Can make a 50MHz PWM output?

[jhwan.2]

MCU : STM32H7, PB0 ​PWM CH2N

TIM1 CLK : 200MHz

Prescaler : 0

Counter Period : 3

Pulse : 2

GPIO Speed : Very High​

​

​

​

I'm testing a 50MHz PWM output.

And my test conditions are written at the top of this question.

​

As a result, I decided that a 50MHz PWM output was impossible.

This is because the square wave is distorted by the slew rate as shown in the attached picture.

​

However, the output buffer timing characteristics of the datasheet states that frequencies up to 220 MHz are possible.

​

So I leave a question to ask if I've missed anything.

​

thank you

[Javier1]

>>GPIO Speed : Very High​

are you just toggling a GPIO or using a proper PWM timer?

What about unwanted capacitances? show us your hardware

[jhwan.2]

hi. Javier

I use TIM1 for PWM CH2N.

Nothing is connected to the PWM output.

I'm just checking the waveform with an oscilloscope.

[Javier1]

>>Nothing is connected to the PWM output.

you think nothing is connected to the pwm output.

Are you usign your own custom board or any nucleo/discovery boards?

Because nucleo and discovery boards have lots of caps and external components

[QSHAO.1]

the capacitance load on PB0 may be a reason of longer rising time of PMW signal, so you may check this point firstly. ​

[jhwan.2]

hello.

I haven't connected anything to the PWM output (PB0).

I'm just checking the PWM waveform with an oscilloscope.

Thank you for answer

[LCE]

... and as your oscilloscope looks exactly like my Tektronix with only 100 MHz Bandwidth:

to check a 50 MHz square wave, you will need much higher bandwidth.

Not to forget the probe, and how it is connected to the test point - mind the GND connection, which should not run over a few inches of extra cable.

[jhwan.2]

..

[LCE]

Nice, okay, I oversaw the 500 MHz!

Still, you can do a lot of things wrong with the probe.

[S.Ma]

Io speed = 3 ?

Probe in DC or AC mode?

Test MCO pin at 50+ MHz.

[AScha.3]

and probe is 10:1 ? how many Pf input ? min. 200 MHz type?

+ square wave is adjusted ?

64Mhz out on MCO1, with my DSO, and with 100MHz probe (so no square possible, just similar sine wave):

 and 100Mhz , 4Vss ...

 

[jhwan.2]

I'm used in x1 mode. and I don't know "Pf input" exactly.

I wonder what is your MCU Model

please let me know. thank you

[LCE]

In GHz-wireless times, measuring some 50 MHz seems like it must be so simple... :D

"pF" input: capacitance of the probe, a good probe has that printed on the probe or its connector.

And do NOT use x1 mode, all the probes I ever used have a much higher capacitance than in the 10:1 divider mode.

[AScha.3]

you have to use probe in 10:1 setting. (1:1 is useless with hi-speed cmos circuits ! )

cpu is stm32H743vit , at 200MHz core clk.

[jhwan.2]

The problem has been resolved.

Probe setup was the problem.

After setting to x10 mode, the waveform was checked normally.

thank you