High current consumption during RF sending

RE: High current consumption during RF sending

Asbjørn — Mon, 23 Feb 2015 12:51:34 GMT

It seems like the capacitor had some effect on the peak, but I do suspect that the distance from the chip and also the overall size of the module would require the capacitor to have a different valuer or potentially a different combination of capacitors to reduce the peak. We haven't tested or qualified this module ourselves, so I'm not able to give you hard recommendations from the top of my head. I would recommend you, if you are able to, try a few different capacitor values, probably with higher value. This in combination with asking Insight SIP if they have tried this themselves on their module. They might have better suggestions for that specific module.

RE: High current consumption during RF sending

Vicnet — Tue, 17 Feb 2015 15:08:59 GMT

It was placed outside the module. Maybe to far from the module.

RE: High current consumption during RF sending

Asbjørn — Tue, 17 Feb 2015 14:31:19 GMT

The additional 4.7 uF was placed outside the module or were you able to actually replace the C9 capacitor on the module itself? The concern is that the capacitor is places too far from the chip and DCDC itself to be able to handle it efficiently on such a small module. From the qualifications we have done on the chip internally we have used larger PCBs where we also place the additional 4.7 uF capacitor right next to the chip.

RE: High current consumption during RF sending

Vicnet — Tue, 17 Feb 2015 07:48:16 GMT

Insight SIP confirm that a 4.7µF external capacitor has to be add on VCC to reduce high current peak.

With 4.7µF the peak current is 57mA -> 42mA.

Does it exist another solution to reduce this peak without disable DC/DC?

RE: High current consumption during RF sending

Vicnet — Fri, 13 Feb 2015 15:35:48 GMT

Yes, my mistake.. I confused C7 (nRF51822 datasheet) and C9 (Insight module). I'll ask to Insight C9 value.

RE: High current consumption during RF sending

Asbjørn — Fri, 13 Feb 2015 14:54:46 GMT

@Vicnet, I wasn't able to find a schematic for the ISP130301 with component values. I do however see that C7 is a different component compared to the C7 in Nordics nRF51822 product specification. In the Insight documentation the component I'm referring to is named C9, but without any value specified. I found the datasheet here. (Document named: "DS130301_R4.pdf). So yes, it it could be that the module has one large capacitor on AVDD, but we recommend to also have a 4.7 uF on the VDD pin itself to reduce the peak you see.

RE: High current consumption during RF sending

Locky — Thu, 12 Feb 2015 23:35:22 GMT

I tested this theory today, I started the HFCLK (32MHz) crystal then drop into an infinite loop with _WFI(); The board consumes 1mA, instead of ~500uA. This could be a contributing factor to the increase in peak current consumption during radio events.

RE: High current consumption during RF sending

Locky — Thu, 12 Feb 2015 19:31:24 GMT

I see the same ~1-1.5mA increase in RX and TX peaks. It is not an MCU running issue as I do not measure the same peak heights on the Dev kit, only on custom hardware. I've hypothesized that it may be somehow related to either the 32MHz crystal I use, or perhaps this particular version of silicon.

RE: High current consumption during RF sending

Vicnet — Thu, 12 Feb 2015 16:50:35 GMT

On Insight ISP130301, C7 - 4.7uF is already connected to pin 1. How can you explained this peak if C7 is connected? Other explanation?

RE: High current consumption during RF sending

Asbjørn — Thu, 12 Feb 2015 16:11:36 GMT

Yes, you should have a 4.7 uF capacitor at C7. See chapter 24 in chapter 11.3.3.

Do you have anything else running on the chip while you are transmitting? The current consumption if everything is sleeping except the radio, the current number you should see should be in the same as the RX/TX current specified in the product specification with/without the DC/DC.

RE: High current consumption during RF sending

Vicnet — Thu, 12 Feb 2015 14:29:33 GMT

Hi Asbjørn, Thanks for your reply.

During RF sending, the MCU is sleeping. We are using the Insight SIP 130301 and a larger capacitor (C7) is already on VDD pin1 (ver nRF51822 QFAA).

When you said: "larger capacitor on the VDD pin1", add another capacitor in addition to C7?

RE: High current consumption during RF sending

Asbjørn — Thu, 12 Feb 2015 08:42:11 GMT

The additional current during TX is most likely due to the MCU running while sending and the difference between 12.2 mA and 15.4 mA is due to the DC/DC. The peaks at the end are due to the shut down of the internal PA and it's power supply. While running on the DC/DC we have seen that there might be a short but high peak as you measure. To counter this we have recommended to add a larger capacitor on the VDD pin1 as seen in chapter 11.3.3 of the nRF51822 product specification.

RE: High current consumption during RF sending

Locky — Tue, 10 Feb 2015 21:48:36 GMT

As for your DCDC measurement, I recall when I used to work with the Ti CC2540 that when using a DCDC, there can be high swing currents at the enable/disable points. This is due to the fact that the Voltage rail switches from Coil Cell level (3.0V) to DCDC level (~2.1V on Nordic) and back during RF events. With this voltage swing you will see an in-rush current when measuring the board with a resistor. Though 57mA seems quite high.

RE: High current consumption during RF sending

Locky — Tue, 10 Feb 2015 20:10:17 GMT

I also see 16.5~17.8mA on TX and 14.5~15.4mA on RX. I do not use the DCDC on my hardware so its inline with your measurements. Not entirely sure of why the deviation from the PS