Any idea if there will be a re-spin of the AM1707 to fix the intermittent reset problem (per errata)?
Any idea how frequent this problem is?
This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
Any idea if there will be a re-spin of the AM1707 to fix the intermittent reset problem (per errata)?
Any idea how frequent this problem is?
Hi Gary
This problem will be fixed in an upcoming revision of the device. I believe the production schedule for that is in September '11 time frame. If you need more comprehensive schedule and information, please do contact your TI sales time/field team.
It is hard to quantify on the "frequency" , when customers reported this issue it was always in the order of 2-5 boards failing out of 200/300 total boards, so the percentage is small, and we do not expect all devices to fail. Additionaly the failure is influenced by several factors like device variability (transistor turn on threshold, leakages), and external variabiles (power on ramp, operating temperature conditions, external components etc) and is essentially caused by some of the logic bits that will get initialized in a non deterministic way.
Needless to say that we would recommend going to production with existing revision , unless you put the recommended workaround (second reset) in place.
Hope this helps.
Regards
Mukul
Very helpful. Thanks.
If it happens in Sept, that would be great for us.
I have an independent inquiry in to our TI Tech Sales rep. She is usually very good with getting accurate information.
Will report back if I get any more details from that path.
Can I get an update to the Rev C silicon? It is already the end of March 2012 and Rev C doesn't show on the AM1707 website "Sample or Buy".
I'd like to suggest that an AM1707 being able to supply a WDI signal from GPIO (per the Errata Workaround) might be difficult if you have a Linux system (that requires 30 seconds to boot).
Is the boot failure exacerbated by temperature? Do I need to test EVERY assembly at high or low temperatures? The recommendation was to go-to-market. How do I ensure a successful product?
Regards,
MikeF.
Hi Mike
The product marketing team has bee notified on your query about Rev C availability. Someone will address this in coming days.
MIKE FARBARIK said:I'd like to suggest that an AM1707 being able to supply a WDI signal from GPIO (per the Errata Workaround) might be difficult if you have a Linux system (that requires 30 seconds to boot).
Understand. The issue applies to the entire product family , which has ARM only , DSP only and ARM+DSP only variants, and AM17 users that could possibly be using RTOS's other then Linux. For DSP only variants the WDI way is still very feasible. AdditionallyF For linux users too, perhaps there are are creative ways to insert the GPIO/CLKOUT etc in ubl etc/prior to the whole Linux boot etc.
MIKE FARBARIK said:Is the boot failure exacerbated by temperature? Do I need to test EVERY assembly at high or low temperatures? The recommendation was to go-to-market. How do I ensure a successful product?
Unfortunately if you do not put one of the recommended workarounds in place, you can expect failures. The nature of the failure is such that it can be impacted by several factors, whereby you could possibly not see a failure at time zero (getting the product out of the door), but it could show up at a later time. Implementing the workaround in your existing board design or moving to rev C will be the only sure shot way to ensure a successful product w/o susceptiblity to this issue.
Regards
Mukul
Edits: Corrections to my previous post