This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMX2492: PLL not locking

Part Number: LMX2492

I have a self made LMX2492 PLL circuit that I cannot get to lock.

1. The SPI communication seems to work OK.

I send SWRST : 0x000204

I set POWERDOWN to "Power up, ignore CE" : 0x000201

I config MUXout_MUX to "output readback" and set to MUXout_pin to "Pullup/Pulldown output" : 0x00273a

At this time I am able to read back the value of any register, ie. if I send 0x801000 (the PLL_N[0:7] register) I can see 0x64 on the MUX_out pin using an oscilloscope.

Just to be complete, after each of these SPI messages the LE line is pulsed.

So far so good.

2. Next I try to read back the R/2 counter to check if the reference oscillator input have propagated to the PFD.

The SPI programming is as follows

SWRST : 0x000204

POWERDOWN : 0x000201

MUXout_MUX set to R/2 and MUXout_pin set to Puup/Pudn : 0x002782

It is my understanding of the datasheet that I at this stage should see 25 MHz signal at the MUX_out pin (my input oscillator has frequency 50 MHz). Sadly it is dead 0 Volt.

I measure the differential voltage between OSCin and OSCin* (pin 9 and 10) to 0.9 Vpp. According to the datasheet section 7.6 VOSCin must at least be 0.5 Vpp.

Hence, my input signal should have sufficient magnitude.

I have also tried to read back N/2 (ie. program 0x002792) with the same result. In this case it is more difficult to accurately measure if the input signal has sufficient power (> -5 dBm)

although I think this is the case.

3. My next effort was to read back the TRIG1 input.

Again, it is my understanding of the datasheet that if I program MUXout_MUX to "Input TRIG1", the logic value on the TRIG1 input should be relayed to the MUX_out pin.

My third SPI message in this case reads : 0x00270a

After this I toggle the TRIG1 input a number of times as I measure the MUX out pin. Once again, absolutely no action.

What could be the problem here ? All power and GND pins seems to have correct voltages.

1. Errors in my SPI programming

2. Soldering problems. I measure the voltages on the nets leading into the chip. It is difficult to measure directly on the pin on a chip scale package.

3. The IC is partly malfunctioning.

Item 2 and 3 requires de-soldering. Before I try that I wanted to ask if there are any other possibilities. Are there any more things I could try first ?

Thanks.

  • Hi Harald,

    I wonder if you were actually able to write to the device. Please note, when Vcc is up, the internal reset circuitry will reset all the registers to the default POR value. I suspect you were reading the default value but not the written value. 

    To verify this, follow:

    1. Vcc power up the device and record the current

    2. Write 0x000200 to powerdown the device, you should see around 40mA current drop.

    If you didn't program PLL_R after Vcc power up, then the R-divider is default to 4. Given OSCin is 50MHz, then fpd = 12.5MHz. You should get 6.25MHz clock from the MUXout pin if you can successfully configure MUXout to R/2 push-pull output.

  • Hi Noel,

    I can confirm there is approximately a 40 mA current drop when the device is programmed into power down.

    Additionally after POR if I read PLL_R[7:0] I can see 0x04 on my oscilloscope. If I then continue to program the same register let us say 0xaa and then try to read back the value I see 0xaa on my oscilloscope. The device clearly responds to the SPI communication.

    But I am unable to get any R-divider out. This could be because the the differential voltage between OSCin and OSCin* is too small. But as I said in my first post I measure 0.9 Vpp, according to the datasheet 0.5 Vpp should be sufficient. Some kind of contact/solder issue could be the problem although everything looks fine when inspecting with a microscope.

    But I am also not able to read back TRIG1 input as I explained in my first post. That is confusing and makes me wonder if this is some kind of programming error after all.

    Thanks for your answer.

  • Hi Harald,

    Looks you are able to write and read the logic correctly.

    I assume you were not able to make it lock to the VCO? Could you measure the CPout voltage?

    I cannot think of anything wrong at the moment.