Is STM32 Nucleo F446 handle HSE bypass from STLINK properly ?

[SNish]

Dear all, I am using Nucleo-F446RE and also Nucleo-F103RB.
As you may know precise clock testing require for some application.
For testing purpose always change RCC as Bypass HSE clock from STLINK on this case.

Recently I have used Nucleo-F446RE as above but behavior is different as failure on startup.
I have tried simple test on both Nucleo-F103RB and Nucleo-F446RE.

F103RB's result is correct. Bypass clock is worked.
I had reflashed STLINK side as latest build on both unit.
So STLINK side has no delta.

 Using environment CubeIDE 1.19.0 with included CubeMX.

Now I am looking F446RE SystemClock_Config in detail.

[SNish]

NucleoF446 has SB54 solderedNucleoF103 has No SB54

[TDK]

F103 and F446 are totally different chips on totally, or at least substantially, different boards. Focus on the one you want.

> failure on startup.

What type of failure?

HSE bypass signal comes from the st-link. You'll need to match the solder bridge configurations as outlined in the user manual. The board doesn't have an HSE crystal.

[SNish]

Dear TDK
Thanks for your comment.
I understand your point. And I had read document of UM1724 also.
SB55 should be off for using HSE Bypass method.

In addition, both chipset has same Nucleo64 board so document is same for setting.
I am using NUCLEO-F446RE now.

I am looking schematics of MB1136 SB55 and SB54 are junction to Morpho connector(CN7) PIN29, PIN31.
My trial is using HSE Bypass setting for getting 8MHz MCO from STLINK into OSC-IN of STM32-F446RE side.
But RCC-CR is not show HSE-RDY bit on this case.

I am wondering about whether SB55 solder bridge ON to OFF now.
Because of this effect is increase some capacitance on OSC-IN only.

But original NUCLEO-BOARD soldering bridge has delta on SB55.

regards

[SNish]

 STLINK MCO output SB16 is closed. 
STLINK firmware has been updated with same latest image.
And NucleoF103RB is working fine, So STLINK should be right on this point.
But it should confirm with oscilloscope in fact.

According to UN1724 SB55 should be off.
Current NucleoF446RE is ON. 
But I am not sure real meaning of this effect.

 

Failure happend at SystemClock_Config line 355 break as Error handler  (HSE_RDY is not raised within time)

 

 

Dear TDK
In addition, I am planning to visit my labs tomorrow.
And try to monitor with oscilloscope on this case.
Now I can use CubeIDE debugging method from SW side,
Viewing RCC-CR bits shows not raised HSE-RDY bit.
I will report tomorrow after trial.
Thanks for your comment.

I had attached CubeMX report.

[TDK · Accepted solution]

I recommend following the instructions in the user manual. At least, try those first before questioning things.

[SNish]

Dear TDK
I have settled setting as UM1724

SB55 off SB54 on

SB16 and SB50 on (SB50 setting had been missed)

R35 and R37 removed

Now it worked thanks for your support.