Skip to main content
LMI2
Senior III
August 23, 2018
Solved

Wrong value of VCAP capacitors in STM32F407VGT board

  • August 23, 2018
  • 10 replies
  • 2141 views

I have 10uF (=10uF+100nF)ceramic capacitors for VCAPs in a custom board. Voltage there is 2.7V. Supply voltage is 3.3V

I have problems getting the STlink to accept the board. First I managed to erase the chip, now STLink does not even recognise the chip. A dead CPU already?

Are 2.2uFs really needed, I try keep number of stored values here as small as possible.

This topic has been closed for replies.
Best answer by waclawek.jan

While the DS is very specific with the 2.2uF, and while some LDO are known to oscillate with higher than specified capacitive load, 10uF doesn't sound to be too far.

Grounds correctly connected to all pins (checked and double-checked by measurement)? Decoupling close to the pins? No excessive voltage or oscillation on 3.3V pins? No >7V on any other pin? No oscillation on VCAPs?

JW

10 replies

waclawek.jan
waclawek.janBest answer
Super User
August 23, 2018

While the DS is very specific with the 2.2uF, and while some LDO are known to oscillate with higher than specified capacitive load, 10uF doesn't sound to be too far.

Grounds correctly connected to all pins (checked and double-checked by measurement)? Decoupling close to the pins? No excessive voltage or oscillation on 3.3V pins? No >7V on any other pin? No oscillation on VCAPs?

JW

T J
Senior III
August 23, 2018

I checked my F417 custom board,

one 2.2uFs only on each of those pins, Both reading ~1.31V

Tesla DeLorean
Guru
August 23, 2018

Should have 1.25V, expect other issues on board

Tips, Buy me a coffee, or three.. PayPal Venmo (See Profile) Up vote any posts that you find helpful, it shows what's working..
LMI2
LMI2Author
Senior III
August 24, 2018

Thank you for info. I have tried to check those connections but it is hard when chip is this dense.

I wonder about that 2.7V VCAP voltage. "No oscillation on VCAPs?" I haven't checked with a scope but the voltage looks stable.

Power comes from 7805 type 3.3V switching regulator. Every CPU + pin has 100nF and most have also a 10uF a little further eway. And there are other capacitors between + and ground. This is a test board so I have room.

I have to check further, swapping CPUs is hard I hope it is not that.

LMI2
LMI2Author
Senior III
August 24, 2018

I resoldered all the CPU leads and it helped. Keil does not detect the board but ST link does. I can download a program but it looks like it is not running. More work.

Edit: And the Vcap voltage is now 1.26V

LMI2
LMI2Author
Senior III
August 24, 2018

And it works.

I had to toggle the LED pin and not something else to get it blinking.

And with 10uF Vcaps

waclawek.jan
Super User
August 24, 2018

Please mark it as solved.

LMI2
LMI2Author
Senior III
August 24, 2018

Like that?

waclawek.jan
Super User
August 24, 2018

Yes thanks.

This forum is horrible in keeping track in what thread did I (or others) start conversation, and this helps at least a little bit to filter out those which don't need attention anymore.

JW

T J
Senior III
August 24, 2018

I guess you are using an LQFP package...

I guess you had a bridge somewhere.

Flux makes the rework 100x better.

Do you use a microscope ?

all my board use LQFPs, for expedient repair of prototyping issues.