Reset Controller IC really required for LPC1114

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: 

Reset Controller IC really required for LPC1114

2,443件の閲覧回数
lpcware
NXP Employee
NXP Employee
Content originally posted in LPCWare by celephicus on Tue Apr 03 19:24:23 MST 2012
Does the LPC1114 require a separate reset controller IC? The datasheet seems to suggest that one is not required. or has anyone experienced situations where the processor locked up and required a hard reset.
0 件の賞賛
返信
5 返答(返信)

2,176件の閲覧回数
lpcware
NXP Employee
NXP Employee
Content originally posted in LPCWare by ukmeterman on Fri Apr 06 09:29:44 MST 2012
Hi,

A brown out is when the AC supply voltage is low and out of spec. Tungsten old school light bulbs run orrange and lit objects look a bit brown.

As regards the LPC1114 do look up the data sheet for the BOD interupt and reset levels. Go and have a chat with your hardware guy and ask him for the scope plots of the supply rails when the supply is turned off. Set the levels of the interupt and reset to levels that are above the hardware reset level. NB there is a bit of an undifined level with the LPC1114 at about 0.7V from memory, if the supply gets quite low then it is best to have some kind of circuit that will properly pull the reset to 0V.
0 件の賞賛
返信

2,176件の閲覧回数
lpcware
NXP Employee
NXP Employee
Content originally posted in LPCWare by Rob65 on Wed Apr 04 22:23:43 MST 2012

Quote: celephicus
The hardware designer is on to me to use the error output from the voltage regulator to reset the chip, and this is a good reason to say no.


Good thing he is 'on to you' I have seen too many cases of designs were things just work by accident :eek: But as explained, there is a similar circuit inside the lpc1114 to prevent the chip getting out of reset before the power is OK.

I was hoping to find some figures in the datasheet telling me when reset becomes inactive. E.g. what the minimum voltage is that must be reached and how long after this reset becomes inactive.
I do remember having seen some diagram showing that the reset was only de-asserted after a certain number of clock pulses from the internal oscillator have been received but it's not in the datasheet. The only thing I could find is that the ramp-up of the power should be max. 500 ms.
[B]
Maybe NXP can help us here: what are the specs of the internal (power on) reset generation ???[/B]

What is important however is to still add an external pull-up on the reset pin when you are using deep power down mode; To prevent the pin from becoming floating, you need to have an external pull-up, so do install this to be 'future proof'.


Quote:

I've never heard of the two transistors in an output state turning on at the same time, but I can well believe it.

Have a look at this picture:
[IMG]https://encrypted-tbn0.google.com/images?q=tbn:ANd9GcQ4kfVqINS4LnkSREPp17EwI4ChF7rYTTy5xw1qzEiSqG2CR...
This is not an lpc111x output stage, but a general totem pole configuration that is used a lot in different ICs. Q3 starts to conduct current when the gate voltage becomes at a certain level (a bit higher than 0V) and Q2 starts when the gate voltage gets below a certain level (a bit below Vdd).
The idea is that Q1 makes sure that either Q2 or Q3 is opening its channel, but when there are strange voltages on the input Q2 and Q3 can be both active at the same time. This could especially happen in old MOSFET circuits (CD4000 series and other) when the power supply was increased slowly from 0 to 5V. Chips would heat up and the plastic might even tan a bit (but I'm not sure that's where the 'brown' in BOD comes from).

The LPC microcontrollers contain extra circuitry to detect this and prevent this from happening. The BOD detection in the lpc controller signals your application when things could go wrong so you can disable all external circuits before there is real smoke.

Regards,
[INDENT]Rob
[/INDENT]
0 件の賞賛
返信

2,176件の閲覧回数
lpcware
NXP Employee
NXP Employee
Content originally posted in LPCWare by celephicus on Wed Apr 04 20:58:58 MST 2012

Quote: Rob65
The fact that the datasheet mentions it has a power on reset seems a bit more than just suggest this ...

I'm not sure about your level of hardware knowledge but BOD (BrownOut Detection) is there to prevent your chip from overheating due to power problems.
If circuits with push-pull (totem pole) outputs are supplied with an improper voltage, there may be a certain stage where both transistors in the push-pull are in an 'open' state, resulting in possible high currents flowing through them and resulting in a defective circuit.

So there is absolutely no [I]need[/I] to use an external reset circuit. I would even state that you should not use an external reset circuit to make the most use out of the lpc1114's internal reset and BOD features - unless of course you have some other circuits that need an external reset that you cannot control from the lpc1114 that is.

Regards,[INDENT]Rob
[/INDENT]



I've never heard of the two transistors in an output state turning on at the same time, but I can well believe it. I've certainly had micros behave unpredictably at low volts and do things like writing random data to battery backed up RAM. The old 8 bit AVR's used to regularly clear the first byte of their internal EEPROM at shutdown.

Thank you for the advice that an external reset will prevent the internal systems doing their stuff, very well put. The hardware designer is on to me to use the error output from the voltage regulator to reset the chip, and this is a good reason to say no.
0 件の賞賛
返信

2,176件の閲覧回数
lpcware
NXP Employee
NXP Employee
Content originally posted in LPCWare by Rob65 on Tue Apr 03 22:31:05 MST 2012

Quote: celephicus
The datasheet seems to suggest that one is not required.



The fact that the datasheet mentions it has a power on reset seems a bit more than just suggest this ...

I'm not sure about your level of hardware knowledge but BOD (BrownOut Detection) is there to prevent your chip from overheating due to power problems.
If circuits with push-pull (totem pole) outputs are supplied with an improper voltage, there may be a certain stage where both transistors in the push-pull are in an 'open' state, resulting in possible high currents flowing through them and resulting in a defective circuit.
To prevent this and to make sure that the CPU starts properly, some micro controllers need an external reset generator to prevent the chip from becoming active when the power is not within a certain range.

The lpc1114 has its own power on reset and BOD detection circuits. During power on, the internal reset signal will be pulled low until after a certain delay period after the power is at a proper level. The BOD circuit will detect a low voltage and send an interrupt to the CPU so it can do some critical stuff before the power becomes even lower.

So there is absolutely no [I]need[/I] to use an external reset circuit. I would even state that you should not use an external reset circuit to make the most use out of the lpc1114's internal reset and BOD features - unless of course you have some other circuits that need an external reset that you cannot control from the lpc1114 that is.

Regards,
[INDENT]Rob
[/INDENT]
0 件の賞賛
返信

2,176件の閲覧回数
lpcware
NXP Employee
NXP Employee
Content originally posted in LPCWare by js-nxp on Tue Apr 03 19:51:52 MST 2012
Any processor with reliable Brown Out Detector circuit built in should not require an external one. Of course it is up to you to make sure that the BOD is enabled and at the correct level.
0 件の賞賛
返信