VDDIO of i.MX28 steps up and down

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: 

VDDIO of i.MX28 steps up and down

ソリューションへジャンプ
2,515件の閲覧回数
hikaruuruno
Contributor III

Dear Community,

We noticed that VDDIO of i.MX28 steps up and down during bootlet.

When 5V is turned on to supply VDD5V, VDDIO becomes 3.1V which is reset value.

Then somehow VDDIO steps up to ca. 3.4V even though the target value is set to 3300mV. (ddi_power_SetVddio(3300, 3150);)

After that, VDDIO steps down to 3.3V for a brief moment and then steps up again.

When bootlet is completed and Kernel is loaded, then VDDIO finally settles to 3.3V.

The following image is the behavior of the EVK (Rev.D).

{"tokenName":"rte.image.token","tokenValue":"1425943492796-HKSZEDFANYWEX5C38NJXIHAERZSM9MN9","images":[{"id":20623,"name":"VDDIO_3V3from5V.png","url":"20623_20623.png"}]}

(Power on -> 3.1V -> 3.4V -> 3.3V(instantaneous) -> 3.4V -> 3.3V(stable))

When we changed the target value to 3150mV (ddi_power_SetVddio(3150, 3000);), VDDIO still became 3250mV.

We reviewed the source codes of the bootlet but could not find a part which is causing the behavior.

Does anyone know what is causing the 100mV offset during bootlet and how to prevent it?

We are aware that the output accuracy is 3%. But since VDDIO becomes 3.3V in the end, we don't like the 100mV step-up behavior.

For your reference, we also find the following behaviors.

1. When VDD5V is not supplied and instead 3.3V is supplied to DCDC_BATT on the EVK, VDDIO acted similar way (initial voltage was not 3.1V but 3.15V.)

2. Our custom board, which is VDD5V source only configuration, acted same way as the EVK.

Thank you in advance,

Hikaru

ラベル(1)
タグ(1)
1 解決策
2,330件の閲覧回数
hikaruuruno
Contributor III

Hi Igor,

I got it. Thank you,

(Though it is quite possible that nobody refers to the thread...)

For later reference;

i.MX28 sets its VDDIO 100mV higher than the target voltage set by ddi_power_SetVddio during the bootlet.

The behavior is temporal, and VDDIO will be settled at the target voltage.

There is no known method to prevent the behavior.

Best regards,

Hikaru

元の投稿で解決策を見る

0 件の賞賛
返信
5 返答(返信)
2,330件の閲覧回数
igorpadykov
NXP Employee
NXP Employee

Hi Hikaru

you are right: output accuracy is 3%, actually this also

includes accuracy tolerance during short time transients (as you are observing)

and this behaviour is within datasheet specification.

Best regards

igor

-----------------------------------------------------------------------------------------------------------------------

Note: If this post answers your question, please click the Correct Answer button. Thank you!

-----------------------------------------------------------------------------------------------------------------------

0 件の賞賛
返信
2,330件の閲覧回数
hikaruuruno
Contributor III

Hi Igor,

Thank you for the reply.

Since the offset is 100mV even if the target voltage is less than 3.3V, I'd say the accuracy is not really within 3%...

And I don't think your reply does not answer my questions, but I suppose you mean the temporal 100mV offset is not a weird behavior.

So I guess there is no known method to prevent it, right?

Best regards,

Hikaru

0 件の賞賛
返信
2,330件の閲覧回数
igorpadykov
NXP Employee
NXP Employee

Hi Hikaru

I am not aware of such methods.

Best regards

igor

2,331件の閲覧回数
hikaruuruno
Contributor III

Hi Igor,

I got it. Thank you,

(Though it is quite possible that nobody refers to the thread...)

For later reference;

i.MX28 sets its VDDIO 100mV higher than the target voltage set by ddi_power_SetVddio during the bootlet.

The behavior is temporal, and VDDIO will be settled at the target voltage.

There is no known method to prevent the behavior.

Best regards,

Hikaru

0 件の賞賛
返信
2,330件の閲覧回数
gonfer
Contributor V

Hi Hikaru,

your findings are really interesting.

Thanks for sharing !!!

Gonzalo.

0 件の賞賛
返信