LMX2594: LMX2594 MUXOUT pin

Hi Jonathan,

My coworker will get back to you in a few hours.

Regards,
Hao

Hi Jonathan,

Here's a few debug questions:

• Does the customer follow all the recommended programming sequence steps, including the initial device reset and writing FCAL_EN=1 an additional time (R0[3]=1) at the end of the programming sequence, before checking the state of R110?
• Have they checked the Vtune voltage with a multimeter to confirm the charge pump voltage is not railed (as indicated by the readback status)? It may be the case that the VCO frequency is fixed, but still too close to one of the rail voltages.
• How is LD_TYPE set? (R59[0])

Regards,

Derek Payne

Hello Derek,

Thank you for the response. 

Here is the response from our customer regarding the questions:

-          Yes, procedure ist R0[3] = 1, then R0[3]=0 then programmimg from Register 78 downto 0. Following a short Delay FCAL_EN =1 before checking the registers.

-          Vtune Voltage is at 1.226V

-          R59[0] is set to 0

Best regards,

Jonathan

Hello Jonathan,

It seems that the registers are being programmed correctly, and the system is truly locked. Under normal circumstances, R59[0]=0 should assert lock detect immediately after VCO calibration. Are they using full or partial assist in their calibration?

When you set R0[2]=1, does the readback value of R110 change?

Regards,

Hi Derek,

They use partial assist for the calibration.

When they set R0[2]=1,  They can’t read back because it shows lock detect. Is that right?

Nonetheless when R0[2]=1  MUXOUT_LD_SEL goes high shortly after I write R0[3]=1.

Regards,

Jonathan

Hi Jonathan,

Can you confirm that you don't always get wrong R110 lock detect at different frequencies?

Make R59[0], LD_TYPE = 0x1 will ensure the MUXout pin report the lock status frankly. As far as I know, the MUXout pin and R110 use different lock detect mechanism, so there is chance that they don't always align. 

Just on little thing that need to be clarify here, to avoid confusion. It didn’t mean MUXOUT_LD_SEL, They meant just the MUXOUT Pin. 

Hi Noel,

Here is the response from our customer. Sorry for the delay. 

----------

So I’ve checked different frequencies (whole range of VCOs, different divider settings with and without the divider included, so basically random frequencies from 10MHz to 15GHz) ,

 With R59[0]=0

Readback basically stays the same except that rb_VCO_SEL (R110 [5:7]) changes as expected. So I always get 0x6X8 as readback where X is the VCO_SEL field.

 With R59[0] = 1

Readback of R110 now shows that it is locked, as I always get 0x4X8 as readback where X is the VCO_SEL field.

 For both cases:

I checked every frequency as well with MUXOUT_LD_SEL R0[2]= 1 and it always says that it is locked. Up to 3GHz (my Spectrum Analyzer can’t go higher) I’ve also checked it with the Spectrum Analyzer and can confirm that it is indeed locked.

 I’m still confused about R110[3] being set and the behavior with R59[0]=0. Is it possible that “Unlocked (Vtune High)” is mixed up with “Invalid State” in the datasheet and it shows that LD_VTUNE is invalid when R59[0]=0?

 Anyway, I think that answered my question in the first place! Thanks a lot for your help!

---------

Thank you and best regards,

Jonathan