1) DIRECTV would have to raise the data through-put rate of a 36 MHz CONUS+ beam tp. from ~40 to 45 mb/s for a 3 UHD channel multiplex.
2) 2 bonded tps. means one tp. used for regular HD/SD reception is inefficiently wasted as sealed off and not recoverable during 4K reception.
1) If they wanted 30 Mbps for each they'd have to raise the transponder throughput to 45 Mbps. That means using QPSK 3/4 instead of QPSK 2/3. That requires only 0.93 db of increased signal margin, so it would hardly affect its resistance to rain fade.
2) Yes, but it would depend on the way they're implementing bonding. They've cooked up their own solution for bonding that's apparently software based, since they are continuing to use DVB-S2 hardware which offers no native hardware support for bonding like DVB-S2X does. I've surmised before that maybe the 'wasted' transponder is a limitation of the way they're doing it, but wouldn't necessarily be a permanent limitation. If they use DVB-S2X tuners in the HS17 follow-on, I'm sure Broadcom and Maxlinear would provide Directv a way to use the hardware bonding in their non-standard way if they requested it, similar to how they support Directv's hacked version of DVB-S2 as well as DSS.
I can't find the quote at the moment, but I'm 100% sure Phil Goswitz specifically said Directv would be bonding two transponders together to carry three 4K channels at 30 Mbps each. Obviously plans can be changed, and maybe they have if K9SAT (and apparently others in the testing group) are correct about this. Has anyone been involved in any actual testing of bonding so this can be treated as a known fact, or were they just told this? I guess what I'm getting at is whether they were specifically told that three transponders would be broadcast in a bonding triplet, or just that the receivers
used three transponders for bonding? The former is not the same as the latter, if the implementation works the way I previously outlined at dbstalk.
I just don't think using three transponders makes sense for a couple reasons. First, bonding three transponders to carry four 4K channels is less efficient. Granted, they only go from 54 to 48 4K channels of capacity, and the way 4K is going nowhere we may never have even close to 48 4K channels so it might not matter...
Second, and more importantly, needing three transponders (and therefore three SWM channels) per 4K channel really limits receiver hardware. What happens when there are actually real 4K channels to watch, and people have all 4K TVs in their house, so people will want to watch/record 4K on every TV? If you need three SWM channels per 4K channel, a RB LNB's 21 SWM channels can serve 7 such channels which fits perfectly with 7 tuners (assuming a follow on to the HS17 that has three 8 tuner chips instead of two, and outputs 7 4K streams) Then what the heck is up with the DSWM30 supporting only 15 tuners per output?? People using those couldn't grab enough tuners to watch 7 4K streams, are those living in MDUs just screwed? OK, I've always said I expect Directv to eventually permit more than one HS17 per account, but until they do we can't assume they ever will.
That said, it seems there's now more evidence for three transponders bonded than two, since that's just based on Goswitz's statement from a few years ago, so that's probably how it is going to be. It doesn't make sense to me, but then a lot of what Directv has done the last few years make no sense. Why does the reverse band LNB drop down to 13 SWM channels when "certain hardware" is connected, even though all SWM capable hardware handles the full frequency range required for 21? Why is the DSWM30 limited to 15 tuners per output even though the chips inside it support more - the same chip as used in the reverse band LNB - so it could easily support 21 per output (more than that, actually...) I have to assume there's some method to this madness, but so far none is evident. Watching their decisions is like reading the news from Washington, so many things just make you say "WTF?"