K64 reset_b pin

[PM2]

Hi,
What capacitor can I use on the K64 reset_b pin? I have read that I shall use 10k pullup and 100 nF to ground. A MCU with empty flash will generate a 70KHz waveform on the reset_b pin. A 100 nF capacitor will destroy that waveform. Is that OK?

Thanks,
Per

[myke_predko]

Hey @PM2 

I've never heard that an MCU with "empty" Flash will generate a 70kHz wave on the Reset pin - do you have a reference for that assertion?  

Going back to your original question, the size of the resistor/capacitor on the reset line will determine the rise time from when power is applied and !Reset becomes active.  

In regards to Kinetis, I believe that there is a 35k pull up built into the !Reset pin and the Voh threshold is Vdd x 0.7(5) - if you're going with the RC route you can use an online RC calculator to figure out your rise time (and when the Kinetix starts).  

Personally, I would be a bit leary of this approach - the Reset pin is used for programming and hanging a fairly large capacitor on the line may cause problems when your updating your Flash contents. 

I use a voltage supervisor (currently the APX803S-31SA-7) which provides a 240ms specified delay between when the Vdd line becomes active and it allows !Reset to go high. 

A voltage supervisor costs a few pennies more than an RC circuit, but it's operation is much more deterministic and will reset the Kinetis in the case of power glitches (which may not be caught by an RC network).  

myke

[PM2]

Hi,

This is a behaviour that has been discussed a lot in this forum. According to NXP this is normal.
See NXP employee answer to question below.

https://community.nxp.com/t5/Kinetis-Microcontrollers/Reset-signal-of-K64/td-p/473694

/Per

[myke_predko]

@PM2 

Interesting.  I just looked at one of my a boards, erased the Flash and saw the waveform described in the referenced thread - thank you for the additional datapoint.  

If you look back at other discussions, I was in one regarding the Trinamic Kinetis control boards where they only have a capacitor on the !Reset pin and we queried Trinamic regarding how it worked with JTAG programming (they used SWD) and never got a satisfactory answer.  A big part of the problem was that we never got a schematic from Trinamic that matched the boards (in some very obvious ways).  

Going back to the original question and with these data I wouldn't change my answer and I would be more emphatic in recommending that you don't put any kind of capacitor on the !Reset pin, look at using a voltage supervisor with a pullup (with this configuration you can have a reset button pulling to ground AND do JTAG/SWD programming without concern).

When you're trying to get any kind of processor running, you need three basic things:

• Power 
• Clocking 
• Reset

If you ensure that each of these are good, any problems tend to be in software/configuration registers which makes bringing up a board much easier.  Knowing exactly how !Reset is operating is a big step in that direction especially if you are going to do programming on the board.  Again, I highly recommend not putting capacitors on the !Reset lilne and using a voltage supervisor.  

Good luck,

myke