Fastest high resolution timer in a microcontroller?

Thread Starter

irmanao

Joined Apr 29, 2017
87
I am doing some research for a quick microcontroller to measure some pretty sort time differences and was wondering if you guys know what microcontroller has the fastest timer. The MSP340 from TI has a 5ps resolution but its development board is really expensive and I am not familiar with TI's microcontrollers. The next fastest I found is the STM32F334 with 217ps resolution and the KV5x (NXP) with 260ps resolution. Are there any other offers that I missed?
thanks
 

MrChips

Joined Oct 2, 2009
19,271
Timers generally derive their clocks from the core MCU clock. Hence you cannot generalize on the speed of any family without looking at the specs of specific components.

Take a look at the STM32H743 that can be clocked at 400MHz (2.5ns resolution).

Edit: I see that STM32F334 has a high-resolution timer that is clocked at 144 x 32MHz to give 217ps resolution.

I have not been able to read read how the MSP430FR6047 can achieved <5ps. I believe this is embedded in the ToF measurement in the USS software.
 

Thread Starter

irmanao

Joined Apr 29, 2017
87
Upon further investigation, the STM32F334 seems to have a resolution of 217ps only for output and not for capture mode... Is that correct? I can't measure time between two events with this resolution?
upload_2018-7-26_21-21-56.png

So the only solution is a TI chip???

thanks
 

MrChips

Joined Oct 2, 2009
19,271
I'm on the trail. When I get back to base camp I have to look into how the MSP430FR6047 can achieve <5ps. That is incredible!
 
I would also like to know how a TI MSP430 has a timer at <5ps and, frankly, I am skeptical. I was looking at this a bit this morning and the D timer on the TIs that you seem to be referring to run at 256 MHz and that is a period of just shy of 4 ns and the CPU is at, what, 25MHz.

So, yes, I would like to know how that is possible and what I am missing.

Here are two interesting threads that might interest you:

https://electronics.stackexchange.com/questions/59216/very-accurate-time-measurement

(note the reference to a 1GHz overclocked RPi?!!)

https://www.researchgate.net/post/is_there_a_processor_which_provides_Nano_second_timer

(discussing the use of counters outside of the microcomputer)
 

MrChips

Joined Oct 2, 2009
19,271
On further research, the TI MSP430FR6047 does NOT have <5ps timer resolution.
That capability quoted by TI is the ToF (time-of-flight) resolution of the USS (ultra-sonic sensing) technique for measuring ultra-low flow rate for metering applications.

http://www.ti.com/lit/ml/swmp003/swmp003.pdf

http://www.ti.com/lit/wp/sway007/sway007.pdf

The ADC samples the data at 3.6Msps. The data is processed to extract the differential ToF (forward and reverse ToF between two sensors). The <5ps quotation is the resolution after processing.
 

Thread Starter

irmanao

Joined Apr 29, 2017
87
Nice catch. I don't know if i'll go that route then unless i find some examples (i'll have to look into that). Another option is combining any microcontroller with the TDC7200 (time-to-digital converter) chip which has a 55ps resolution and is accessible via SPI. So the fastest timer is on the new stm32h743 with a 2.5ns with 4.5MSPS. The TI MSP430FR6047 has 8MSPS. Also, for the KV5x from NXP, the 260ps resolution is also for PWM.
thanks guys
 
Last edited:
Top