SIM5320 CMUX problem, Having problem implementing multiplexer with SIM5320 Опции

[bernarduh]

Hello.

I'm prototyping a board that uses a SIM5320. Currently i have a prototype that is working with the uart interface without multiplexer.

My next and crucial step is to get the uart interface to use the multiplexer function so that i can have 3G/GPRS and GPS at the same time.

The module is connected through a null modem connection (RTX -> TXD/ TXD -> RXD).

To implement the 07.10 Multiplexer protocol i'm using the Linux kernel driver n_gsm. Currently i have already modified this driver in order to get some comunications out of the SIM5320 in CMUX mode.
I'm able to establish the DLC0 but after that the only thing that the SIM5320 sends is sequences of F9 F9 F9 F9. After a lot of digging i found out that the module can send this flags in a few situations:

Indicate the connection is established.
Module has wake up from sleep mode.
Module is out of sync and is discarding the frames until i send it F9 F9 F9 F9.

I already tried to send it back the F9 F9 F9 F9 but the module send them back :s.

ATI

Manufacturer: SIMCOM INCORPORATED
Model: SIMCOM_SIM5320E
Revision: SIM5320E_V1.5
IMEI: 861311000924364
+GCAP: +CGSM,+DS,+ES

OK

Currently this is the only issue stopping me from sending the board for PCB production so this is kind of urgent.
I'm really need help on this topic.

Thank you for your attention.

Sorry for posting in English but i Portuguese and can't speak Russian.

[insektazz]

Hi. What OS are you using? Or maybe it's just a micro controller?
In case of Linux, here is a good example how to run and use CMUX mode
https://github.com/Rtone/cmux

[bernarduh]

I'm using Linux. I have tested that precise script but no luck. I finally did some progress by using another daemon (http://sourceforge.net/projects/gsmmux.berlios/). I think that the problem with the Linux driver n_gsm is that it creates the DLC0 (Data Link Connection for control of the MUX) but then only creates the other DCL on demand (when we try to open the virtual ttyGSM1/2/3/4) and the SIM5320 module expects the DCLs to be open right away.

[insektazz]

I'm using mentioned script and it work pretty good.
In my configuration ttyGSM1 used by pppd, ttyGSM2 for some scripts to gather signal quality, send SMS, etc - all this thinks work simultaneously. From time to time I manually use ttyGSM3 and 4 to do some experiments.

Could you describe your issue from the user's perspective?

[bernarduh]

I launch the script and it creates the ttyGSM1/2/3/4 but when i try to use any of the ttyGSM i get "[Errno 51] Level 2 halted".
So i started digging the n_gsm driver source and found out that the tty_open return that error if the DLC for that port was not establish.
I enabled the debugging to see what was been sent and received from the module and found out that the n_gsm sends:

F9 03 3F 01 1C F9

to the module to establish the control DCL0 and the module replies with:

F9 03 73 00 00 A4 F9

accepting the connection and also sends:

F9 F9 F9 F9 F9 F9 F9 F9

after that.

At this point the n_gsm driver does nothing until we try to open a MUX channel (ttyGSM).
When we do that the module only replies with:

F9 F9 F9 F9

What module are you using? Is it the same SIM5320E?

[insektazz]

I'm using SIM900R.
I do not have deep experience working with CMUX. I wish I could help you more, but have no idea what these bytes patterns mean. =(
If any idea come to my mind I'll let you know.

[bernarduh]

Thanks anyway. At least i manage to get it working with the order daemon. May be later i'll get back to the n_gsm driver and run some more tests to figure out why doesn't the n_gsm work and the order deamon does.