CDCE706 evaluation module: jitter on reference clock

Hi,

Have you checked the register settings of CDCE706? There is a setting to add some amount of phase shift or no phase shift to the output.

It may have been turned on.

Regards,

Sid

Thanks for this idea, Sid. The registers are set to no phase shift. And in any case, my understanding is that this setting is for consistent phase shift relative to other outputs. The signal I'm getting is not regular! (There are roughly the right number of periods in any one second, but the variation in period of any one cycle is large.)

Hi Roddy,

I was using CDCE62005 in one of my projects and had similar problems. It turned out to be 2 things

1. There is a SYNC bit which synchronizes the output, which needs to be turned ON. I think it is register 26 in CDCE62005, not sure abt its equivalent in CDCE706

2. The power-up sequence is also important. The power-up sequence ensures that the output is phase-locked and also ensures that output is not jittery.

Hope this helps.

Regards,

Sid

Hi Sid

I've spent quite a bit of time messing around in the lab with this board this morning. It appears that the frequency is also sometimes too low by about a factor of 2.5 (independent of the frequency I've programmed the device at).

Thanks for your suggestions: in reply:

1. There doesn't seem to be a sync pin for the CDCE706.

2. It specifically states on p.36 of the CDCE706 datasheet that "There are no power-supply sequencing requirements, as the three power nodes are separated from each other. So, power can be supplied in any order to the three nodes". However, I still tried powering up the supplies in different orders, but without any change.

What was weird is that during about an hour in the lab messing around with this board, there was about a 5 minute interval when it seemed to work okay -- frequency correct and negligible jitter. Then the trace flickered intermittently for a little while and then the signal went back to being jittery (and wrong frequency) as before.

Regards,

Roddy

Hi Roddy,

As you are usign the evaluation board of the CDCE706, could you post here the setup saved in the device? .hex will do.

Certainly -- attached (both .hex and .csv).

(P.S. I have tried changing the power supply, oscilloscope, parallel cable and SMA cables and nothing changes to the board's output.)

@mapefer3: If you happened to look at my setup, is there anything wrong it?

I have checked your setup and if you are using a 27MHz XTAL (not a cristal oscillator) it sould be ok. Do you have any screenshots that show the problem that you could share with us?

I'm using the 27MHz crystal that the board comes with. (I haven't reconfigured it.) I attach a PDF file showing screenshots of the configuration screen, and data from the TDS1012B scope for two of the signals. You'll notice immediately that the frequencies aren't correct, but additionally (and no less importantly) the jitter is significant. (In the meantime I've also put the board in a metal enclosure in case there was a problem with RF interference, but this hasn't changed anything.)