20:16 <abws> hi I have random oops and reboot with helios64 with a cpufreq switching test for big cpu of the rk3399. I have been hinted on #u-boot that the helios64 board likely is not well designed.

20:18 <abws> for one the cpu b regulator is a syr837 but the design around it is supsciously the same as the rk3399pro design around the tcs4525 regulator for cpu b. But the tcs4525 does not have the same requirements (it has internal pulldown on EN and VSEL, the typical application has 10uF capacitor on VIN while syr837 requires a 22uF one

20:25 <abws> the syr837 datasheet typical application has a 0.25uH inductor on the switching node and one 22uF capacitor before it goes to cpu b, and direct line from SW to VOUT sense while the helios64 has 0.47uH on SW and 4 x 22uF capacitors before cpu b and 133ohms resistor before VOUT sense

20:29 <abws> Do you know a place to get tcs4525 and tcs4545 datasheets (I see that rockpro64 has an even closer design around the tcs4545 to regulate the cpu b voltage)?

21:39 <abws> Could one with a Rock960 board test this testcase which crashes Helios64 with a similar design for the syr837 cpu_b regulator (except the inductor which is 0.24uH on Rock960 but 0.47uH on Helios64) https://forum.armbian.com/topic/24005-helios64-u-boot-does-not-build-anymore-after-we-bumped-to-202207/page/3/#comment-178800

21:40 <abws> Helios64 schematics https://wiki.kobol.io/helios64/files/schematics/Helios64_Schematics_Rev1.2.pdf