9160 Continuous Current Drain during Cell Tower Connection -- Is this expected?

[alcornb]

I'm having some issues with the 9160 on battery and wondering if we're doing something wrong or if this is expected behavior.  During the initial connection to the tower, I'm seeing an average continuous current pull of ~40mA on battery supply (which we directly feed to the 9160).  This goes on for seconds or even minutes.  Is this normal?  If so, I can't imagine any small batteries hold above the 3.3V minimum no matter how big the super cap...  With two brand new lithium AA lithium batteries in parallel (3.6V nominal) and a 100mF supercap, we can only stay above 3V for a couple seconds.

Here are a few scope captures for clarity:

1. Zoomed out (1 second per div).  Blue is battery voltage; starts at 3.6; cursor shows final level of ~2.95V.  Yellow is Vbatt current consumption (20mA per div... converted from voltage).

2.  Zoomed in on radio transactions (5ms/div).  Blue is battery voltage.  Yellow is current consumption.  Scales are same as above.  Current peaks ~80mA, and never turns off fully.

Note that the current consumption measurements in yellow are made with a custom current sense resistor and amplifier setup.

My power supply scheme and layout were reviewed by Nordic in a private post and the feedback was that the 100mF super cap may not even be necessary.  That makes me think something is going wrong with the modem or there is some other unexpected behavior on my hardware.

Is this what I should expect for current consumption during an initial connection to the tower?

[Hakon]

Hello,

 

I'm seeing an average continuous current pull of ~40mA on battery supply (which we directly feed to the 9160).  This goes on for seconds or even minutes.  Is this normal? 

 Yes, as far as I know it should be around 35 mA when scanning. Have you considered using the %xbandlock command? It could speed up the connection time if you know what bands your network uses.

[alcornb]

Thanks Hakon.  That's all I needed to know.  We'll updated the supply scheme to handle this.  For future reference, that might be a nice typical spec to add to the datasheet.  Coming from other low power radios, I had an incorrect expectation that there would be a longer duty cycle associated with the current pulses.  Maybe it's already there and I missed it.