Icom launches their second radio with digital predistortion liniearization

[w-u-2-o]

It looks to be a hell of a radio. They missed the boat with that obsolete DVI video connector and lack of balanced mic input, but other than that it's pretty sweet (although it undoubtedly has that nasty Icom 2.7KHz TX bandwidth limitation).

DPD discussion starts at 3:45...

[N4QNT]

I understand the pre-distortion feature is proprietary and only works within the radio and the ICOM Amplifiers...

[K1LSB]

It appears they also missed the boat by not phase-syncing their dual receivers. Diversity Receive (as implemented in hardware by Apache Labs) requires true phase synchronous receivers, which Icom doesn't mention anywhere in their blurbage about their new 7760.

Mark

[dj1yr]

of course the two receivers are phase syncronous, they have a reference, two LTC2208, maybe they go for the 2209.

Apache Labs has no diversity, the PowerSDR/Thetis software makes this possible

DPD would also work with other power amplifiers, for this you need the CIV commands of the PW2, which nobody has yet.
Since they work with lookup tables, this is also safer and more stable than under Thetis, where you try to control in “real time”.

René

[w-u-2-o]

Apache Labs has no diversity, the PowerSDR/Thetis software makes this possible

Would you please explain what you mean by this statement? Both ADCs in the Apache hardware use the same reference clock. They do remain phase locked during operation. From the perspective of "diversity" (more properly named two element coherent beamforming), this is merely a small matter of software given the availability of two, separate, RF data streams that are phase locked. Thus there are requirements on both the hardware and software for beamforming and both the Apache hardware and Thetis software meet those requirements.

I'm not super familiar with the 7610 or the new 7760. Do they allow the two receivers to be assigned to two different antennas at the same time? If not, then that would preclude beamforming. If they can be so assigned then it becomes merely a matter of a software update for the radio.

DPD would also work with other power amplifiers, for this you need the CIV commands of the PW2, which nobody has yet.
Since they work with lookup tables, this is also safer and more stable than under Thetis, where you try to control in “real time”.

René

That is very, very interesting info, René. Would you be able to point me to more detailed information on how the Icom DPD implementation works?

PureSignal is perfectly "safe" and stable, and works with any chain of amplifiers automatically. Even within a single amplifier there are many possible operating conditions, all of which could not be characterized in a set of lookup tables. IMHO Warren's adaptive predistortion approach is far superior in it's flexibility.

[K1LSB]

of course the two receivers are phase syncronous, they have a reference, two LTC2208, maybe they go for the 2209.

René

I stand corrected, I do see diversity receive mentioned on Icom's website:

Capture-1.png (45.08 KiB) Viewed 10900 times

Mark

[w-u-2-o]

What kind of diversity reception? Each receiver on a different antenna? If there is no summing of the signals with controllable phase and amplitude then this is merely antenna diversity, which can be spatial, polarization, or some combination of the two. But without the phase and amplitude adjustments what it is NOT is beamforming as found in our radios.

Yes, it's a pet peeve of mine, but they should never have called the openHPSDR beamforming function "diversity", because it is NOT diversity, it is beamforming

[K1LSB]

Scott,

100 percent agree!

Mark

[dj1yr]

With the Apache HW it is the same as with the 7610/7760, both ADCs have a reference source, and different antennas can also be assigned to the ADCs.
Without the right software the Apache HW can't do more and even that it couldn't, because Apache Labs doesn't have its own software and relies on the many software writers out there.

PD per lookup never has the problem that it regulates incorrectly, everyone who is in this scene knows the problem that it can happen that the pure signal generates a very wide signal due to misregulation, which is usually due to timing problems of the network packets.

You don't have this problem with Look up, as the preset offset table is used and is not regulated, but this comes at the cost of slightly poorer IMD regulation.

There were the first versions of PS at HPSDR (Atlas System) times, at that time it was alpha and you had to adjust it manually, at that time I had Pennylane + 3 Mercury EU receivers running synchronously.

There was an OM (HB9EPU), he had created his own version with lookup table and used it with very good results, that must have been around 2014/2015.
I also have a pdf with his experiences, all in German, if anyone wants it PN.

by the way, back in the days when K5SO was developed, it was also called beamforming...

René