I have a 5V GSM shield running off 18650 batteries.
Initially I had this 5V 2A step up converter but it would cut off as not enough current could go through.
So I bought this 5V 3A step up converter but despite turning the GSM module off and sleeping the arduino it seems to churn through power really fast.
I also tried this 3.7V GSM module as it runs directly off 18650 batteries, but it needs 3.7-4.2 V to power it so it cuts off before the 18650 gets anywhere near being partially discharged.
Can you guys see any clear cut best option?
I'm at work at the moment but when I get home I'm going to try using 2 x 2A step up modules in parallel to essentially give me 5V 4A output and see if that works better than the 3A module. Or should I add a capacitor after a 2A module to handle the peak current?
For the 3.7V GSM module, is there a component I'm missing to maintain the 3.7V? I feel like that was designed to be used with lithium batteries but it seems a bit useless if it can't drop as low as 3V input.
I currently have 2 18650 batteries set up in parallel. I did have them in series with a voltage regulator but the end use will be solar powered. Charging 18650 batteries in series becomes considerably more difficult. I may have to revisit this though
If you use solar panel with charging voltage less than 8V then you-ll either use a step-up converter to charge two 18650 in series, or a step-up converter to power the GSM module from 3.7V battery voltage.
In both cases the step-up converter has a base power consumption, by just being powered.
I think the first choice is better, because the step-up converter draws its current from solar panel. When there is no solar voltage, the step-up converter doesn't consume power.
If the step-up converter is powered by battery, it will keep drawing precious mAs all the time, regardless the GSM module is in use or not, until battery is exhausted
I have 12 and 6 V solar panels. So if I went with series I would use the 12V solar panel. Just when I started researching solar panels and lithium batteries most things were related to using the TP4056 charging unit for 1 or multiple batteries in parallel. There is not much in the way of cheap series charging options and the few options there are seem to have their own draw backs. All that is to say that I thought 3.7V with a step up voltage converter to 5V was the better choice. If I do use 2 batteries in series I will still be running the power through a voltage regulator to drop the output of the batteries down to 5V which would also inducing power loss into the circuit, albeit maybe not as much.
A GSM transmitter takes about 1.6A peak for 270uS with a 1/8 duty cycle - I've designed them.
The PA's are generally specc'd to 3.6v - so should be ok to end of life of your Li cell.
Suggest you add a large capacitor ( 2000uF) across the supply AT THE GSM SHIELD -
the large current pulses can cause undue ripple otherwise - particularly because the battery impedance increases towards the end of it's discharge cycle.
Hi Allan, How do you select the voltage for the capacitor? It has always confused me because the voltage in the capacitor drops so quickly. I've played around with supercapcitors and the voltage of the whole system drops to the capacitor voltage, so wouldn't it have a similar effect wiring it in with the GSM module? It would drop the voltage at the GSM as the capacitor discharges.
Also an 18650 battery is fully charged at 4.2V and cuts off at 2.8V, so 3.6V is still missing out on a significant portion of the batteries cycle, not to mention they recommend not fully charging the batteries to increase the life of the battery. If I charge the battery to 4V as often suggested it doesn't actually allow all that many mAh to be pulled from the battery.
At least 20% higher than the highest voltage it will see.
So for your 5v supply to the GSM shield, 6v or higher would be fine.
And don't use a supercapacitor - their internal impedance is far too high. An ordinary electrolytic would be OK.
by the time the Li is down to 2.8v it's really dead -and you won't get much between 3.6 and 2.8 - see some trials I did on an 18650 a while back...
-and be careful about 18650 specs - I ran these tests because I bought them as 6000mAH and wasn't convinced that they performed to that. It works out about 1250.
The most you can cram into that size is about 2500mAH - and reputable suppliers such as VARTA and SANYO claim no more.
hmm good point, I think I'll go back to the 3.7V GSM module with only the internal regulator. I'll also try a capacitor on the module as well because the power consumption really jumps around when connecting which can't be great for the batteries. Although I couldn't find any 6V capacitors (on tayda electronics) so I picked up the following.
My understanding is that the capacitor gets charged to the battery voltage and then can rapidly discharge saving the batteries need to have current spikes, but as the cap discharges it drops voltage lowering the voltage of the entire system.
What prevents it from dropping the voltage below that required to run the GSM module (is that why I needed a large capacitor?)
Why doesnt the battery see a similar spike in current to refill the capacitory, because the capacitor can charge rapidly and an 18650 can discharge rapidly so I would have thought they would almost charge and dischare at the same rate making the capacitor not all that valuable.
What brand of 18650 batteries are you using.
There are a lot of fakes on the market, especially if they are coming from China.
Fakes have very low AH capacity and very low C ratings.
jremington:
Your understanding about capacitors work in circuits could be improved.
haha that's an understatement. Thanks for the point in the right direction. I feel like I've just discovered what was causing every unexplained glitch I've had since I started this project! Is there such a thing as adding too many decoupling capacitors? I feel like my circuit would benefit having a these after my solar panels (before the TP4056 charing unit), immediately after the batteries, right before the GSM unit and right before my arduino mini (although there are internal capacitors). I can't find anything that suggests you can overdo it but I'm probably oversimplifying it.
Well it wouldn't do any harm, but not much good either.
The reason for a capacitor at the GSM shield is that it can supply the very sudden pulse of current taken by the transmitter because its internal resistance is very small - a hundredth of an ohm would be typical.
It acts as a small but very good battery.
The 18650, particularly when nearly discharged, has an internal resistance much higher than this , so when you draw a sudden large current it's voltage drops significantly, which may upset the operation of the electronics.
The capacitor helps the 18650 for the short duration of the transmit pulse.
It supplies most of the large current during the pulse, then recharges from the battery during the gap between pulses.
allanhurst:
In parallel with the battery - AT THE PINS OF THE GSM SHIELD - this is important as the impedance of the wires is not trivial with these currents.
Be sure you get the polarity right : + to +, - to - ............
the 2200uF shoould be fine. 16v is plenty.
regards
Allan
Thank you so much!! I know this is an old post but I just wanted to report for anyone else having this same issue that IT WORKED for me as well! I only had a 1000uF 16V capacitor on hand so i connected that to the pins of the GSM shield as you advised and problem solved!
Thanks again!