Board Restarts continuously with External Antenna

[Dumken]

Hi Good day guys, I need your help. I have a board running a custom linux distro. It has a SIM800C as part of the circuitry. When ever I insert the external antenna and try booting the board, the board will continuously be restarting. If i remove the antenna, the board boots well. Please what could be the problem?

 

[dknguyen]

What are you using to supply everything? Maybe it's browning out because with the antenna connected the radio is trying to do it's thing which is drawing more power than can be supplied.

 

[Dumken]

What are you using to supply everything? Maybe it's browning out because with the antenna connected the radio is trying to do it's thing which is drawing more power than can be supplied.

I am using 4.2V 3A. I have reviewed that option it seems not to be the problem.

 

[Nigel Goodwin]

I would concur, most like a power supply, poor decoupling, or PCB layout problem - GSM modems take high current spikes (2A or so), these pulses can easily cause a brief drop in the power rail and reset the processor.

We use a 1F 'BEST' capacitor across the supply rails, and short connections, to avoid such problems - and also (In our case) because we're using long life Lithium batteries which can't supply the high current pulses.

 

[Cicero]

Similarly my instant gut reaction is also a power supply issue. Place sufficient caps as close as possible to the SIM800C , even mod a board to supply a different power supply to the SIM800C to further pinpoint a power problem, scope out the current spikes etc.

Its the transient nature of the power requirements, not the average which causes problems. Although 3A may be enough for normal operations, large current spike requirements can cause voltage sag for transient current requirements and thus cause brown outs. That sag is exacerbated by insufficient track widths and layouts, insufficient decoupling. That 3A capability might be reached under perfect, slow resistive conditions, not taking into consideration heat derating and so on.

Not saying you haven't thought of all these, just throwing out thoughts.

 

[Nigel Goodwin]

OK, as an example, here's the board I'm currently playing with, the capacitor to the left of the SIM800L is a 1000uF, with a 0.1uF across it, close to the supply pins for the module. The tall capacitor at the right is a 1F 'BEST' capacitor, to provide the current pulses that the battery itself can't, and is directly across the incoming battery supply. The power to the SIM800 is switched using a dual-FET, which is a SM device under the module - the 1000uF is on the output side of that switch. The other 8 pin SM chip is an I2C FRAM, used for storing various settings, which can be updated via a webpage and MYSQL database - when the board connects to the Internet it downloads the time (to reset it's clock) and checks for a settings update, and downloads it if there is one - once that's done, it marks the update as done in the database so as not to do it again. So far it's all going pretty well - I've currently got one transmitting every 8 hours, running from a single 18650, to see how long it lasts.

But as for the OP, it's almost certainly a power problem, which is always the case with GSM projects.

I took this picture with my phone, I'm pretty impressed - you can read the number on the PIC really well

 

[Cicero]

Not gonna lie Nigel, the way that board is routed and fits so snugly in that housing is satisfying.

 

[Nigel Goodwin]

Not gonna lie Nigel, the way that board is routed and fits so snugly in that housing is satisfying.

I'd like to take credit for it - but I copied the layout from an existing board, manually as I didn't have the files for it, but I did have a bare PCB of the right size and shape so simply measured it and made my own version of it. It's a bit tricky getting the corners dead right, but I managed it after a couple of tries.

I then saved it as a template, so I can use the same outline whenever I want.

I must admit when the boards came back from China I was VERY pleased when it dropped in perfectly

 

[Dumken]

Thank you guys. I will review the power design but currently i changed the antenna and the board seems to be working fine.

 

[Nigel Goodwin]

Thank you guys. I will review the power design but currently i changed the antenna and the board seems to be working fine.

Was the antenna shorting out?, that could well produce even higher current demands.