- Ask a related questionWhat is a related question?A related question is a question created from another question. When the related question is created, it will be automatically linked to the original question.
Hi,
my problem is similar to this topic. I succesfully generate 125 MHz and 1 Hz outputs, but I cant get lock (i suppouse that chip is in free-run). R14 register status have LOFL, LOPL and HLDOVR flags. Register R13 is clear. I have Tics Pro from january 7 2022, i tried to use fix from this topic with same result.
Please let me know if you need any more information.
Can you help us understand why it would not be locking and what can be done to resolve this?
Hello Maksymilian,
Can you please inform me whether the PRIREF_VALSTAT and SECREF_VALSTAT are high or low? These two bits will determine whether your input reference is valid.
Regards,
Kia Rahbar
Hello Maksymilian,
Please enable the amplitude detector and set the amplitude detector to CMOS Slew Rate Detector Mode. For a 1PPS input, the amplitude detector must be configured in this fashion.
Regards,
Kia Rahbar
Hi Kia,
still the same. I tried already increase validation timer and modify 1 PPS Phase detector and its still doesnt lock.
My schematic hardware is simple and tested with other PLL from Silicon Labs:
1 PPS signal on PRIREF_P pin :
Status regs from running hardware:
R14 = 240 (LOPL_DPLL, LOFL_DPLL, HIST, HLDOVR is high, the rest - low)
R20 = 240 (same as R14, so i think that LOR never get high)
R13 = 8 (LOL_PLL2 only - i dont use APLL2)
R411 = 0 (PRIREF and SECREF are not valid)
My configuration:
My configuration file from TICS Pro:
Hello Maksymilian,
Apologies for the delay. I spoke with my team and we believe we have determined your issue.
The cause for your issue is that your XO input must have a maximum ppm accuracy of 4.9 ppm when using a 1PPS reference and a XO frequency of 12.8 MHz. The maximum allowable ppm error of 4.9 ppm is calculated using the following equation: max XO ppm = 1PPS cntr value (shown below)/ XO frequency (in MHz) = 63/12.8 = 4.9 ppm
If the ppm accuracy of the XO does not meet this requirement you will not be able to lock to the 1PPS input.
Also please note that I was incorrect about enabling the amplitude detector control. Please disable this feature.
Regards,
Kia Rahbar
Hi Kia,
we try to use other generator LINK and we multiple frequency x3 with this LINK (we use rigol dg5252 synchronized to external GPS earlier just for test). Now my configuration looks like:
Still cant lock and no valid reference in R411. We check raw register setting and recognize that R46:
is set to 0x11 (AC-Differential (external termination)) for both reference. I think that it should be CMOS or DC-differential, but i dont see option to change it in TICS Pro. Im not shure that i can change it raw without consecuences in other settings (when i change it in raw registers and try to change some other options in GUI TICS pro overwrite it everytime).
We're figthing with this about 3 weeks and we really need to make this work. Our clients wont wait forever.
[PS]
We're already validate our generator and its much better that 1ppm ( device does not allow for a more accurate measurement). 1PPS phase detector fails all time - when we disable it R411 says that signal is valid but stil doesnt lock. Witch flag says that 1PPS phase detector is ok?
Hello Maksymilian,
I have created a new configuration and tested it in our lab.
Please follow these steps and your device should be locking properly:
1. Load this tcs file into TICS Pro:
2. Once the file has been loaded, press the soft-reset chip button shown below:
3. Then, readback the status bits and everything should be locking properly as shown below in my setup:
Please note it will take 10-60 seconds for the LOFL_DPLL to go low and may take a few to many minutes for the LOPL_DPLL to go low as it takes a significant amount of time for the device to lock to a small frequency such as a 1PPS signal.
Regards,
Kia Rahbar
Hi Kia,
i load and program your configuration and its better now. PRIREF is valid (R411), chip switch to PRIREF (R20[3]), but stil cant lock (LOFL and LOPL flags - valid reference is working for 3 hours).
When i load your file TICS Pro show error:
There is a lof of non-exists register in TICS Pro . Can u send a proper device profile?
Hello Maksymilian,
You're welcome!
A possible reason for why you may not be able to relock is that there is a drift in the accuracy due to the tuning word history when the device is in holdover. Please try turning off the tuning word history as shown below and see if that solves your issue.
Unfortunately, we do not have a flag for valid history.
To determine if the device has entered holdover, the HLDOVR flag can be used. If the flag is high the device is in holdover. The register associated with this flag is R14[4].
Regards,
Kia Rahbar
Hello Maksymilian,
Unfortunately, the only way to force relock is with a soft reset.
Yes, a 19.44 MHz OCXO will provide a greater ppm valid range (3.24 ppm (63/19.44)) than the 60 MHz XO input (1.05 ppm (63/60)), so it will be much easier for the device to relock to the 1PPS input.
I would recommend making the changes on your end first since you have it operating correctly on the initial lock. Please follow these steps:
1. Load the working configuration and get the 1PPS to lock.
2. Navigate to the XO page and input the new 19.44 MHz XO frequency as shown below.
3. Navigate to the set outputs page and click calculate frequency plan as shown below.
4. Perform a soft-reset and readback the status bits to see if everything is locking properly. If it is, save the file and load it again to try your relock test again.
I can make the update and test the configuration next week if you are not able to get it working on your end.
Also another test you can perform is to increase the tuning word history count and delay. This will allow for the holdover frequency to be much more accurate when you are in holdover and will allow you to relock much easier.
Regards,
Kia Rahbar
