DS90UB954-Q1: REFCLK Spread Spectrum Feature Problem

Hello Carlos,

What I/O level are you using? Does it match the I/O level for the VDDIO of 954? I see that at 1.8VIO the spreading modulation is +/-0.7% which is exceeding the specification of < +/-0.5% spread per the datasheet. 

Also are you using synchronous mode for the connection to 953? or non-synchronous mode?

Thanks,

Casey 

Hi Casey,

Sorry I forgot it, we are working at 3.3V, I have checked this parameter before. So we should be inside...

I will check the mode of the 953, but I am almost completely sure that we are working in the 953 side also in Sync mode (In the MODE pin) but let me check this...

Could be possible that the time which the SS clock takes to stabilise the frequency is bit higher and we are sending or we are trying to communicate with the 953 in this settling time?

I will write here as soon as I will check the 953 Mode.

Thanks!

Hello Carlos,

I don't think it has to do with settling time. What is the modulation frequency for the spreading and the spread shape? Also I'm a little concerned because the +/-0.4% spread is listed as a typical number. Please confirm min/max specifications are within the datasheet limit. 

Best Regards,

Casey 

Dear Casey,

I will check the Modulation Frequency with the clock supplier. What do you mean with: "I'm a little concerned because the +/-0.4% spread is listed as a typical number.?", that the real value is not 0.4%?

I will check this Modulation Frequency because I think is the only value which is not in the Clock Specs.

Thanks!

Hello Carlos,

What I mean is that datasheet typical numbers imply that the min/max could be different. For example, the typical spread could be +/-0.4% but the max value could be > +/-0.5% which would violate the device capability. So please verify that the maximum spec for the spread percentage is < +/-0.5%

Best Regards,

Casey 

Dear Casey, 

I will try to get other oscillator but with lower Modulation Output Deviation. 

Anyway I have been talking with the Oscillator supplier and they sent us measurements, and the real deviation is 0.38%, so should be inside the requirements. 

I will check the I2C, to check when is failing the comunication, and I will give you the information about when is failing the link "backchannel". 

Do you have any recomended oscillator with Spread Spectrum feature tested by your team? To have a reference. 

Thanks and regards!

Thanks! I have just ordered the same P/N, we will check, and I will come again with the results. 

Thanks and regards!

Dear Jiashow,

We have checked our system with 0.25% Si9005 as you recommended and It is working. And we received new oscillator of our main supplier, with 0.1% deviation, and also with this one is working...., I know that the clock is very critical in the FPDlink, but I didn't know it how critical It is this deviation parameter..., we received a report from the supplier with the Deviation measured with a Modulation Measurement Device, and looked in the range, but it seems for some reason was not working...

Thanks and regards!