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.

CDCE72010: power up time

Part Number: CDCE72010

I am using this IC in a new design.

The IC outputs clocks properly, freq, etc .... but it doesn't power up well.

Being powered off from an infinite amount of time, once powered up, the device is lasting to output a clocks about 30 or 40 seconds.
Once this happen, if powered off and powered up again, it outputs clocks almost instantaneously.

The values of the registers I am writing to the EEPROM are:

REG0 = 0x682C0230
REG1 = 0x68840241
REG2 = 0x68840002
REG3 = 0x688402A3
REG4 = 0x68840004
REG5 = 0x680C00A5
REG6 = 0xE9040006
REG7 = 0x21840317
REG8 = 0x210C0078
REG9 = 0x69400059
REG10 = 0x0EA41D4A
REG11 = 0x0000089B
REG12 = 0x0000180C

The VCXO is 600 MHz, and the loop filter is C1=100nF, C2/R2 = 22uF/7K, R3=1K and C4=100nF.

Digital control lines, PD and RESET are OK, and the device is in SPI mode.

Could you help me with this problem?
Thank you in advance!
Jose

  • Hi Jose

    Looks like you are running with very low bandwidth. What is the charge pump current and what is the gain of the VCXO? Can you plot the VCO/VCXO control voltage with an high impedence probe in both cases? Also, can you please check the power up ramp time in both cases?

    Best regards
    Puneet
  • Hi Puneet,

    Thank you for the quick response.

    The charge pump current is the default one, 2.2mA and the gain of the VCXO is +100ppm/V.

    The control voltage signal is difcult to plot due the amount of time needed, but it is always the same, a ramp from 0V to 0.55 V aprox, and the time it needs to get to that value seems to be the same, a few seconds, maybe 10s or so ....................... BUT, in the first case, the IC is not outputing clock until 30-40 seconds later, and in other cases, the IC outputs clock even when we are measuring 0.35V in the VCO/VCXO control voltage.

    Hope this helps to make you an idea, looking forward your answer.

    Jose

  • Another circumstance is tthat at the moment we are working without any reference.

    Regards,

    Jose
  • Hi Jose

    What does that mean "at the moment we are working without any reference". Are you not applying any rederence clock input?
    If that is the case, i would propose to check with the reference connected.
    Otherwise, i could only think of difference in supply ramps. Please check that. I can also review the schematics if you share with us.

    Best regards
    Puneet
  • Hi Puneet,

    Thanks for your previous responses.

    Because of the nature of our application, we need to power up without any reference clock input signal ..... in our case, this is a must.

    We have checked the ramps, not only in power supply nets, but also in reset and power down pins, and they seem to be ok.

    We are experiencing the same behaviour explained before, the IC is not outputing the clocks until, may be, 1 minute later than power up. 

    We know the device is working, because it gets warm, but the outputs present no clock.

    Once the device has reached certain temperature, it powers up well, i mean instantaneously, in every retry.

    We think this is a temperature related problem ... 

    Can you help us in any way?? ... are these known effects??

    Thank you in advance,

    Jose

  • Hello Jose

    I have not heard about such chip behavior from any other customer.
    Another thing to check is the RESET pin. How is the signal at RESET pin in those two cases?
    Have you checked this on single unit or multiple units shows such behavior?

    Best regards
    Puneet

  • Hello Puneet,

    We have experienced the same behaviour in the only two prototypes we have at the moment.

    The RESET pin is connected to a pull-up resistor, but we think this is not affecting to the problem, because is the first power up when we experience problems.

    Also, the CDCE72010 starts getting quite warm, or even hot very soon, even when it will put out clocks a minute later .... so we think it is internally operating, configuration has been loaded, and it is not in reset.

    Another experience, we have assembled one of the prototypes with the enclosure, which acts as dissipator, and it is not powering up until almost 15 minutes later..

    Without the enclosure, it was powering up as described above.

    We think this is a temperature related problem. As I described in the previous posts, once it has powered up well, if the circuit is switched off and switched on again, the time to power up depends on the time the circuit has been off ..... i.e. if I keep the board off 5 seconds, the cdce72010 will output clocks almost 5 seconds after been powered ....

    Also, as I stated before, we need to start up without any clock reference signal .... can this be a problem??

    Thanks in advance!

    Jose Luis

  • Hello Puneet,

    Any ideas on this topic?

    Thanks in advance!

    Jose Luis

  • Hello Jose,

    Let me check in our LAB.
    By the way, have you tried powering up by applying an input clock?

    Best regards
    Puneet
  • Hello Puneet,

    That is not possible in our system, because we use the CDCE72010 as main clock generator.

    Thank you in advance for checking it in your lab, looking forward your conclussions.

    Best Regards,
    Jose Luis
  • Hello Puneet,

    I have another question in addition to the previous ones .............. we are using a VCSO (https://www.vectron.com/products_av/vcso/VS-702.pdf, ) instead of a VCXO ....

    We think this should not be a problem, could you confirm?

    Thanks in advance.

  • Hi Jose

    Using this VCSO should be fine.

    Best regards
    Puneet