LMX2582EVM: LMX2582EVM

ifte hossain

Part Number: LMX2582EVM
Other Parts Discussed in Thread: LMX2582

Two problems:

1st: When unpacked the evaluation board and following the evaluation guide with on board 100 MHz Oscillator, Output at does not lock at 1800 MHz. I tried with default setup. I have asked distributor and they are telling me, they would not be able to help and I have to go to this forum.

2nd problem:

I used external 100 MHz and now it locks; but lot of close in spurs, like 400 Hz and multiple. Is there any setup I should worry about.

In both cases i am trying to use default settings.

Thank you for your quick help.

regards,

Ifte

over 5 years ago

0 Noel Fung over 5 years ago

TI__Guru* 81230 points

Hi Ifte,

Were you using TICS Pro to program the evaluation board?

0 ifte hossain over 5 years ago in reply to Noel Fung

Prodigy 40 points

Hi noel,

Yes, I am using TICS Pro.

0 Dean Banerjee over 5 years ago in reply to ifte hossain

TI__Mastermind 27581 points

ifte,

In general, you should not expect 400 Hz spurs from the LMX2582 in default mode. Only if you had some fine fractional channel spacing should you expect this. However, this is not the case for default mode.

I suspect that these spurs are actually from the power supply.

In the USA, power supply spurs are typically at multiples of 60 Hz, but other countries it can be at multiples of 50 Hz, and in aircraft it is often at multiplies of 400 Hz. You can try setting MASH_ORDER=Integer on TICSPro to be absoultely sure it is not fractional spurs.

Regards.
Dean

0 ifte hossain over 5 years ago in reply to Dean Banerjee

Prodigy 40 points

Hi Dean,

Thank you for your reply. I initially thought about that power supply spurs since they were so close-in and multiple (seems) of 60 Hz. So I actually solder a 2200 uF leaded onto the eval board and tried it. Did not help. Next I tried using one of those charger's that is well regulated(I believe) and isolated; I took that charger 5.5 volt output and used in a LDO which has output of 3.3V. still didn't help.

Then, one of other problem is EVAL board with on-board oscillator doesn't lock even though lock detect light is Green.

I am going to go try your suggestion and see if I have any luck. I will also try to divider in R-register and see.

Regards,
Ifte

0 Dean Banerjee over 5 years ago in reply to ifte hossain

TI__Mastermind 27581 points

lfte,

2200 uF will probably do very little for a 400 Hz spur. Also, aside from the power supply, sometimes they can come from the equipment including the signal generator or spectrum analyzer itself. Also, sometimes they couple through the cables. The PSRR of the outputs is 0 dB as they are pulled diret to Vcc.

For the not working board with green lock detect light, try cycling power, go to the bits/pins page and toggle the RESET bit, and then load all registers (Ctrl+L) and see if this helps.

Regards,
Dean

0 ifte hossain over 5 years ago in reply to Dean Banerjee

Prodigy 40 points

Hi Again Dean,

That cap was actually a 4700 uF not 2200. I had some long lead to connect it to PLL EVAL. So I assume 1 ohm and that should give cutoff around 34 Hz.

Anyway this made me think about our 100 MHz CLK board (something I trusted to be clean) where the 3V3 LDO resides.

I have lot of measurements and plots that I took; I own't bore you with all that. Summary is

It was very clean up to 500 Hz (may be a very small one at 120 Hz, ~-140 dBm)
From around 1 kHz to 600 kHz quite a few spurs
I believe some of these could be modulating on to the PLL output. So I will rest on close in spurs with our own Clock.

But with on board Clock, I tried two eval boards, both are same and do not LOCK, very slow drifts. I also tried your suggestion of RESET, didn't help.

Thank you for all the help.It's nice to chat with an Industry guru on PLL.

Regards,

Ifte

Clock & timing

Clock & timing forum

LMX2582EVM: LMX2582EVM