In some countries the answer would be a definite “no” - because the 2G network has been switched off - in order to redeploy the frequencies for other things perhaps. You should check that for your own country.
The low speeds that 2G offers would be fairly painful for many use cases.
are the LTE-A EM05/06/16/20 available for purchase by the general public?
Currently, for M.2 for module, our main stream product was EM06/12/20, EM06 and EM12 already have sample. EM20 need to wait for to July.
Blockquote
Are the Linux drivers for the previous mentioned modules open source or could they be open sourced?
You just need to modify system some file to install the module driver for Linux. I think it should belong open source.
Blockquote
Can all those modules support voice and VoLTE?
Yes, support.
Blockquote
Is there firmware to be loaded into the device by the operating system and if so is that open source?
Currently, the module will loaded firmware from product line. But it was not open source.
Blockquote
What variants and bands for the EM20 (and the EM16)?
I suppose you ask about EM20 and EM12. Please refer to below detail:
The documents suggest that the EM20 is compatible with the EM16 but what about compatibility between the EM06, EM05 and the EM16/20?
Almost all the Pin functionality are compatible as they have the same form connector.
Most of EU member countries plan to have (by Decision (EU) 2017/899) assigned the 700MHz band (known as APT700 or B28) to mobile operators by June 2020. Therefore, EU is being very positive to this implementation started by the Asia-Pacific Telecommunity.
For example, from what is to be understood (between the lines), Republic of Croatia plans to leave (free up) LTE 700 MHz spectrum by the 30th June of 2020 as it is currently used by its digital terrestrial television (still transmitting 576p, a common “square-pixel” representation of 625/50 PAL/SECAM, under 10-years binding concession agreement). To sum up, coexistence of new services (like IoT) in the 700 MHz band with DTT is (only) possible if DTT is limited from 470 to 694 MHz leaving 703-788 MHz spectrum free for B28 band.
Back to topic, with appropriate firmware support my eventual choice for B28 would be affordable Telit LN941A6-E1 or, if something better needed, Telit LN940A11/LN940A9. @Torrone, yes, for L5 developer board Quectel EG25-G module (Qualcomm MDM9207 Cortex A7 @1.3GHz running Linux kernel) looks promising (including B28 and VoLTE) and good testing choice with Mini PCIe form factor, just like Simcom SIM7100E-PCIE (Qualcomm MDM9215). Price for Quectel EG25GGB-MINIPCIE-S of around 50.00 EUR looks acceptable as well, yet firmware might be questionable, I don’t know. Anyway @36w4r6, I can certainly cope with the fact that Quectel EM20 module for 699.00 USD is not something I need.
I guess my question should have been clearer. Compatible means a lot of things: physically compatible, electrically compatible, software interface compatible.
To minimise the workload on Purism, you would want them to be as compatible as possible.
I think we’ve established that the EM20 is too fast and hence pointless. So probably more interested in the lower spec modules e.g. EM16. (The EM06 already has reasonable documentation for blueskying.)
The Telit LN940A variants do look promising. Thank you for linking them. I couldn’t find any reference to them handling PCM audio. I will admit it has only been a quick look.
Our biggest catch is that we need an M.2 NGFF modem card that handles PCM and VoLTE and works on the bands in our respective countries.
If we could have a global standard that would simplify this whole process.
I have come across many modems that would work brilliantly and be reasonably priced, however, they would need to be mounted to an M.2 daughter board and I believe that to be bound the scope of those needing these alternatives.
The M.2 form factor is standardise. Power pin locations, voltages, board size and mounting points are defined.
Our hurdle is the drivers determine what data comes through which of the data/address pins. The locations of the antenna connectors may also be different.
As I suspect the modem card being used by Purism isn’t custom it should follow most practices relating to M.2 modem cards.
I have the contact details for the local representative from Quectel. Hopefully I can call tomorrow and get some more details.
I doubt many people would need more than the EM06. I like the EM12 because of its global nature. The EM20 would be pointless. The pricing on these cards represent the market they are intended for. These are used in M2M devices.
Here are same User Guides (under Downloads) for Telit M.2 (NGFF) card. Please skim some if any helps in anyway, even to eliminate Telit modem card I would consider this as progress as well. Here are Telit EU Declarations of Conformity so if any precise info needed maybe you may try to get it from Telit Communications S.p.A in Trieste, Italy. @36w4r6, thank you for your understanding and support.
Just to clarify, Any alternative modems that we come up with that work for our carriers will be an additional cost. The EM06 looks to sell for ~$200, the EM16 is ~$399. Will the Librem 5 be sold without a modem at a lower cost or will I have to buy a phone not for my region and spend an additional 200 bucks to bring it to a functional state?
I really like the idea of the Librem 5 and want one but I don’t think I will be able to afford one that works in my area.
That really is a question for Purism. As there is currently 2 options for modems with the Librem 5 I don’t see any problems offering a version sans modem.
This may however be a cost problem for Purism as the price they offer may be dependent on a minimum purchase order of modems from Gemalto.
Do we know the cost of the Gemalto modems for comparison.
I would be good if someone from Purism could weigh in on this.
It has been implicitly suggested that to get good global options, customers go out and research alternative cards. This has occurred but we don’t have enough info to take it further than that.
Edit: I would happily have my shipment delayed in order to get a phone that works in my country.
That’s true but prices fall over time and this project is 6 months late with no end in sight. No one issues open-ended price quotes good for that long.
I’m hoping the Gemalto is a placeholder - a starting point they can revisit as they make progress killing other snakes in this project. No 3G for GSM carriers and few 4G bands is a non-starter.
With all the different carriers in all the different countries they’ve sold the Librem 5 in it’s unlikely they’ll find one or two modems that will make everyone happy. The smart thing to do would be to let knowledgeable users do the research to find the best cards for their use case, come up with a list of cards for carriers by country then bundle our card orders together to make up an economic ordering quantity in a, “modem sold separately” arrangement. I’d be much happier with that arrangement than paying for a Gemalto modem that provides minimal functionality.
@36w4r6 This repo should be what your looking for, if you wanna know details about the devkit electronics. There are no such details for the final phone that i know of. The dev kit has e mPCi slot for wwan but i think they will keep things like the used buses as similar as possible. For example the sim card slot is also extern on the dev-kit.
i’d be interested if there are going to be any L5s sold without the celular modem. it would be nice if the design in future models could accomodate a fully airgapped version (no celular/wifi/bt) - just make it fully modular so people can add whatever they want inside.
There is also an offering using an Intel modem its the Fibocom l830-gl that supports VOLTE Data sheet this offers 300Mbps down, and 50Mbps up.This is on Amazon for ~50$
EDIT: As said, the VOLTE is a future feature, and no mentioned Linux driver.