AD9548 USB driver problem.

Hi,

To make the board be "found" by the PC, go into UG-639, the user guide of the evaluation board. You can download this guide from here: 

https://www.analog.com/en/design-center/evaluation-hardware-and-software/evaluation-boards-kits/EVAL-AD9548.html#eb-documentation

Please go to page 6 and follow the instructions at bullet 6:

This will allow the PC to recognize the board, which will make the USB Programmer to identify the board. Use the instructions on the page 6 of UG-639  to program the USB EEPROM.

Petre

Hi, Yes I have read this, and tried it, but the only thing you can do from the Windows Device Manager is to update the driver. But were is the driver? 

I managed to update the driver, I had to use an old web browser to be able to see the instructions in the Driverguide. Anyhow, there is still no difference, The board is not shown in the USB Eval Board Programmer or the AD9547_48 Evaluation Software.

HI,

 I remember doing this procedure some time ago and it worked very well. I had the USB programmer recognizing the board like in Figure 3 in UG-639, Cyusb-1.

Are you sure the old AD9548 evaluation board is still functional?

Maybe you should order a new board or try the install procedure on a different pc.

Please also note that if you want to control 1PPS  GPS clock, we now recommend the AD9545 instead of the AD9548. For other control tasks, better use the AD9542. Please take a look at the AD9545 family of devices: AD9545/44/43/42.

Petre

Ok, thank you. Probably the board is broken. I can see the LED CR2 is red (steady) and also the two green LEDs CR3 and CR5 is on. 
I will check the AD9545.

/Per

Hi Petre, 
Can you please guide me to the right choice of IC?
I have a project where I intend to discipline a Rubidium Oscillator using a PPS from a uBlox GNSS receiver. When starting up the PPS from the receiver will be available as long as needed but in normal mode the PPS is only switched on for shorter periodes during a day. 
This means that the circuit needs to have good holdover performance and take into account the Rubidium aging etc. The system must also produce a low jitter PPS.

Is AD9545 the right choice? 

/Per

HI,

I do believe the AD9545 is the right choice. It can lock onto a 1PPS clock with a DPLL loop bandwidth of 50mHz. The 2.4GHz system clock will be created using a 52 MHz crystal resonator which will make the AD9545 outputs to have the best phase noise performance. You will need to use a OCXO with around 10MHz or 20MHz to stabilize the effect of the system clock wonder (induced by the 52MHz crystal wonder). The OCXO will make the DPLL to permanently lock in phase and frequency onto the 1PPS.

In holdover, the outputs will be as stable and as precise as the OCXO.

I do not know how you will account for the Rubidium aging. If you give me more info, maybe we can find a way.

One more note: some Ublox GNSS receivers have a quantization noise that appears at the 1PPS output as a spur within the 50mHz bandwidth. This means this spur will be seen at the AD9545 output because the DPLL tries follows the 1PPS. Ublox provides a measure of this error and the controller can update the AD9545 to compensate it, but we need to create the 1PPS loop in a different way.

When you'll have an AD9545 eval board, please give me exact info about what you want the AD9545 to do and I'll create a configuration file for you to start the development.

Petre

Hi Petre,

Thank you for your support. I will come back to you in this matter. 

Per

Hi Petre,

Is it possible, via the AD9545, to measure the time difference between two PPS? My idea is to measure the difference between the GPS PPS and a PPS created by the Rubidium Oscillator. This value will then be used by our MCU to correct the frequency of the Rubidium oscillator.

Per 

HI,

Yes, the AD9545 can do this. See section Timing skew measurements using two TDCs, page 130 in the rev C data sheet.

Petre