats off to the Icom engineers for having the foresight to have the necessary hardware in the 7610, just waiting for a firmware update to turn it on.

- Glenn W9IQ

[AC2RY]

From what I understand, the 7610 in a standalone situation is not equipped to predistort its own signal. This is because it lacks RF sampling of its own output. Not that this should matter as the 100 watt output is not likely to wake the IMD purists

- Glenn W9IQ

You are incorrect here. 7610 has sampling circuit. This is the same circuit that is used for external amplifier sampling. There is a software switch that selects which sampling is used. Look at https://icomuk.co.uk/files/icom/PDF/newsFile/IC-7610_Technical_Report_Vol2_Eng.pdf  Page 10 in Block Diagram shows both sampling inputs: PW2FB and RFMON. The latter comes from tuner board with signal taken right after band pass filter (Page eight) and likely can be used to sample output of internal amplifier.

What it might not have is an ADC fast enough to have closed loop correction. But it should be enough to have open loop pre-distortion with periodic calibration. It may have enough capabilities to remind you for required re-calibration when it observes too much deviation from expected results.

[W9IQ]

From what I understand, the 7610 in a standalone situation is not equipped to predistort its own signal. This is because it lacks RF sampling of its own output. Not that this should matter as the 100 watt output is not likely to wake the IMD purists

- Glenn W9IQ

You are incorrect here. 7610 has sampling circuit. This is the same circuit that is used for external amplifier sampling. There is a software switch that selects which sampling is used. Look at https://icomuk.co.uk/files/icom/PDF/newsFile/IC-7610_Technical_Report_Vol2_Eng.pdf  Page 10 in Block Diagram shows both sampling inputs: PW2FB and RFMON. The latter comes from tuner board with signal taken right after band pass filter (Page eight) and likely can be used to sample output of internal amplifier.

That's great. Better yet!

Thanks much.

- Glenn W9IQ

[KA4WJA]

Glenn, Paul, Alex, et al,

1)  Wow!  Thanks to everyone here for actually looking at the '7610 closer!
It looks like there is some good news here and some bad news as well.   
(and, a bit of embarrassing news for me...I ignorantly thought that Icom might be incorporating pre-distortion into their PW-2 amp, when it looks like they're just doing what other manufacturers are doing ---- using some dramatic-license in marketing/ad copy?)

{BTW, I'm not surprised that Icom had the forethought to equip the '7610 with hardware(?) for future use / expansion....especially for pre-distortion, as some Japaneses manufacturers (JRC) have been making commercial (for aviation ground station and maritime coast station, etc.) HF rigs with pre-distortion for a few years now, with IMD3 of better than -45dbc (and the JRC's maintain that at 1kw, 5kw, and 10kw output Solid-State)...} 

The IC-7610 does have a predistortion input.  The ALC jack is dual-purposed and is directed by an internal relay (see IC-7610 Service Manual and follow signal line "PW2FB").  Presumably so named for "PW2-Feedback."

You made me look at IC-7610 manual. What I found there is even more interesting: the same circuit that processes PW2FB signal, is processing sampled RF signal from embedded amplifier called RFMON. Thus if software can implement pre-distortion for external amplifier, it should be able to do the same for power amplifier inside transceiver. Looking at measured IMD numbers I do not think we have that feature now, but if developers will do it for PW2, there is nothing stops them from offering it for standalone transceiver.

But I am still concerned if ADC in 7610 is fast enough (130Msps) to be used in digital feedback loop for 50MHz amplifier.

7610 has sampling circuit. This is the same circuit that is used for external amplifier sampling. There is a software switch that selects which sampling is used. Look at https://icomuk.co.uk/files/icom/PDF/newsFile/IC-7610_Technical_Report_Vol2_Eng.pdf  Page 10 in Block Diagram shows both sampling inputs: PW2FB and RFMON. The latter comes from tuner board with signal taken right after band pass filter (Page eight) and likely can be used to sample output of internal amplifier.

What it might not have is an ADC fast enough to have closed loop correction. But it should be enough to have open loop pre-distortion with periodic calibration. It may have enough capabilities to remind you for required re-calibration when it observes too much deviation from expected results.

It looks like the bigger news is that Icom should have pre-distortion-equipped IC-7610's for sale sometime next year! (and, their new 65v LDMOS, 100% duty-cycle, 1kw amp will have a RF sampler port built-in as standard equipment, for use with the '7610!)



2)  The good news here:

a)  it appears that Icom will have pre-distortion-equipped IC-7610's sometime next year?  Slated to be ready for roll-out at Dayton 2023....and available for purchase 2nd / 3rd quarter 2023... (perhaps these will be an IC-7610MkII? with more powerful DSP processor?  or maybe the current '7610's will just need a firmware update?)

b)  it also appears that Icom does have this working in the lab, as they show 20m two-tone test spectral scans of with and without their DPD ("Digital Pre Distortion"), in the IC PW2 brochure.  (they just have to make it "ready for primetime" and reliable enough / easy enough for the masses of ham operators, before rolling out for sale)





3)  The confusing news:

a)  it appears that the Icom PW-2 solid-state amp (while it looks like a nice 1kw, 100% duty-cycle, SS amp)....it looks like it has an RF sample output that a connected transceiver, etc. can use (just like many other modern ham amps have)....but does not actually possess pre-distortion circuitry / software? (I sort-of feel naive for thinking otherwise / believing the marketing/ad copy)

b)  If this is the case (as I'm fairly sure it is), then why add confusion and state that the PW-2 is

"the world’s first DPD as a linear amplifier for amateur radio in combination with the IC-7610."

   I suspect this is because other amp manufacturers also market their amps with RF sampler ports as "pre-distortion-equipped", when in fact they have an RF sampler built-in but it's the transceiver ahead of it that would need to have pre-distortion

[btw, just last month I got into it all here with some of you, 'cuz I balked at the casualness that many of my fellow hams accept those manufacturers using these marketing phrases....as I said back then, sure we here know what WE are talking about, but when the manufactures start using these terms ("pre-distortion-equipped", etc.) and less knowledgeable hams start making assumptions that if they use xyz amp, etc., they're good-to-go 'cuz their amp has "pre-distortion", then we've got another problem to deal with]



4)  So, with this new info regarding the '7610 (thanks Paul and Alex), and reading the PW-2 brochure closely with this new '7610 info in mind....it looks like the PW-2's "pre-distortion" is like other modern ham amps ---- it has a built-in RF sampler port to send the amp's output sample back to a connected transceiver's pre-distortion sample input.


73,
John,  KA4WJA

P.S.  Alex, I wonder if what you wrote above might be one reason I've heard some friends with ANAN 7000's having occasional issues with their PureSignal?

What it might not have is an ADC fast enough to have closed loop correction. But it should be enough to have open loop pre-distortion with periodic calibration. It may have enough capabilities to remind you for required re-calibration when it observes too much deviation from expected results.

I'm not informed enough to know if the ANAN's (PowerSDR or Thetis) PureSignal is run by the connected computer or completely inside the ANAN?  But, either way may be an issue if the ANAN (or computer) isn't quite up-to the task, yes?

I know this is drifting off topic from Icom PW-2....but, your words just made a light go off in my head, as I personally hear at least once / twice a week, one of my friends [running ANAN's] say their PureSignal is off or needs re-booting, etc., or has stopped working, etc.?

[W9AC]

If they use digital feedback loop, then both exciter and amplifier will be inside the loop. Thus pre-distortion will be applied for the whole chain. If pre-distortion does not use feedback, but LUT for forward correction, then only amplifier non-linearity will be compensated...

It's a shame Icom is not clearer on these points but regardless of implementation, their predistortion graph (assuming 10 dB vert. scale) looks comparable to what PureSignal can achieve with ANAN 7000 or 8000 transceivers, Thetis software and LDMOS amp within the correction loop. 

With PureSignal active, I typically see -70 dBc intermod products from my grounded-grid amps with optimal plate loading.  And, about -55 dBc from my SPE 1K-FA amp at full power, including 6m.  Measurements taken with an Anritsu spectrum analyzer match Thetis' own post-correction transmit spectral display.

Paul, W9AC