LMX2595: Phase synchronization of two devices

Anjan Kumar.N

Prodigy 100 points

Part Number: LMX2595

As per the data sheet doubler can be used in phase sync mode. The LMX2595 is configured for the following setting:

Fvco = 9000 MHz,

VCO2X_EN = '1',

Fout = 18000 MHz,

Fpd = 50 MHz,

N = 180,

Two LMX2595 devices are mounted on separate PCBs, The input reference signal is synchronized for the both the device. the devices are used in the phase synchronized mode. Part to part delay variation is nullified using the MASH_SEED. Now the temperature difference between the is about 2°C. As per the datasheet the delay variation with respect to temperature is +2.5 ps/°C. So the delay variation between the device will be 5ps. Due to the above delay variation, How mush will be the phase difference between the two outputs?

over 4 years ago

0 Noel Fung over 4 years ago

TI__Guru* 88320 points

Hi Anjan,

The delay variation spec was measured without VCO2x and was measured at 100MHz output.

If the delay remains the same when VCO2x is enabled and at 18GHz output, then the delay variation between the two LMX will be 5ps. That is, their phase will be differ by 32.4deg.

0 Anjan Kumar.N over 4 years ago in reply to Noel Fung

Prodigy 100 points

Hi Sir,

Thanks for the reply,

1. Is the delay variation with respect to temperature same with VCO2x enabled?

2. Our requirement is generation of 18 GHz at two different PCBs with constant phase difference between power ON. Is this possible?

3. In this case, since the VCO out is 9 GHz, the phase difference due to 5ps delay will be 19.4deg. After the x2, what is the phase difference? should it be same...

0 Noel Fung over 4 years ago in reply to Anjan Kumar.N

TI__Guru* 88320 points

Hi Anjan,

1. I don't have the data with VCO 2x enabled.

2. According to the flow chart of Figure 33 in the datasheet, for Cat.1 sync, the devices will in-phase. For Cat.2 and Cat.3 sync, we need a sync pulse to make them in phase.

3. The phase difference is measured at RF output, so if the difference is 5ps, their phase difference is 32.4deg.

0 Anjan Kumar.N over 4 years ago in reply to Noel Fung

Prodigy 100 points

1. Can the delay variation with respect to temperature with VCO2x enabled be tested? Can you provide the results?

2. As per data sheet the temperature coefficient for propagation delay is 2.5 psec/deg C. What is the tolerance/variation of temperature coefficient from device to device?

3. As per the data sheet, The delay through the device varies from part to part and can be on the order of 60 ps. Does this value vary/change over the time/age?

4. As per the above reply, "delay variation spec was measured without VCO2x and was measured at 100MHz output", Can this be tested at higher frequencies like 18 GHz?

0 Noel Fung over 4 years ago in reply to Anjan Kumar.N

TI__Guru* 88320 points

Hi Anjan,

1. this involves huge amount of work, we don't have bandwidth to do it right now.

2. We don't have this data.

3. I think this is possible but the change should be small. However, I don't have this data.

4. We don't have test equipment to test the delay at this frequency.

0 Anjan Kumar.N over 4 years ago in reply to Noel Fung

Prodigy 100 points

The answer provided does not resolve the issue, and needs further attention.

0 Noel Fung over 4 years ago in reply to Anjan Kumar.N

TI__Guru* 88320 points

Hi Anjan,

Sorry that we cannot help. Please understand that the data in the datasheet is the only data we have right now.

Clock & timing

Clock & timing forum

LMX2595: Phase synchronization of two devices