Electronic – Need help finding pin layout diagram for GSM module (SIM800c RPi development board)

First things first. Grab yourself a copy of the BISS0001 datasheet.

To understand what was going on I traced the signal line from the microwave IC to the BISS

1. signal comes from top left corner (green)
2. then enters preamp stage on pin14 (green)
3. goes from pin 16 to pin 13 .. 2nd amplification stage (blue)
4. goes from 2OUT to internal comparator

The comparator does following ..

• signal should be around V2 ~ (1/2 VCC)
• if the signal goes above VH or VL (i guess 0.3*VCC and 0.7*VCC) the output is triggert

But .. what the hell is wrong? Why does the output stay always high? The problem (in my oppinion is the output amplitude of the 2nd amplification stage which higher then expected, maybe due to some noice or whatever).

To understand what is going on the 2OUT I got my arduino and connected 2OUT to an analog input via 1megohm resistor.

The following picture already shows my modified circuit.

• startup and stabilization is .. fine .. this would explain the first long phase of the led staying high
• 2 = min value
• 3 = max value
• 4 = avg value (all during 50ms period)
• 1 = output value (aka LED)
• 5 = std-deviation (just for testing)
• scale 0-1023 is analogRead (aka 0-5V)

During the retrigger-inhib period the signal gets very eradic, even after replacing the amplification resistor (2nd Amplification stage) from 105 to a 150k (154). Before doing this the max signal would spike VH (0.7*VCC) and the output would instantly toggle again at the end of the inhibition period.. rendering the module useless.

EDIT: Try a variable resistor with maybe 200k Ohm to find the best value for your application

After replacing the amplification resistor I got the module working!

SOLUTION:

• Replace amplification resistor from 105 to eg. 154 (lowering the amplification) or put one in parallel
• (optional) Replace output pulse time resistor from 103 to 471 (10kohm 470ohm) (30sec -> 6 seconds)
• (optional) Lift pin 1 and connect to ground to prevent retriggering

Some additional ideas that worked very very well for me:

• Power the FC1816 VCC and UDP through an RC-Section (100Ohm Resistor and after that (on the side of the FC1816) an capacitory ~100-220µF)

My final solution I am currently using:

1. Replace the 2nd stage-amplification resistor with a 150k resistor
2. Disable the trigger (PIN3 low) all together and grab the signal at 2OUT and feed it directly to an arduino for further processing
3. Use a 150 ohms series resistor and after that an 1000µF cap. The 150 ohms are crucial .. i tried 80 ohms and this was way too low.

To do so I used the statistics library and had very good results with max/min/std-dev

Summary

Everything you are reporting is consistent with a voltage drop at the module, when they require extra current to connect to the network (up to 2A pulses during RF transmit).

Explanation

I tried to connect to network with two SIM800C and two SIM800L modules (they have the same breakout pins). SIM800C tends to turn off immediately after 7 blinks while SIM800L restarts in this case.

So the problem is not just with one module, but affects several - which suggests problems with your setup rather than the modules.

That behaviour of the modules is consistent with voltage drops causing erratic functioning or reset. Therefore I would not continue with your plans for doing a firmware upgrade, as I believe that would be wasting your time, and might damage what are probably functioning modules (power interruptions when writing to flash, could leave the modules without a working, bootable, firmware).

Read these questions describing similar power-related problems:

How to properly power sim800l module?

GSM Module gets into reboot loop

When either one is set to a wrong frequency on which my operator doesn't transmit or has its phone functionality disabled (AT+CFUN=4) they don't turn off/restart themselves.

Again, that fits completely with a power-related problem. AT+CFUN=4 disables the RF transmitter (and receiver) which causes the brief spikes of higher power-drain, which causes the problem. The same is true when the module can't "hear" a cell tower to talk to.

Thanks for the photos you added. They have been useful in confirming some problems:

• The battery is simply too far away from the module. The photos don't show the full length of the battery-to-breadboard wires. However the resistance & inductance of those long wires + breadboard "tracks" will seriously affect current flow to the module during high-current drain e.g. RF transmit.

• The wires going to the battery holder also appear quite thin - thinner than the normal "Dupont" jumper wires. This is a concern, as thin wires will cause greater voltage drop under load, than thicker wires.

• The 1500 uF capacitor is also too far away from the module, so the resistance & inductance between it and the module reduces its ability to supply the additional energy required under the extra load during RF transmit.

  The capacitor must be close to the module - "close" meaning within a few mm, not 15+ cm away (look at the length of the wires carrying power between the module and the power tracks on the breadboard, and then the length of the breadboard tracks to the capacitor and battery). And those cable lengths are effectively doubled, as the return current has to travel the same distance.

• I've never tried using 18650 batteries for powering such modules, but beware that not all 18650 batteries meet their specification; there are many fakes which will measure 4.2V after charging (so they might seem OK) but have small capacity and higher internal resistance.

• When you tried using a bench power supply, this likely added extra cable length to the power network, so if your problem is related to extra resistance & inductance in the power cabling, then even a stable power source may not have been good enough.

Using an oscilloscope to measure the power supply voltage on the module itself during RF transmit should confirm the problem. The fact that you see a voltage dip during RF transmit even on a multimeter, is a good indication to do further investigation.

Changes which may be part of a resolution could include:

• Move the battery holder closer to the GSM module and shorten all wires carrying power to the module to an absolute minimum (i.e. not using the standard length "Dupont male-to-male" jumper cables you are currently using);

• Use thicker wires connecting to the battery holder;

• Mount the additional capacitor (or preferably an SMD alternative) onto the module itself with minimal length leads - e.g. as a piggy-back to the existing tantalum capacitor visible on the lower-left corner of the module (I expect that tantalum capacitor is across VBAT, but you should double-check).