ICS factory calibration accuracy (QD4)

キャンセル
次の結果を表示 
表示  限定  | 次の代わりに検索 
もしかして: 

ICS factory calibration accuracy (QD4)

1,494件の閲覧回数
P_B
Contributor I

Hi everyone!

 

If one would want to use the factory calibration/trim value, what kind of accuracy could someone except to get?

 

In the datasheet it says that the total deviation is +/- 2% over the full voltage and temperature ranges.

This is of course for a trimmed ICS. But does it include Freescale's factory trim values? Or is the deviation bigger with those values? The data sheet does not say. It only says that the "Fine trim value" should not be read. I assume that is because the accuracy from the Freescale factory is not high enough.

 

I know that programmers can do the trimming when programming the chip, but let's not talk about that. Please keep the discussion to the Freescale factory trim values.

 

I appreciate all answers! :smileyhappy:

ラベル(1)
0 件の賞賛
返信
3 返答(返信)

1,037件の閲覧回数
bigmac
Specialist III

Hello,

 

When the factory calibration is done, this will occur at a specific, but unspecified, ambient temperature and supply voltage.  These conditions may no longer exactly apply for your application, particularly with the supply voltage.  Additionally, it is possible that variations in the device mounting on the circuit board will affect stray capacitance, and therefore may also have some affect the calibration of the internal reference frequency.

 

These factors may result in the "initial accuracy" of the calibration to be somewhat outside the specified limits.  The use of the factory determined trim value will typically ensure that the baud rate for SCI serial communications will have sufficient accuracy.

 

What accuracy do you actually need to achieve for your application?  I would strongly recommend that the trim value be recalibrated when the firmware is programmed, to achieve the best initial accuracy.

 

Regards,

Mac

 

0 件の賞賛
返信

1,037件の閲覧回数
P_B
Contributor I

 


bigmac wrote:

Hello,

 

When the factory calibration is done, this will occur at a specific, but unspecified, ambient temperature and supply voltage.  These conditions may no longer exactly apply for your application, particularly with the supply voltage.  Additionally, it is possible that variations in the device mounting on the circuit board will affect stray capacitance, and therefore may also have some affect the calibration of the internal reference frequency.

 

These factors may result in the "initial accuracy" of the calibration to be somewhat outside the specified limits.  The use of the factory determined trim value will typically ensure that the baud rate for SCI serial communications will have sufficient accuracy.

 

What accuracy do you actually need to achieve for your application?  I would strongly recommend that the trim value be recalibrated when the firmware is programmed, to achieve the best initial accuracy.

 

Regards,

Mac

 


 

This question is not for any application...yet :smileyhappy:

 

I'm just comparing different microcontrollers (AVR, PIC, 78K...etc)

Most likely the MCU would be recalibrated during programming, as you suggested.

 

But I wanted to know what accuracy one can expect over the whole voltage and temperature ranges when fresh out from the Freescale factory.

 

Do I interpret you answer correct, if think you mean that the Freescale factory trim is not within the +/- 2%?

But most likely not much higher?!

0 件の賞賛
返信

1,037件の閲覧回数
bigmac
Specialist III

Hello,

 

When the device is freshly trimmed during programming, the trimming resolution should be within 0.2 percent.  However, due to the circumstances previously mentioned, the factory trim setting may possibly fall outside this limit.  The total frequency accuracy would be determined by the initial trim resolution, plus the worst case 2 percent variation due to temperature and voltage.

 

Regards,

Mac

 

0 件の賞賛
返信