Sorry for the cross-posting. It was recommended the question be posed here instead of the Low Energy Forum.
We are having trouble explaining a roughly 6 mA, what appears to be DC, current draw from the CC2564-PAN1326 in our custom designs and we are hoping the community may have some ideas to help us debug. We have made several breakout boards that allow us to breadboard the part for testing purposes as well as a custom PCB design. We see the same problem (plus or minus a few mA) for all the breakout boards as well as the custom design. The current flows no matter what state the Bluetooth chip is in. We even see the unexplained current when there is nothing but the CC2564-PAN1326 powered and held in reset. The interesting part is that the chip works perfectly and can communicate with others without problem.
Above is an oscilloscope plot showing the current draw of a GATT Read request. Channel 3 (blue line) shows the output of an INA210, which is amplifying the voltage drop across a .47 Ohm shunt resistor between the power supply and the CC2564-PAN1326. Channel 1 (yellow line) is measuring the reference voltage of the INA210. As suggested by the datasheet to help keep the INA210 in a linear region for small currents the reference is set to Vcc and the shut resistor connected in reverse, therefore the one must take Vcc - Vout to get the real amplified signal. The pink line is a MATH trace which takes Vcc-Vout and scales the voltage to directly read off the current in the shunt resistor. The cursor measurements on the right hand side of the screen are given in mV; however, because of the MATH they should be interpreted as mA.
As you can see with cursor Y2, the sleep mode current is roughly 6.125 mA when we were expecting a current in the microamps. Also notice the current drawn during the Read request is offset by the 6.125 mA. As said before, this 6.125 mA is seen flowing even when the chip is held in reset. To make sure the problem was not caused by the current measurement setup we verified the average current draw from the built-in ammeter of the power supply with the INA210 and shunt resistor removed.
It seems unlikely, but does anyone know if this is a known problem? If not, then our working theory is that we are somehow consistently damaging the chip. To do the soldering we use a hot plate with two different temperature zones. We try to follow the heating curves in the CC2564-PAN1326 datasheet as best as possible; however, there is only so much control one has with a hot plate. Is it possible that overheating or incorrect heating could cause this type of problem? If we have solder joints under the chip does it seem reasonable that we would only have 6 mA of current?
Thank you for any help or thoughts you can offer.
Andrew