[Gta04-owner] Invitation to indicate interest in next batch of GTA04A5 boards
Dr. H. Nikolaus Schaller
hns at goldelico.com
Tue May 21 10:44:23 CEST 2013
Am 19.05.2013 um 14:31 schrieb Sven:
> Hello Nikolaus,
> some concrete recommendations for GD, resulting out of my explanations:
> a) Design the plate of the GTA04A5 so, that the variant that is equipped with B2B connectors instead of display also has one further connector instead of the modem.
> b) Make a little board equipped with the modem so, that it can be connected to the modem connector of this GTA04A5.
> c) Offer the handheld variant of the GTA04 ("OpenPhoenux 3704") in this way. There should be enough place within the case in order to do so.
Originally we had experimented with modems from Telit which have such B2B connectors. But there was simply no space and there was no connector that we could have used for the antenna (you can't have RF signals go over the same type of connectors as the digital signals). So we did choose the OPTION GTM601 which is a LGA soldered to the PCB.
> d) In any case where you advertise the GTA04A5, never forget to mention that it is intended to develop the gravity to attact a new community of developers, who replace your little modem board by a self-made one, equipped with a new open source LTE modem.
> e) Also never forget to express the dream, that then this device would have the potential to develop the gravity to also attract a community of developers, who write the open source firmware for it. If this dream would come true, then place the offer that GD is willing to take the results to create a GTA05 out of them, to get the needed certificates and in the end: To sell it to everyone who wants to have it.
> f) Never forget to argue, how good the initial situation for such a project is: GTA04A4 equipped with QtMoko v55 should be convincing, that a very big amount of work to get such a new GTA05 phone has already been done!
I don't want to discourage you to develop an open source LTE modem.
But there are some things to consider.
1. you need a baseband chip. MAX2580 and MAX2553 look good - but it starts with the datasheet being available on request only. This means that you have to convince them that you are a customer with big quantities so that it is interesting enough for them to talk to you.
2. LTE protocols are not easily written in one weekend. Expect a project with first results after at least 2 years, maybe 5.
3. at that time there will be 5G
4. you can't develop such protocols and test them on the real air waves. I.e. you need a network simulator e.g. from Rhode&Schwarz. Or buy a Ferrari from the money you need to invest.
5. to legally operate such a device you need a licence and certification. Expect an amount equivalent to several Ferraris to finally get through this process...
So if you organize to develop and certify such a device, we will happily integrate it into a future GTA0x.
It is hard to say, but for me tinkering or opening the modem is a nice dream but unrealistic. And therefore not my (or GD's) goal.
IMHO it is better and much easier to make the system secure and safe by design. E.g. if there is no connection with GPS and if all communication is going through ssh, the modem (and the network) has no chance to do anything harmful. It can only *transfer* the encrypted data... And as long as we have a 2-processor architecture, the modem can't control the application processor. I.e. the (unknown) firmware inside can try to do strange things - it may only drain the battery but not my data.
> Best regards
> On 18.05.2013 16:46, Sven wrote:
>>>> What we could do is have a special offer at the next Fosdem. If
>>>> _everyone_ (thousands of people...) (or 90% of the people there)
>>>> pledges an $x amount of money, they will get a Free (as-in-speech)
>>>> phone. That way, you can probably cut down the production cost to a
>>>> 'reasonable' amount (300-250 EUR apiece?). If there's not enough
>>>> interest, nobody loses money, but if there is, the cost per phone is a
>>>> lot lower than in a batch of 200-300 phones.
>>> Yes, sure. I think there are 5000 participants every year and 90% would
>>> likely allow to reduce the price to 300 EUR most likely including an injection
>>> moulded case.
>> taking into consideration all facts and arguments from this list, I come to the following personal market analysis:
>> Sold GTA01/GTA02: approx. 18k
>> Percentage of those, who have been frustrated by them so much, that they will never again buy an expensive phone, that in fact is of prototype state: at least 70%
>> Percentage of those of the rest 30% or less, who are willing to get an upgraded phone with more power, more features and more reliability by the cost of forgetting the idea of having a really free and open modem: Take 300 sold GTA04 as basis of assumption and come to the result: 5%.
>> I claim that's realistic.
>> Now estimate possible sales rate of GTA05. Take two variants:
>> a) GTA05 that is based on GTA04, but with OMAP5, DVB-T, external accessible interfaces like I2C, perhaps also another display, but again with closed source modem: Less than 100.
>> b) GTA05 that is mostly the same as GTA04. Only improvement is to have an open source LTE modem (hardware and firmware): Approx. 2k or 3k.
>> Both variants already include the assumption that GTA05 also has a better soldered MMCX connector.
>> So I assume that it's clear that within ordinary capitalistic ways of thinking we'll never get any GTA05. But before making more thoughts, I'm asking you to comments to these statements.
>> Best regards
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
More information about the Gta04-owner