10MEG resistor to set switching frequency of Buck?

[cupoftea]

Hi,
We are doing a 214kHz Buck with 24vin and 15vout at 3a(out).(as attached in LTspice)
Using the LTC3878.
-Though it needs a 10 Meg resistor into its RON pin in order to give the 214kHz.
That means a current of just 2.3uA into the RON pin of the LTC3878.
Surely using a 10 MEG resistor to set the switching frequency is asking for trouble? (with noise)

LTC3878 datasheet

https://www.analog.com/media/en/technical-documentation/data-sheets/3878fa.pdf

 

[D.A.(Tony)Stewart]

Avoid 1 pF coupling to Vin and avoid 0.1 pF coupling to gate drives.
Keep the loop small for R9 between Vin and Ion by placement.

Then its ok.

 

[cupoftea]

Thanks, i also doubt that there is anything in that chip that can detect the 2.3uA.

 

[KlausST]

Hi,

I don´t see a problem in detecting 2.3uA. To keep noise low, just make the node as small as possible.
--> place R9 close to the IC pin.

In the datasheet there is a chart "On-Time vs I_ON Current" ... which shows I_ON go down to 1uA


Klaus

 

[D.A.(Tony)Stewart]

Thanks, i also doubt that there is anything in that chip that can detect the 2.3uA.

Not true. Haven't you ever made a CMOS Relaxation Oscillator with a 10 Meg resistor and < 10 pF capacitor?

 

[cupoftea]

Thanks, Yes but i wasnt worried about accuracy.

 

[D.A.(Tony)Stewart]

I see you haven't mastered how to simulate parasitics.

My advice was a result of verifying my advice in LTspice. The consequences increase 50 mVpp output ripple on Vout-15 or worse , overshoot, freq. and damping factors by changing loop gains from +ve or -ve x pF feedback to Ion from each nearby signal.

If you can imagine it, you can simulate it.
Tony