TMS320F28375S: Errata - PLL: May Not Lock on the First Lock Attempt

Jordan Nelson

Part Number: TMS320F28375S
Other Parts Discussed in Thread: C2000WARE

We have a product that uses the TMS320F28375S and would like to understand the likelihood or probability that a device susceptible to this issue would lock after 5 attempts. It is mentioned in the referenced post that the workaround "has had an exceptional success rate" after up to 5 reattempts. We are using the latest version of the workaround from C2000ware.

Can this success rate be put in terms of percentage, for example there is a 99.9% change that a device within the general population would lock within 5 attempts? If not, any data on this topic that can be shared would be greatly helpful for us to understand the probability the device would not exit the reattempt loop within ~25 reattempts.

over 3 years ago

0 Hareesh Janakiraman over 3 years ago

TI__Guru* 95315 points

Jordan,

I am referencing 3 posts that discuss this. Bottomline: The rate of occurrence is so low that it is hard to quantify.

https://e2e.ti.com/support/microcontrollers/c2000-microcontrollers-group/c2000/f/c2000-microcontrollers-forum/632936/tms320f28377d-about-errata-pll-may-not-lock-on-the-first-lock-attempt

"The workaround is effective and there are not any reports of failures in systems when the workaround is implemented. In applications without the workaround implemented, the number of reported field failures is very small and it is difficult to quantify occurrence rate with such a low number of occurrences".

https://e2e.ti.com/support/microcontrollers/c2000-microcontrollers-group/c2000/f/c2000-microcontrollers-forum/963741/tms320f280049-pll-errata-pll-may-not-lock-on-the-first-lock-attempt-question

"As for the PLL locking errata, this happens only on a very low volume of chips if the PLL errata workaround is not implemented. If the PLL errata workaround (lock several times + use DCC to check the SYSCLK frequency) is used the number of units not locking would be further reduced (very very low PPM)".

https://e2e.ti.com/support/microcontrollers/c2000-microcontrollers-group/c2000/f/c2000-microcontrollers-forum/948732/tms320f280048-pll-may-not-lock-on-the-first-lock-attempt-errata

"The occurrence rate is very low, so much so that it is not really quantifiable."

0 Jordan Nelson over 3 years ago in reply to Hareesh Janakiraman

Prodigy 250 points

Hey Hareesh,

In the second post which you reference it discusses the F28002x and F28004x having different PLL designs. Based on this I was not able to determine if the mention of "very very low PPM" was applicable to the F2837x chips. I was going to use this as my answer and tie it to the peripheral reference guide, but the C2000 Peripheral Reference Guide (spru566q) does not list the PLLs as a peripheral.

Is there anything more concrete that can be given?

Thanks,

-Jordan

+1 Hareesh Janakiraman over 3 years ago in reply to Jordan Nelson

TI__Guru* 95315 points

Jordan,

The F28004x is impacted by this issue, as mentioned in www.ti.com/lit/SPRZ439. The issue has been fixed in F28002x device. The F2802x device uses a completely different PLL design and is not impacted by this issue.

Jordan Nelson said:
Based on this I was not able to determine if the mention of "very very low PPM" was applicable to the F2837x chips.

It is applicable.

Jordan Nelson said:
but the C2000 Peripheral Reference Guide (spru566q) does not list the PLLs as a peripheral.

Agree SPRU566 does not list PLL as a peripheral. All C2000 MCUs that operate with a 1.8v core voltage use the same PLL. MCUs that operate on 1.2v core voltage use a different PLL.

C2000™︎ microcontrollers

C2000 microcontrollers forum

TMS320F28375S: Errata - PLL: May Not Lock on the First Lock Attempt