ATTiny88 overclocking, or is it?

Go To Last Post
7 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0


ATTiny88 CLKI (PB6) declared to be 12MHz, but works on 16MHz.

 

My ATTiny board looks like this:

 

 

There is 16MHz external clock  (the square in the middle), the board works fine on this frequency. Declared for Arduino (but you can remove it), enjoy the beauty of 26 I/O pins. Mind, the sink/source current is generally less -except PD3210- than in other Tiny or Mega.

 

Chip looks like 32 QFN, marking is :

T88 - 15MZ

1536    TH

A3KBAA

 

Problem is that datasheet specifies only 12MHz on CLKI.

I know for overclocking, I did 17MHz on M16, but here the difference is significant, how it can be, please.

 

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

grohote wrote:
... but here the difference is significant, how it can be, please.
Conservative SOA (margins)

Push Performance and Power Beyond the Data Sheet White Paper (Texas Instruments)

[second paragraph]

The result is that we need to find new ways to satisfy our continuing demand for more performance, and new ways to achieve that performance at a lower power level. One way we can do this is by understanding how performance and power dissipation depend on other variables under our control. It is also important to understand how IC manufacturers use guard bands to guarantee the performance of their products. By understanding these factors, we can uncover hidden performance.

more :

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The clock frequency limit is not an "exceed, then fail: limit.  It is the limit at which The Ones at Atmel/Microchip are willing to say that it will always operate and do everything over all temperature values. For many AVRs, the maximum frequency was limited by EEPROM behavior at temperature extremes. If your application for those devices does not use the EEPROM, then you can get away with using a higher frequency clock. Its just that The Ones will not guarantee it.

 

So, the behavior you report is no great surprise. I would simply not use it that way for a commercial product. For "play", do as you wish.

 

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I definitely will not use EEPROM. What I am afraid is that 16MHz is too commercial, and that some instructions may fail.

 

Decision: use 16MHz as is. Put some extra code to monitor the safety, restarts... change to 8MHz (and get PB6 pin) eventually.

Hope that this thread will not have a sequel.

Last Edited: Sat. May 14, 2022 - 05:02 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You will hardly ever find that "instructions" fail with such mild over-clocking. EEPROM read/write might fail at -20C. There might be ADC problems at +80C. 

 

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ka7ehk wrote:
... and do everything over all temperature values.
and SOA VCC (CMOS hysteresis)

ka7ehk wrote:
... at temperature extremes.
Another is any CMOS oscillator (transconductance); a very cold oscillator may not start or may otherwise malfunction (inadequate amplitude, frequency out of tolerance, excessive phase noise [jitter])

ka7ehk wrote:
I would simply not use it that way for a commercial product.
... whereas I would if necessary though the customer would receive a probability of failure for each estimated failure (risk = probability of failure * consequence of failure)

Some AVR have FMEDA data.

 


Abracon | 18pF Crystals May Not Oscillate with Energy Saving MCUs

[mid-page 2]

These circuits [oscillators] run the risk of failing to startup due to total capacitive loading, changes in temperature & bias levels, etc. The sustainability of oscillations can be quantified with the following ratio:

[gain margin is proportional to transconductance]

 

[page 3, end of paragraph below table]

It is widely accepted that the gain margin needs to be greater than (5) for robust oscillations, when accounting for board parasitics, part-to- part variation, operating temperature range and other unforeseen problems that could cause oscillation to cease.

Wafer fab process variation is more known after an MCU has completed characterization (significant number of sampled die over time [wafer batches], frigid and boiling)

 

Push Performance and Power Beyond the Data Sheet White Paper (Texas Instruments)

[page 1, beginning of last paragraph]

Before we get into the details, let’s look at a few examples of ways TI customers have taken advantage of hidden performance.

[200°C, defective die, mild overclock]

https://onlinedocs.microchip.com/?find=functional%20safety%20package

Functional Safety? | AVR Freaks

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ka7ehk wrote:
You will hardly ever find that "instructions" fail with such mild over-clocking.
Exceeding the VCC limits (magnitude, dv/dt) is one way to malfunction the CPU or NVM controller; another way is excessive phase noise (AVR have a maximum ΔTCLCL)

 

XMega SRAM slow turnaround? - Solved (Glitchy Power Supply). | AVR Freaks

"Dare to be naïve." - Buckminster Fuller