I am powering a SIM800 module which looks like this: enter image description here

This module is rated between 3.6v and 4.4v and can draw upto 2A. I have this hooked up to a single 18650 LiON battery. According to the datasheet these are designed to run from this kind of battery. The problem I am having is that when it powers up, the moment it joins the GPRS network, the module will reset. This I presume is because of the extra power requirement as it joins the network. Reading the datasheet it suggests a decoupling network. At the moment I have it like this with the SIM800 next to the battery and the decoupling network off to the side. enter image description here

This doesnt seem to make any difference and the SIM800 just resets as soon as it joins the GPRS network. I am wondering if it should be like this with the network at the end of the decoupling capacitors or doesnt it make any difference? I have soldered this part so it's not easy to try this configuration so wanted to ask the question first. If it doesnt make any difference what am I doing wrong?? enter image description here

If I have it hooked up to a bench power supply I get the same problem but if I push the voltage up to 4.3v it seems stable so I am sure my problem is with the decoupling somewhere.

  • 2
    \$\begingroup\$ Try catching this dip on oscilloscope \$\endgroup\$
    – user76844
    Jun 20, 2017 at 20:31
  • \$\begingroup\$ I wish I could. I dont have a scope at the moment. It's on my list! ;-) Do the two schematics make any difference in their wiring? \$\endgroup\$ Jun 20, 2017 at 20:36
  • 1
    \$\begingroup\$ Then we can only guess. How close are the caps to the device? Can you take a picture? \$\endgroup\$
    – user76844
    Jun 20, 2017 at 20:38
  • 4
    \$\begingroup\$ That decoupling network is already included on the board, it's the row of caps on top. This is done because such a network has to be connected to the circuit with as shortest traces as possible, or it won't work. So you have to look for a different problem. \$\endgroup\$
    – Janka
    Jun 20, 2017 at 21:40
  • 2
    \$\begingroup\$ You may well have trouble solving this with capacitors. Typically it is done with a low impedance connection to a cell capable of high currents. 18650's have good energy density, but they aren't really a high current design. Also beware of some holders for them - coil spring contact ones can have surprisingly high losses when you try to pull an amp or two. \$\endgroup\$ Jun 21, 2017 at 3:19

1 Answer 1


This is a fairly typical problem for modern radios. If you check the datasheet, all their recommended power supplies contain a 330uF cap. The same datasheet says the module makes impulse loading of 2 A at 5 ms period.

Now consider this: 2 A at 4V is a 2 Ohms load. If you have just a 100uF cap, the RC is 2e-4 s, or 200 us. The load duration is 577 us. This means that your 100 uF capacitor will discharge nearly completely during the RF transmission pulse. Of course the feed from battery will somewhat help, but still... You need to increase the local bypass capacitance at least 10-fold, with low ESR caps, maybe several 100-uF ceramic caps, then you might make it work.

  • \$\begingroup\$ Thank you for explaining some of the maths behind it as well. That helps me a lot on what I am looking out for. I will try with more caps and report back. \$\endgroup\$ Jun 21, 2017 at 9:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.