LMK01801: Clock Distributer

Terry Liu70

Part Number: LMK01801

Hi,

In my application, input = 1GHz sine wave signal with -150dBc/Hz phase noise @100k offset. I like to introduce a frequency divider then output a 100MHz signal with better than -165dBc/Hz phase noise @ 100k ofs. I looked into LMK01801, but the noise floor doesn't seem to be good engouth.

Please advise a proper moldel number that meets above requirements.

Thanks,

Terry

4 months ago

0 Michael Srinivasan 4 months ago

TI__Genius 14381 points

Hi Terry,

I am confused. If your signal source has a noise floor of around -150dBc/Hz @ 100k offset, then any device you instantiate following that device will be subject to the noise floor of the signal. A noise floor of -165dBc/Hz @ 100k offset is only going to be possible if you have a source with a noise floor lower than that, and then - depending on the frequency - your noise floor can be satisfied by the whole signal chain.

Thanks,

Michael

0 Terry Liu70 4 months ago in reply to Michael Srinivasan

Intellectual 500 points

Hi Michael

Input 1GHz signal with phase noise of -150dBc/Hz, when frequency divided by 10, output a 100MHz signal, the ideal phase noise should be -150 - 20 * log (10) = -150 - 20 = 170dBc/Hz. Thus, I think -165dBc/Hz target makes sense.

Again, please advise if TI has any part that meets this spec, Thanks,

Terry

0 Michael Srinivasan 4 months ago in reply to Terry Liu70

TI__Genius 14381 points

Hi Terry,

You would actually add that 20 * log (10), meaning you would be seeking a target of -135db/Hz.

This part would certainly meet that spec.

thanks,

Michael

0 Terry Liu70 4 months ago in reply to Michael Srinivasan

Intellectual 500 points

Hi Michael

You mean,

Device input = 1GHz, -150dBc/Hz @ 100k

(LMK01801) Divide by 10

Device Output = 100MHz, phase noise will add 20dB to -130dBc/Hz?? Thus -135dBc/Hz floor is good to use...

This is beyond my knowledge, I never saw phase noise getting worse when dividing frequency.

Terry

0 Michael Srinivasan 4 months ago in reply to Terry Liu70

TI__Genius 14381 points

Hi Terry,

I just realized I made a mistake - I was confusing multiplication for division.

The noise floor for this device will be roughly -162dBc/Hz at 100MHz. We do not have a device with a built in divider that also has a noise floor of -165dBc/Hz at that high of a frequency. Will that work?

Thanks,

Michael

+1 Terry Liu70 4 months ago in reply to Michael Srinivasan

Intellectual 500 points

Hi Michael

Thanks for your kind assistance. I just found an ADI part that offers -170dB noise floor.

Terry

Clock & timing

Clock & timing forum

LMK01801: Clock Distributer