WiWavelength

Here is a graph of the Lower 700 MHz A/B block spectrum (in the top 100 markets) that VZW *may* sell.

If VZW does voluntarily divest, I highly doubt that Sprint will participate in the open sale. The spectrum available is too limited (both in bandwidth and in number of markets) and the competition (read: AT&T) is too fierce for Sprint to come away from the sale with spectrum worth 1) the money outlay and 2) the addition of another band class.

AJ

So the final consensus is that the EVO LTE does not support SVDO?

I am puzzled. How did you determine that the consensus was negative? The article states that the EVO 4G LTE supports SVDO. And the table four posts above yours lists all of the simultaneous transmission possibilities.

AJ

But for those that want to see the SVDO discussed in the doc its in the main "SAR Test Report" listed there pretty clearly on page 7.

Here is the extracted chart:

AJ

For reference with respect to ERP, here is how the 4G LTE compares to the original HTC Evo 4G...

Nate, I was hopeful that someone would post that ERP comparison data because it is ripe for discussion (and it saves me some of the work of doing so myself).

That said, you posted Cellular 850 MHz ERP figures. (CEL 0384, 0777, and 1013 are all common CDMA1X/EV-DO 850 carrier channel assignments.) Cellular 850 MHz is not particularly illustrative, as the average EVO 4G LTE will spend relatively little of its life roaming on Cellular 850 MHz.

If instead we compare PCS 1900 MHz ERP, the original EVO 4G has max ERP of 25.0 dBm (CDMA1X) and 24.9 dBm (EV-DO), while the corresponding figures for the EVO 4G LTE are 22.98 dBm (CDMA1X) and 18.44 dBm (EV-DO). The EV-DO 1900 ERP difference, in particular, is massive.

For CDMA1X, EV-DO, W-CDMA, and HSPA mobile uplinks, I generally consider 23 dBm ERP to be standard, <23 dBm to be low, and >23 dBm to be high. In this case, the EVO 4G LTE falls into the low ERP category in all but CDMA1X 1900.

In the next day or so, I will follow up with more ERP comparison data from other devices.

AJ

Don't jump off the bridge, yet.

Rickie, if you do jump off the bridge, please take Roger Entner with you.

AJ

What will raw data speeds be like between tower and device in optimal conditions at 5Mhz

A 5 MHz LTE downlink contains 25 resource blocks for a total of 300 subcarriers. If all subcarriers are operating at 64-QAM, then peak, raw data rate in total is 25.75 Mbps. For typical 2x2 MIMO on the downlink, two spatial channels doubles that to 51.5 Mbps.

But that is including coding and overhead. Peak, actual data throughput will be considerably lower, likely ~35 Mbps.

AJ

no speedtest?

How could we conduct a speed test if Sprint has announced several LTE handsets but has not yet released any of them?

AJ

My main concern is the phone's shape. I loved the wideness of the OG EVO 4G but when they went to the EVO 3D, they reduced the width of the phone. I hope the EVO 4G LTE brings back the wider base. But I'm afraid looking at the pictures that this might also have the smaller width.

The original EVO 4G has a 5:3 or 15:9 (1.67:1) aspect ratio screen. The EVO 3D and EVO 4G LTE have 16:9 (1.78:1) screens. So, yes, the EVO 4G LTE will be narrower.

AJ

Are there any other lte phones on the market that use a single chipset? Or is HTC really breaking new ground with this? In my house we have a slogan, "4G no batt-ery." Like you already said, hopefully a single chipset helps that reputation.

No, not to my knowledge, at least not on the 3GPP2 (cdma2000) side of things. Some AT&T LTE handsets, however, may utilize only one chipset because a dirty little secret in the 3GPP (GSM) ecosystem is that it does not generally support SVLTE. Absent VoLTE, when a GSM based LTE handset utilizes simultaneous voice and data, it falls back to W-CDMA. This is one reason why AT&T will not go directly to LTE; instead, it has to deploy W-CDMA first, even in rural areas, and that is a highly inefficient, wasteful use of spectrum.

AJ

The LG Viper does do SVDO. None of the other handsets do. The Viper will join HTC's Thunderbolt and Rezound as the only SVDO handsets in North America. Implementing SVDO requires 2 CDMA baseband processors. It appears the EVO will rely on the single CDMA baseband built into its S4 SoC processor so no SVDO, but SVLTE should be supported.

Yes, it appears that the CDMA1X/EV-DO and LTE modems in the Qualcomm MSM8960 chipset can operate separately and simultaneously. So, that should enable SVLTE. We had initially assumed that HTC would throw in another Qualcomm MDM, a la the Rezound and the LG Viper. But that second modem would not really add anything beyond SVDO.

Furthermore, because the EVO 4G LTE will rely on only one chipset (and a 28 nm process chip, at that!), it should have excellent uptime on its 2000 mAh battery. The dual chipset configurations for LTE handsets (and WiMAX handsets, too) have been a real drain on batteries, so the single chipset, 28 nm die, and extended battery in the EVO 4G LTE should be multiple steps in the right direction.

AJ

When was band class 26 approved? I only see 1-25 and the TDD classes 33-43 on the official lists... The other question is will the FCC allow a superset of ESMR and cellular since the two bands are governed by different rules (part 90S vs. 22H), which impacts power output limitations that would have to be enforced in some manner.

A few comments...

I do not believe that 3GPP has approved band class 26 quite yet. But that is no matter, as 1) it is slated to be standardized this spring and 2) Sprint already references band class 26 in internal LTE documents.

The FCC does not standardize band classes -- 3GPP and 3GPP2 do -- so the FCC will not have a direct say in the matter. Also, some current band classes already have differential power output limitations. An example of that is band class 12, which applies different ERP regulations on the Lower 700 MHz A block because its uplink is directly adjacent to UHF DTV channel 51.

But none of that, too, should be an impediment because I find no evidence that Part 90 rules will unduly limit uplink ERP for band class 10 CDMA1X or band class 18/26 LTE. True, many Sprint CDMA1X/EV-DO handsets do seem to be biased toward higher ERP with band class 1 (~25 dBm), lower ERP for band class 0 and/or band class 10 (~20 dBm). But that is highly variable from handset to handset and seems to be a function of antenna optimization. For example, the following is an FCC test report for a Sprint BlackBerry that outputs 27 dBm ERP for band class 10.

https://apps.fcc.gov/eas/GetApplicationAttachment.html?id=1521343

So, obviously, band class 10 can allow healthy ERP, and I think that the SMR 800 MHz, Part 90, band class 10 (CDMA1X), band class 26 (LTE) concerns are much ado about nothing.

AJ

I have seen reports that the FCC is looking into why phones cannot use LTE BC 12 in lieu of BC 17. Band 12 (regional carriers) is a superset of Band 17 (AT&T). I see no reason why these two band classes have to be separate.I don't see any way they can do anything about BC 13, however. It is backwards (uplink frequencies higher than downlink frequencies) in addition to having a different duplex spacing (-31 instead of +30 MHz).

Agreed, band class 13/14 (Upper 700 MHz) and band class 12/17 (Lower 700 MHz) will remain separate because of the duplex inversion.

The perfectly solvable problem is VZW and/or AT&T throwing around their anti competitive clout to get OEMs to build devices that are band class 17 (instead of the original, inclusive band class 12), band class 2 (instead of the superset band class 25), or band class 5 (instead of the superset band class 26).

Those are little more than predatory, exclusionary practices designed to squeeze out the competitive carriers by making it more difficult for them to procure compatible devices and nary impossible for subs to churn and take their VZW or AT&T devices to other carriers.

AJ

In theory similar frequencies should work with similar-length antennas, so ESMR and the traditional North American cellular blocks at 850 (and GSM 800) can share an antenna

Indeed, see nascent band class 26, two posts above.

...similarly, everything on 1900 A-H blocks (CDMA, GSM, LTE bands 4 & 25) should work with the same antenna.

Careful, you are confusing band class 2 and band class 4. See previous post.

AJ

Can Band 25 be used in the space of Band 4? All Band 25 is Band 4 + G Block.

Nope, band class 25 is a superset than contains band class 2. Band class 4 is AWS 2100+1700 MHz.

AJ

Slight problem: band class 2 is PCS blocks A-F only; band class 25 is needed to support the G block.

Keep in mind that this is just some analyst's sketch. I would take everything with a huge grain of salt, as I do not trust analysts to fully understand the technical intricacies.

Furthermore, band class 25 as a PCS superset will soon replace band class 2. As the FCC turns its attention to interoperability concerns, VZW and AT&T should not and will not be allowed to continue their parochial, predatory practices to prevent interoperability and stymie competition.

Also, band class 5 is Cellular 850-- it does not include ESMR 800.

Enter band class 26, which is another superset that encompasses both SMR 800 MHz and Cellular 850 MHz.

I think 8 is European 900 Cellular and 1 is MSS.

Correct, band class 8 is GSM 900 MHz, but band class 1 is UMTS 2100+1900 MHz, not MSS. The most prominent MSS ATC bands are band class 22 (DISH) and band class 23 (LightSquared).

AJ

HTC owns the trademarks for all of the EVO branded devices....

Indeed, HTC, not Sprint, controls the EVO trademark, and it is important to make this distinction, as it renders the Droid comparison less relevant. On the other hand, VZW controls (licenses from Lucasfilm) the Droid trademark. So, in the first case, the OEM holds the trademark; in the second case, the carrier holds the trademark. And, like it or not, that makes a big difference in the use of the EVO name vs the use of the Droid name.

AJ

Sgt., an upcoming USCC Samsung handset supports CDMA1X/EV-DO 850/1900 + LTE 700/850/1900/2100+1700 (i.e. AWS).

AJ

Oh, by the way, the MSM8960 does natively support TD-LTE, not that the HTC Jet will likely take advantage of that capability, however.

AJ

Sgt., for a relatively recent parallel (that admittedly may or may not apply until we get the final world on the S4 MSM8960), VZW's HTC Rezound uses the S3 MSM8660 for its CPU + modem chipset and the MDM9600 for its second modem.

In the case of the Rezound, the MDM9600 handles LTE duties, since the MSM8660 lacks LTE capability. If Sprint's HTC Jet requires a second modem, however, the roles could be reversed, with the MSM8960 taking the LTE lead.

AJ

Sgt., unless circumstances have changed with the release of the Qualcomm S4 SoCs, SVLTE and SV-DO always require two radio pathways, hence dual modems. If the MSM8960 can actually support SVLTE and SV-DO without a second modem, then that will be a very nice development toward greater simplicity and lower power consumption.

AJ

And if I am being too coy, feel free to call me out it. Otherwise, if you are enjoying the game, we will let it play on...

AJ

Yes, muffinman. I will stand to be corrected, but I know of no LTE handset, not AT&T nor otherwise, that is capable of LTE simultaneous voice/data (outside of VoLTE handsets, of course). That said, AT&T LTE handsets are still capable of simultaneous voice/data. It is a bit of a paradox, and I made this same distinction above, but no one seems to have seized on it yet. So, I am still waiting for an explanation...

AJ

No, this has no real connection to VoLTE, other than that VoLTE does allow for simultaneous voice/data, though not through MIMO. MIMO is just a way of increasing throughput, a special type of spatial reuse, and simultaneous VoLTE/data does not require MIMO nor separate antennas. In LTE, everything is packetized, and the packets get sorted out via addressing and potentially QoS.

As for a GSM ecosystem version of SVLTE using two separate modems (a la CDMA1X SVLTE), it might be possible. But I am not certain that there is a GSM ecosystem set standard for dual modems, nor do I know of any such devices.

Also, keep in mind that several VZW handsets now support SV-DO, but the iPhone 4S is not one of them.

And the mystery lingers...

AJ

awesome.. glad this was actually news to you guys. cant take all the credit..

Yes, the SID list that you linked was news, not because I had never seen it before, but because that was so long ago (>5 years) that I had completely forgotten about it. There just has been little call for the SID list, as we have not had the geeky excitement of important new SIDs for a major US carrier in a long time. So, I greatly appreciate you returning the list to my attention, and watch for your name up in lights sometime in the next few days.

AJ

lynyrd, no such standardization seems to exist. Even in FCC filings, you will find, for example, "5x5," "5+5," and "5/5." To remain clear and consistent, I always try (outside of Twitter) to utilize this format: "5 MHz x 5 MHz." By referencing the MHz unit of measurement twice, my format avoids the 5x5=25?, 5+5=10?, or 5/5=1? arithmetic misinterpretations.

Also, for reference, FDD spectrum pairings are typically "up x down," not the other way around, as the uplink spectrum is almost always lower in frequency than the downlink spectrum. The notable exception to this in US mobile spectrum is VZW's Upper 700 MHz C block 22 MHz license, which is "down x up." LightSquared, too, is/was going to utilize a "down x up" configuration in its L band ATC 1600 MHz spectrum.

AJ