LMX2595: CHDIV Power On Settling Time

Dylan g

Part Number: LMX2595

When switching from the VCO output to the Channel Divider output using the OUTA/B_MUX, what is the settling time to power up the channel divider?

The datasheet mentions that when the VCO output is selected for both channels, the channel divider is powered down and is only powered up when SYSREF or CHDIV is selected as the mux output.

What is the approximate time it takes for it to power up?

over 3 years ago

0 Noel Fung over 3 years ago

TI__Guru** 108130 points

Hi Dylan,

We don't have data on this.

I could setup a bench to take some data probably in early May, will that works to you?

0 Dylan g over 3 years ago in reply to Noel Fung

Prodigy 170 points

It appears to be on the order of 11µs from what we are seeing from our results.

Thank you but I think we won't need specifics anymore, we are just going to keep the divider on at all times in order to keep consistent settling times across our spectrum

0 Dylan g over 3 years ago in reply to Noel Fung

Prodigy 170 points

Hi Noel,

I have a follow up to this question - would putting the output to the SYSREF option with the SYSREF powered down keep the channel divider on? Also what would the output on the RF port be for a case like that?

Thanks

0 Noel Fung over 3 years ago in reply to Dylan g

TI__Guru** 108130 points

Hi Dylan,

If you use only one RF output, you can set the unused channel to "Divided output", that will keep the channel divider ON all the time.

0 Dylan g over 3 years ago in reply to Noel Fung

Prodigy 170 points

Yes, this was our plan and what we are currently implementing

What I am curious is what would happen is we were to use the SYSREF option with the SYSREF powered down? We are currently trying to improve spurs and figured the divided VCO on the unused RF output might be affecting the spurs in unwanted ways

0 Noel Fung over 3 years ago in reply to Dylan g

TI__Guru** 108130 points

Hi Dylan,

We can set OUTB_MUX to 0x2 (Sysref) but set SYSREF_EN = 0 and OUTB_PD = 1. This will also keep channel divider ON without enabling the sysref blocks.

0 Dylan g over 3 years ago in reply to Noel Fung

Prodigy 170 points

Hi Noel,

We tried this and there was no difference/change in our spurs.

0 Noel Fung over 3 years ago in reply to Dylan g

TI__Guru** 108130 points

Hi Dylan,

Is the spurs frequency equal to the "divided down" frequency?

Also try set CHDIV = 0x0; SEG1_EN = 0x0; SYSREF_DIV_PRE = 0x1; SYSREF_DIV = 0x0.

0 Dylan g over 3 years ago in reply to Noel Fung

Prodigy 170 points

Yep, at least it is predictable where it is going to be. It appears that even with the RFB output muted, there is crosstalk between the two

I currently have all those set and we are still seeing the problem

0 Noel Fung over 3 years ago in reply to Dylan g

TI__Guru** 108130 points

Hi Dylan,

Did you try set OUTB_PWR = 0x0?

Please also try set R33[12] = 0, this is a hidden register, make this bit = 0 shall disable the channel divider driver to RFoutB.

0 Dylan g over 3 years ago in reply to Noel Fung

Prodigy 170 points

Noel - will disabling the channel divider driver to RFoutB keep it powered on when selected? Or will it be powered off when RFoutA is on VCO and RFoutB is set to Channel Divider?

I am currently implementing the pwr = 0x0 change and the R33[12] = 0 right now to see if it helps