Practice with a Sim800L

[Giuseppe]

Hi everyone, I'm trying to do some practice with a Sim800L red PCB version. 4V power supply with 2200uF capacitor in parallel on the rst pin, a 5.6k pullup resistor + 470uF capacitor. I notice a continuous flashing of the LED and the module does not switch to its normal operation which sometimes it reaches and then loses and then continues to emit a fast flash. Do you have experience on this?

[Bravo]

Hi Giuseppe,
I presume that you have a SIM card plugged in & your area has 2G GSM operating.
Your 4v should have a capacitor like about 470uF as close to the SIM800 as posible.
NOTE that your AT commands should be at baud rate 9600 until you change that.
Also your TTL communication with the SIM800 should be at 3.3v, so you may as well run your PIC at 3.3v.  When the led on the SIM800 board is flashing fast, it has not yet logged on to the network yet. When it is flashing slowly it has logged on to the network. Apart from setting up the baud rate & putting the SIM800 in to AT & text state, do not send any AT commands until the LED is flashing slowly. This can take up to 30 seconds on a bad day.
It is important to have a good 4v supply, although the capacitor should take care of
short term high current demand. I measured the peak current draw of a SIM800 logging
on to the network using my Keithley DMM6500 in digitizing mode. The peak current
was over 3 Amps !!!! 
Good luck with this.

[Giuseppe]

Thanks for replying. On the module soldered onto the pins of the board I put a 2200uF and then I tried a 1000 uF. I already knew that the peak current can exceed 2A. Instead on the Rx module I put a resistive divider to reach a maximum of 2.8-3V so I don't ruin the module. the problem is that the module is unable to connect stably to the network. Sometimes the connection remains stable for minutes, sometimes it's a continuous cycle of reconnection attempts. I also tried another module same problem

[trastikata]

Hello Giuseppe,

- where is the SIM card connected and how is it decoupled
- SIM_VDD needs 10uF decoupling cap,
- SIM_RST, SIM_CLK and SIM_DATA need 10-20 Ohm resistors in series
- SIM_DATA needs 22pF capacitor to ground for stability
- SIM_RST needs 100nF cap to ground for RF noise suppression

- large electrolytic capacitors are too slow, you need to add in parallel few tantalum caps in the range of 330-680uF and one 100nF for RF noise

- long battery power supply cables can serve as an antenna and pick signal which will induce RF noise, I used coax cable for the power supply

- these modules have peak RF signal about 3W - you need a lot of ground for proper operation at these RF levels

- if powered from the same source, the power surges are large enough to cause sufficient ripple in the LDO regulator for the MCU to reset. I usually put a diode in series and a large tank capacitor before the LDO regulator

- at 3W and small ground plane you can have enough noise to the MCU to reset it, place power decoupling capacitors as close as possible to the MCU power pins

[Giuseppe]

Thanks for your precious advice, I'll try later and let you know how it went

[Giuseppe]

Did a bit of testing and put everything together but still not resolved. Every 10 seconds or so the rst automatically goes to 0 even with a 100nF capacitor. I'll post photos and waveforms

https://postimg.cc/xqW5D0Cg (breadbord)
https://postimg.cc/bG7rYFtX (Pin Rst Sim800L)
https://postimg.cc/MfMvW2s8 (Pin Vcc Sim800L)

[John Lawton]

Giuseppe, what is "Vcc" in your third trace, is that "Vbatt" on the Sim800 datasheet (pins 1,42)?  According to the datasheet it should be 3.4V to 4.4V (recommended 4V) rather than the 3.24V - 3.44V shown on the trace.

John

[trastikata]

Did a bit of testing and put everything together but still not resolved. Every 10 seconds or so the rst automatically goes to 0 even with a 100nF capacitor. I'll post photos and waveforms

https://postimg.cc/xqW5D0Cg (breadbord)

Giuseppe, 3W RF output and a breadboard is a no-no, IMO there is not enough ground plane and every pin or wire will serve as a RF antenna and shift the ground. Can you place the module over a large ground plane and connect it with short ground wires?

I am not so experienced with RF designs and there are people here with a lot of knowledge, please correct me if I am wrong but I think this might be the main reason for your resets.

Edit:

Also can you check if the MCU is not being reset every time when the module attempts to connect i.e. transmits at full power.

[Bravo]

Giuseppe,
On my projects which are on PC boards with wide tracks connecting the earth & a 4 amp P channel FET right next to the + power pin with 100uF tant capacitor before the FET.
You will be better off to not use a bread board for the power. Just solder wires to the earth & power pins & put the cap there as well. Maybe you have to have 4v to the SIM800 & 3.3v to the PIC with separate power supplies for testing.

[Bravo]

Giuseppe,
Note that about 20% of those boards are faulty, which is why I test mine on a Rhode and Schwarz CMU200 test set before using them.

[Giuseppe]

Hi, what do you think is it better to focus on this hardware

https://it.aliexpress.com/item/1005003348671279.html?spm=a2g0o.productlist.main.5.127564d28YHAjN&algo_pvid=f16fb3de-b6d2-4dd9-81a4-4a937d1ee895&algo_exp_id=f16fb3de-b6d2-4dd9-81a4-4a937d1ee895-2&pdp_npi=4%40dis%21EUR%215.08%214.63%21%21%215.48%21%21%402103893b17042068264127317e8009%2112000025346021872%21sea%21IT%21871626989%21&curPageLogUid=9n7TpumdNwxK

[trastikata]

Hi, what do you think is it better to focus on this hardware

Hi Giuseppe, I find A6 much easier and less demanding to work with, however if I remember correctly they lack http support and only TCP can be used to connect remotely with a PC.

[evoortman]

Note that more and more providers are phasing out 2G networks.
Are you sure there are still 2G networks at your location?

[Giuseppe]

What other hardware alternatives do you recommend? Always remaining economical but reliable

[david]

Note that more and more providers are phasing out 2G networks.
Are you sure there are still 2G networks at your location?

Indeed.   I thought New Zealand was one of the last locations to still support 2G.  Most of the networks here pulled the plug on it a few years back but one maintained it and picked up a heap of legacy security applications.   
I made a widget with the SIM800L that hooked up to our security system and would txt me every two weeks to confirm it was still on line but would phone me if there was an activation.   It was just built on Veroboard.
Definitely check that your local service providers are still offering 2G service and that you're not just briefly connecting to the network and then getting flicked.

David

[Giuseppe]

Yes 2G service is still provided in my area. In fact, sometimes the module holds the line for many minutes

[Giuseppe]

It seems that feeding directly onto the module pins establishes a stable connection. I'll try carefully and then I'll let you know

[JonW]

These modules are terrible and misleading being so small and supplied with that helical antenna.
As mentioned they are intended to be mounted on a much larger ground plane for thermal and RF properties.

Even mounted on the DIL Carrier and then onto another larger pcb will yield poor RF and thermal performance.  Likelihood is that the module will ramp to max RF power that may cause thermal shutdown in the SOC due the RF PA efficiency.  Probably close to 8W peak dc power during transmit bursts. Depending on the base station location, duty cycle and Antenna Efficiency the module may get toasty quickly.

Also RF at peak power can cause all kinds of problems to nearby electronics if you have flying leads, if these are even close to the 1/4 wavelength of multiples of you can induce large transients onto other electronics during transmit bursts.

Another issue will be the VSWR with that setup, I very much doubt it will be any good and this will cause major issues with range and can stress the PA because RF power will be reflected back into the PA and depending on the phase it can easily pop the PA.  I read many posts complaining  that the reliability of these modules is poor but it's likely to be the users causing the damage and not the chipset failure.

Think you will really struggle getting a robust solution on breadboard with such a small module, no heatsinking, poor antenna/RF setup. You're probably better designing a pcb with good heatsinking and RF microstrip traces leading to a decent antenna with a tested VSWR or try to source a larger EVK/EVB with a module mounted to a SMA connector or UFL and buy a decent Antenna (not a stubby with a crappy gnd plane).  In the long run it's likely to save cost buying a known working EVK versus time spent debugging.