• State Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 3 answers
  • Subscribers 23 subscribers
  • Views 1658 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

LMK61E2 External Ref Input

BILL RESS
Prodigy 60 points
Other Parts Discussed in Thread: LMK61E2, LMX2581, LMX2582, LMK03318, LMK03318EVM, USB2ANY, CODELOADER

I've been evaluating the LMK61E2 and while the phase noise meets my project requirements, the stability does not. Is there any chance that this device might be available (someday) with an high stability external reference (50 MHz OK but 10 MHz better)? Or - - - does TI have another device in a similar "crystal" type package, which has the phase noise of the LMK61E2 but also has the external reference capability? I searched but nothing jumped out....Thanks Bill

  • 0
    TI__Mastermind 24205 points

    Hi Bill,

    Unfortunately as you state here, LMK61xx (oscillator device family) is only speced to provide a reference clock that is accurate to within +/-25ppm at best. You could perhaps consider using LMX2581 or LMX2582 family of RF PLLs  with an external 10 MHz OCXO/TCXO reference that can be programmed via microwire. Although not available in standard oscillator footprint, these devices have excellent phase noise and wide range but do not integrate the resonator (crystal).

    If your end application volumes are significant, we might consider releasing a LMK61xx variant without an internal reference clock. Please provide us more details of your end application requirement.

    Regards

    Arvind Sridhar

  • 0 in reply to Arvind Sridhar
    Prodigy 60 points
    Hi Arvind,

    Thanks for the prompt, although disappointing, reply. The LMX2581/82 phase noise at 1 to 10 KHz doesn't come close to the phase noise performance of the LMK61E2. The LMXxxxx devices also lack the EEPROM feature which is important in our application. With the addition of the external reference capability in the LMK61E2 you would solve the accuracy and stability issue and yield, in my opinion, an outstanding device for high-end communication signal source applications.

    Unfortunately we do not have the volumes to justify a custom version, but I would hope your design/marketing team would see the "outstanding performance setting" potential of the device by just adding the external reference capability.

    I'll keep my fingers crossed that the situation might change...Bill
  • 0 in reply to BILL RESS
    TI__Mastermind 24205 points

    Hi Bill,

    Thanks for providing good feedback. The reference input path for LMK61xx was in fact highly optimized for improving the close-in phase noise performance. I will discuss this use-case further with marketing. This is not the first time this request has come up.

    Regards

    Arvind Sridhar

  • 0 in reply to BILL RESS
    TI__Mastermind 32575 points

    Would the LMK03318 work for you?  It can accept a 10-300 MHz external reference (either ext. clock or ext. crystal) and achieve flexible frequency translation/clock generation using ultra-low noise Frac-N PLL & VCO, and it also has on-chip EEPROM.  It has up to 8 differential outputs with 6 individual output channel dividers, but any unused output channel can be powered-down

    Regards,
    Alan

  • 0 in reply to Alan O
    Prodigy 60 points
    Hi Alan,

    Thanks for the reply. I have looked at the LMK03318 but it had a few disadvantages compared to the LMK61E2 for my specific application. The main one is the high current consumption compared to the LMK61E2 and then its board volume (48 pins versus 7x5 mm crystal). But I will look at it further. It seems for in-system programming I would need the TI USB2ANY device. My browsing indicates I need to go to clock_support@list.ti.com but that gives me a URL error. So maybe you can help. Here's the info the LMK03318EVM Guide said I needed:

    1. Request/Reason: 1 pc. USB2ANY module for LMK03318 in-system programming/prototyping
    2. Company Name: High Sierra Microwave
    3. Application/End-Equipment: Microwave Communication Equipment
    4. LMK03318 Est. Annual Volume/Year: 2000
    5. Ship-To Address:
    High Sierra Microwave
    86 Rancho Del Sol
    Camino, CA 95709
  • 0 in reply to BILL RESS
    TI__Mastermind 32575 points

    LMK03318 can be programmed by any I2C master/host, so USB2ANY is not required.  USB2ANY is just a suggested option for device programming since it is a readily-available, low-cost USB platform developed by TI to support the various communication/control needs across many different devices/EVMs.

    LMK03318EVM already has USB-to-I2C circuitry on-board (it actually uses the same USB uC circuitry as the USB2ANY module), so a USB2ANY module is not required for evaluation of the LMK03318EVM.

    If you still want to request USB2ANY, please send your request via EMAIL to clock_support@list.ti.com.  This is an EMAIL address, not a Web URL.

    Alan

  • 0 in reply to Alan O
    Prodigy 60 points
    Hi Alan and Arvind...Taking your advice I have incorporated the LMX03318 to the design and are attempting to program it in system using the Code Loader 4 GUI and the USB2ANY. We are communicating just fine but have not be able to work through the myriad of register options (which are always well explained) using the LMX03318 EVM guide, the Code Loader guide and the LMX03318 data sheet. I have spent many hours so far fighting through without success.
    I believe what would really help me is you sending me a EEPROM file which I can upload to the Code loader and tehn use it to program the LMX03318. The EEPROM file should be for the simple case of only the Clock0 being programed for 100 MHz output (CMOS) with a 10 MHz input at PRIREF. Perhaps I can then see what's going on and move on.
    Thanks a bunch...Bill
  • 0 in reply to BILL RESS
    TI__Prodigy 20 points

    Hi Bill,

    I've attached the EEPROM (.epr) file you requested, configured with a 10 MHz input on PRIREF and a 100 MHz CMOS output on OUT0. I committed the registers to each page of the EEPROM so any of the pages should give you the desired output after loaded onto the EVM and reset. I've also attached a .mac file for the configuration, if you would like to load the register settings in the CodeLoader GUI before programming the EVM.

    Regards,

    Jesse

    lmk03318_config.zip

  • 0 in reply to Jesse Baker50
    Prodigy 60 points

    Hi Jesse,

    Thanks for the prompt posting to the ,epr and .mac files. While the weekend effort wasn't entirely successful, those files helped give me a needed sanity check. Everything seems to go OK right up to PROGRAM EEPROM< SRAM step. When the button is clicked, the bottom  status tray says "EEPROM Programmed...DONE" but it it appears very quickly. I'm loading Page 0 only. Not like the "Write GUI Map > SRAM" button click which takes many seconds to load the GUI Map. So I'm not convinced the EEPROM was indeed programmed, since when I place a 0 on "HW_SW_CTRL" and 0's on GPIO3 and GPIO2, then PDN - I get no output signal on OUT0_P (10 MHz is connected). I must have some sequence or "switch" screwed up still. The device current is nominal and I can see a 200 MHz signal on OUT0_P when a 50 MHz REF signal applied (I assume this is loaded in ROM). Oh - I found that Soft Pin Mode switches must the left "open" while doing the EEPROM programming right?

    What have I not done right?

    Thanks...Bill

  • 0 in reply to BILL RESS
    TI__Prodigy 20 points

    Hi Bill,

    ROM page 0 and 1 (HW_SW_CTRL=1, GPIO[5:0]=0 & 1) will output a 200 MHz AC-LVDS signal on OUT[0-1] with a 50 MHz LVCMOS input on PRIREF or 50 MHz XTAL input on SECREF, as shown in the LMK03318 datasheet, page 65, which may be the output that you are seeing.

    Try this sequence to see if you get the desired output:

    1. Find the EVM as an I2C device using Scan I2C bus in Port Setup tab of CodeLoader

    2. Restore the .mac file provided using File -> Restore

    3. Use <CTRL+L> to Load all device registers from GUI (updates the device from the GUI)

    4. Apply a valid 10 MHz LVCMOS clock (0 - 3.3 V levels) to PRIREF_P input

    5. Toggle RESETN_SW bit in CodeLoader to trigger device startup routine (top left of Inputs/PLL, Outputs, or Status tabs)

    6. Observe device outputs to confirm proper configuration/operation

    - 200 MHz LVCMOS clock on OUT0_P/N on oscilloscope

    - STATUS LEDs indicate PLL lock and valid PRIREF

    7. Once device operation is confirmed, program EEPROM

    a. (1b) import EEPROM file

    b. (2) Write GUI Map -> SRAM - may take a few seconds

    c. (3) Program EEPROM <- SRAM - should happen instantly

    8. Set the following control pin jumpers to LO position ( HW_SW_CTRL, GPIO3, GPIO2 to hardware reset/restart device in Soft Pin mode, EEPROM page 0

    9. Set GPIO0 to HI position to enable outputs (de-assert SYNC) <- this could be why you saw no output in EEPROM mode

    10. Toggle PDN button to restart device in EEPROM Page 0

    11. Find I2C device using Scan I2C bus in Port Setup (same as in step 1)

    12. <CTRL+R> to read all device registers (updates GUI from device)

    13. Confirm that device configuration is the same (should look like configuration restored in step 2)

    Regards,

    Jesse