MAX3232: Heating problem with RS232 side connected at startup.

Andy,

Sorry to hear your problem with MAX3232. Can you give more information about your power design? Do you have two supplies for MAX3232 and RS232 device? Does the supply have any current limit? When you observed 400mA, was the total power from both MAX3232 and RS232 device? When the current was 400mA, what's the voltage level of the supply? As you power up the system with MAX3232 and RS232 device connected, what's the status of pin 7,8,9,10?

Regards,

Hao

Hello Hao:

We have one power supply for MAX3232 and a Atmega MCU (Linear regulator) we also tryied to power with our laboratory power supply to discard power supply problems. RS232 device is powered with another power supply with common grounds between RS232 device, MAX3232 and MCU. We tested with different RS232 devices.

If we power MAX3232 and MCU with RS232 connected current reach 460 mA at 5V (MCU and MAX3232) and doesn't work fine.

If we power MAX3232 and MCU with RS232 disconnected, current reach about 45 mA, then, after power upMCU+MAX3232 we connect RS232 device and current is maintained in 45 mA (And works fine)

We also test with MAX3232 to MCU TTL connection disconected with same result.

Best regards.

Andy,

I assume pin 10 (DIN2) and pin 9 (ROUT2) are connected to the MCU. What's the load of pin 7 (DOUT2) and pin 8 (RIN2)? When the power consumption is 460mA, what's the states of these 4 pins? Can you also check V+ (pin 2) and V- (pin 6) for their the voltage level? In normal operation, V+ is about 5.4V and V- is -5.4V.

Regards,
Hao

Hao:

Yes, pin 9 and 10 are connected to MCU(TTL side), and 7 and 8 (RS232 side) to a RS232 laptot computer port 

When power consumption is 460 mA i attach voltage levels of each pin:

pin2          4.1v

pin6          2.5v

pin7          2.9v

pin8          2.2v

pin9          3.6v

pin10        4.2v

When system works fine and power consumption is 50 mA (MCU+MAX3232) in stand-by mode i attach voltage levels of each pin:

pin2          9.5v

pin6          -9.1v

pin7          -8.5v

pin8          -8.5v

pin9          5v

pin10        5v

Best regards.

Andy,
Thanks for your measurement. Can you advise if MCU side of pin9 and pin10 both have pull up resistor? If so, they're internal or external? If you don't mind, can you point me the datasheet of the MCU? It looks like when MAX3232 is hooked up with the MCU, it creates some kind of positive feedback loop that pulls lots of power and prevents MAX3232 output from reaching the right voltage level. You can do some experiments to further pin down the problem. 1) power up MCU+MAX3232 with pin7 and pin8 not connected to each other, or leaving them float; 2) do the same test as 1) with only pin 9 connected to MCU; 3) do the same test as 1) with only pin 10 connected to MCU.
Regards,
Hao

Hao:

We test with TTL side disconnected (pin 9 and 10), then power MAX3232 and MCU: Doesn't work, power consumption is 450 mA

We test with all connections except RS232 IN connected (pin8): power consumption is normal, about 50mA (MAX3232+MCU)

Regards.

Andy,

Do you have any decoupling cap from Vcc to ground, how many, where and what the values are?

What's the RIN connection? Is it pulled high or low? By default, that pin is biased to ground internally. Can you try with setting RIN high? From your description, the abnormal power might not come from the interaction between MCU and MAX3232 (at least from TTL side).  

Do you have any schematics? If so, it would be helpful. I draw a simplified diagram.

Regards,

Hao