DS90UH948-Q1: internal oscillator reference frequency variation

Yoshikazu Kawasaki

Part Number: DS90UH948-Q1

Hello,

I understand UH948-Q1 uses internal test pattern generation based on AN-2198 as shown in the datasheet on page 104. The AN-2198 shows table 2 and it says below the table that "internal oscillator reference frequency variation of 140MHz(min), 200MHz(typ) and 260MHz(max) respectively". It should include the variation of PVT(Process, Voltage, Temperature). However my customer sees 139MHz even at room temperature. Is it normal? Or is the unit bad one? Does it really include PVT? Would you please tell me the actual min/typ/max?

Best Regards,

Kawasaki

over 6 years ago

0 Charley Cai over 6 years ago

TI__Genius 12866 points

Could you let me know how your customer is measuring the internal oscillator frequency?

Best Regards,
Charley Cai

0 Yoshikazu Kawasaki over 6 years ago in reply to Charley Cai

TI__Expert 3510 points

Hello Charley,

Thank you for supporting me on this. I checked how they measured the internal OSC frequency. They configured PGCDC(0x03) shown in AN-2198 as 0x03[5:0]=0x03, so it is expected to show 140MHz/3=46.667MHz(min), 200MHz/3=66.667MHz(typ) and 260MHz/3=86.667MHz(max), but it did aournd 46.5MHz. Threfore they thought the frequency was 139.5MHz(46.5MHz x 3 = 139.5MHz).

Best Regards,

Yoshikazu Kawasaki

0 Charley Cai over 6 years ago in reply to Yoshikazu Kawasaki

TI__Genius 12866 points

The 200MHz +/- 30% variation is not under PVT. It is normal seeing close to 140MHz internal clock.
If the customer need precise clock, they could use pattern generation with external clock.

Best Regards,
Charley Cai

0 Yoshikazu Kawasaki over 6 years ago in reply to Charley Cai

TI__Expert 3510 points

Hello Charley,

Thank you for your reply. Would you please tell me what the conditions of 200MHz+/-30% shown in the AN-2198 application note are then? I think such variation should include PVT in most of the documents.

How about the variation including PVT if +/-30 doesn't do PVT? Will it be +/-40% or so?

Best Regards,

Yoshikazu Kawasaki

0 Charley Cai over 6 years ago in reply to Yoshikazu Kawasaki

TI__Genius 12866 points

This variance is due to internal VCO stability. We typically only recommend using internal clock for debug purposes.

Best Regards,
Charley Cai

0 Yoshikazu Kawasaki over 6 years ago in reply to Charley Cai

TI__Expert 3510 points

Hello Charley,

Thank you for your support. I understand there is no way to make precise clock by using internal 200MHz OSC and can't be used on actual applications.

Best Regards,

Yoshikazu Kawasaki

Interface

Interface forum

DS90UH948-Q1: internal oscillator reference frequency variation