Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

help needed on MCU "latch up"

Status
Not open for further replies.
C

cdcll

Full Member level 3
Joined
Feb 15, 2002
Messages
163
Helped
8
Reputation
16
Reaction score
1
Trophy points
1,298
Activity points
1,199
what do you mean by latch-up ?

after lightening, microcontroller did not work, and it is very hot, can not be reset, the only way to recover the system is power off and then turn on the power.

the MCU is SM5964, is it "latch up"?

anybody can tell me how to solve this problem?

Thanks a lot!
 

causes of microcontroller latch-up

hi cdc

it could be a latchup, but it depends on what's connected to it. If it's I/O is connected to a couple of loads that source/sink a high current and the firmware froze with those I/O active, it could lead to an overheating.

However, you must make sure that the MCU can withstand the environmental conditions it will be in and that it will reset properly if it loses its power for a few seconds.

Hope this helps,

cyberblak
 

capacitor couple mcu wake up

MCU latch up can occur due to EMI, RFI and noise. If the MCU has a WDT, it is better to enable it so that when a latch up occurs, the WDT will reset the uC and the uC can start executing the code again.

Cheers

Ravi
 

zener latchup

Hi, This is clearly a latch-up situation in which all CMOS ports inside your MCU are conducting ant that is why the unit is so hot.

If you have long cables connected to I/O pins use small resistors between your I/O lines (1 - 10K) and your external device (only on input pins !). You can also connect a diode to power supply rail or a zener diode of 5.1 V to ground from I/O pin (input).

If your reset line of your MCU is floating make sure it is connected to a logical high or low level whatever is needed.

This will solve your problems.

Paul.
 

microcontroller latch up

cyberblak said:
hi cdc

it could be a latchup, but it depends on what's connected to it. If it's I/O is connected to a couple of loads that source/sink a high current and the firmware froze with those I/O active, it could lead to an overheating.

However, you must make sure that the MCU can withstand the environmental conditions it will be in and that it will reset properly if it loses its power for a few seconds.

Hope this helps,

cyberblak
Click to expand...

Thank you!

we did the surge test in the lab, and have gotten the approval. when the MCU hangs, it can not be reset manually, but it becomes normal after turning off the power and power on again.

I do not know how it happens, is there something wrong with the circuit when MCU reset?

Added after 7 minutes:

ravimarcus said:
MCU latch up can occur due to EMI, RFI and noise. If the MCU has a WDT, it is better to enable it so that when a latch up occurs, the WDT will reset the uC and the uC can start executing the code again.

Cheers

Ravi
Click to expand...

thank you!

the WDT is already enabled in the code,.

we found that the reset does not work when the MCU is latch-up, even manually reset can not take effect. so the WDT may not work in this case, how can we do something with the circuit to prevent the latch-up happens?

Added after 1 hours:

PaulHolland said:
Hi, This is clearly a latch-up situation in which all CMOS ports inside your MCU are conducting ant that is why the unit is so hot.

If you have long cables connected to I/O pins use small resistors between your I/O lines (1 - 10K) and your external device (only on input pins !). You can also connect a diode to power supply rail or a zener diode of 5.1 V to ground from I/O pin (input).

If your reset line of your MCU is floating make sure it is connected to a logical high or low level whatever is needed.

This will solve your problems.

Paul.
Click to expand...

Hi Paul,

thank you so much for your suggestion.

there are so many I/O pins for the MCU, do you mean I need to serial a small resistor and connect a diode to power supply rail or a zener diode of 5.1 V to ground for every input pin? and why it is no need for the output pins?

regards,

cdcll
 

Hi,

WDT or other reset systems inside the MCU are not working anymore in a Latch-up situation. Only use this network on I/O pins that have long wire's or are high impedance. Also place a capacitor of 100nF accross power supply rail close to your PIC.

Normally input pins are more sensitive but output pens can also be affected. MCLR pin on the PIC is the most sensitive of all I/O pins. Use a direct connection or through a 10K resistor to VCC on the PIC.

regards,

Paul.
 

thank you, Paul!

I am using SM5964, not PIC microcontroller, do you know which pins are the most sensitive?

regards,
 

Hi, Looked at the datasheet but did not find any pin I know that would be more sensitive than others. Do you use an XTAL or an OSC ???.. or maybe external clocking ???.. IF so than the left-open OSC pin could give problems.

regards,

Paul.
 

Thank you, Paul!

I am using an XTAL with 2 capacitors.

could you tell me how to confirm which pin causes the latch-up when it happens? is there a way to detect it?

thanks & regards,
 

Hi, Post your complete drawing and I will have a look.

Paul.
 

cdcll - use external watchdog connected to mcu reset line or to power control switch (can be done on mosfet bjt ).
 

Measure Vdd during the latch-up condition.
The MCU is consuming lots of current during the latch-up (you said it was hot) so
possibly your power supply has lowered the voltage below the value that the
MCU needs to make a proper reset.

There was a similar question some time ago, that I answered, but there was no reply:


So let us know how it works out for you.
 

Hello People,

A True Latch-up can't be stopped by pulling the RESET line LOW !!!!. That does not work. Please wake up. Pulling the reset line only brings some part of the CPU/MCU to rest but the current will keep flowing and the device is locked completely.

so. Power down is the only way to get out this situation.

Paul.
 

I have the same problem in my circuit.

I have a lm7805 that powers up a PIC. this PIC controls 2 12V relays by 2 npn transistors. one to power up dc motor and the second for motor polarity inversion. sometimes after relay switching lm7805 becomes hot and output voltage drops down!
I tried filtering power source by varistor and bidirectional Transient suppressor as P6KE18CA parallel with 12V motor and relay solenoid.

but i can't solve this problem!

do you think a 5v1 zener on pic output can help me??

could it be a problem if i use the same ground for mcu and relays??

is there a technique to separet ground?? my ground is made of simple lines and hasn't a ground net.
 

yes for sure ground to switching transistor's emitter has to be taken from power connector to your board and from same place but using separate wire or plane it should go to the PIC's ground . Put also choke together with low ESR electrolythic ceramic 1 uF and 100 pF capacitors to PIC's power.

Just one question - did you properly identified your problem as glitch from power line to PIC which make
PIC hung ? Or do you supply your motor from same source ? If so - you must put chokes to motor supply together with capacitors as well .
 

Dear all,

I have modified the PCB layout, it seems better than before, but still got the latch-up.

Then we short the signal ground and the earth, everything seems OK in the lab testing!

could you tell me why? and can I do like this?

BR,
cdcll
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top