We have two older product designs that use Core i5 processors. The older of these uses the i5-7260U and the later one an i5-1145G7E.
We are now seeing support escalations for prematurely dead CMOS batteries in the products with the i5-1145G7E.
Both designs implement the basic CMOS battery circuit shown below (source) with a few minor changes:
In the course of investigating the cause of the early battery depletion seen on the 11th-gen designs I noted a very large difference in the current into the VCCRTC pin(s) between the two processors when the system is turned on:
- the '7260 showed about 2.6uA currentat a Vbat of ~3.2V
- the '1145G7E showed approximately 280uA for the same test.
In normal operation this current should be sourced from the top-leg via "3.3V_Sus" but I have reason to believe that not all of the current is coming from there, with some coming from the battery and, in the process, killing our battery capacity.
When the systems are off the RTC current appears very similar at about 4uA.
My questions then are:
- Is the shown reference circuit for the RTC battery (sourced from AP-728, dated April 2012) applicable to the 11th gen CPU or is there a new design?
- Is the two-orders-of-magnitude increase in VCCRTC current for the 11th gen over the 7th gen normal for this part?
- If this is normal but out of spec are there any errata or workarounds for the problem?
Link Copied
