This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

MAX232 Overheat

Other Parts Discussed in Thread: MAX232, MAX202, MAX3232

Hello.
I have some problems with the MAX232.

In some cases, when I pluga and turn on the system (with contains the max232), the IC will be inoperated and overheat too much.
In other cases, the RX will work OK, but TX not work.

I don't know wich's the cause of my troubles...

I hope help me.

Best Regards
fede

  • Hello Federico,

    This is very odd; please supply more information.

    How many MAX232 units have overheated? Are they permanently damaged?
    Does this happen when the RS232 cable is unplugged?  Do the 'system'  with MAX232 and device on other end of the cable share the same ground?  

    The MAX232 should not overheat unless it is damaged by electrical overstress or electrostatic discharge.

    Regards,
    Ron Michallick

     

  • Hello Ron!


    Some max232 are overheated.... it's very frequent . Of 10 system, one or two have this problems.


    If I turn off the system at time (when overheated), the max232 aren't permanently damaged.
    I take few minutes, turn on again, and max232 will work ok.

     

    This happen when I turn on my system with the RS232 cable IS PLUGGED.

    The system with Max232 and other device share the same ground.

     

    I checked that the DB9 Chassis are not grounded.

     

    OT: Today I replace the max232 with the TRS232ECN, but my system have some problem at start up. In my system have an RTC. When I change the Max232 for TRS232ECN, ocurred the RTC data loss. The RTC datasheet note the negative voltage glitches... but it's is another problem.



  • Federico,

    Do you have a oscilloscope to measure to MAX232 pins.

    Let me describe the expected signals.

    C1+ is square wave 5V to 9V
    C1- is square wave 0V to 5V

    C2+ is square wave 0V to 9V
    C2- is square wave -9V to 0V

    V+ is +9V
    V- is -9V

    VCC is 5V without large ripple

    DOUT pins +7V or -7V
    RIN pins unknown

    All numbers are approximate; look for differences between good and bad devices.
    Is your application close to that in data sheet figure 4?  The +/-8.5V labels are for information only, they do not connect to external circuitry.

     

    Regards,
    Ron

  • Hello Ron

    Ok, i will check it.

    My application is the same at the figure 4, but I have ommited the Cbypass 1uF.

     

     

    Do you know why my system don't work when I replace the max232 for TRS232ECN??

     

     

    Best Regards

    Federico

  • This is most interesting. Only today, I have been working on an application where I too experience the MAX232 overheating along with the 7085 that is providing power. Initially I observed it to be intermittent. I stripped out the capacitors one by one but still intermittent heating up when i power up the circuit. Further investigation found that if you place a short on the capacitor between pin 2 and ground for a fraction of a second and the thing takes off. I removed all the caps and it still does it. I then checked with a circuit i built several years ago and placed the short on ground and then quickly removed it. Sure enough the MAX232 takes off and heats up, drawing considerable current and making the 7805 way too hot to touch.

    I suspect that my intermittent fault may be down to a bad cap but I am now concerned that I could be sitting on a small time bomb. I haven't experimented yet but perhaps a resistor in series with the supply might limit the current if it starts to do whatever its doing??

    Dave - G4UFS

     

     

  • The plot thickens. After a little more digging and looking at other peoples circuits. I noted that all of them had the capacitor on pin 2 going to Vcc. This could be the problem with MY circuit and could be root cause of others.

    The datasheet I have been working to was the first hit on google (see below) and it clearly shows the cap going to ground :-

    http://focus.ti.com/lit/ds/symlink/max232.pdf

    Yet a further search which brings up another datasheet shows the cap on pin 2 going to Vcc :-

    http://www.chipswinner.com/ends/MAX232.pdf

     

    I will check my circuit later but perhaps it could explain the observation of others??

     

  • David,

    The C1 capacitor can be terminated to ground or VCC. The charge pump works equally well either way.

    I will test the temporary short effect in the lab later this week.

    Regards,
    Ron Michallick

  • I tested shorting the V+ pin with and with out capacitors installed. The short current was very high, over 600mA, because there is an internal diode from VCC to V+.
    There was not any current gain that would suggest a possible latchup condition. When the short was removed the current level returned to normal.

    I was not able to recreate the overheat condition.

    Regards,
    Ron Michallick

  • Ron,

    This is most interesting. I have reproduced the fault on three chips now. Just to be clear, this is pin 2 to ground. Something else i noted, when the chip heats up, removing power and then repowering has to be done after it cools otherwise it heats again

     

    Any one else have the problem?

    Regards

    Dave - G4UFS

  • Hi!

    I experience the MAX232 overheating too.

    My application is a little different that Figure 4. I have nonpolarized ceramic capacitors (1 uF) and the capacitor C3 (Vs+) is connected to VCC.

    I turn off the system at time (when overheated). The max232 aren't damaged because if a turn on again, the circuit works!

    Anyone else have the problem?

     

     

  • I too can confirm even with the cap connected to Vcc, I experience overheating.

    That said, I have also managed to combat the problem by fitting a 20R resistor in series with the power feed. The resistor was initially intended to limit the current draw if the intermittent fault manifested itself but testing after installing has show the fault to disappear completely. Even when I induce the fault by momentarily shorting the capacitor, I observe no latchup.

    Please let me know if this works for you.

     

    Regards

    Dave

    G4UFS

     

  • I have the same problem with MAX202. It occurs randomly when power is applied to the circuit. Connecting the other end of the capacitor on pin 2 to VCC or GND does not make any difference.

  • Will anybody from TI suggest a real solution? I have many boards with this problem and the MAX202 part is exactly the same as the circuit in the datasheet.

  • Kmz Tan,

    I have tried multiple times to create this effect in the lab unsuccessfully. Would you be willing to send in a couple units that are prone to overheating?

    Regards,
    Ron Michallick

     

     

  • Hi Ron,

    Unfortunately, I cannot send any of these units since they are all on the boards assembled. We used 1 uF capacitors with MAX202 based on the statement "The MAX202 requires 0.1 uF capacitors, although capacitors up to 10 uF can be used without harm." on page 7 of the datasheet. When we replaced the capacitors (all of them) except the power supply decoupling one with 100 nF in a few boards, we observed that the problem does not occur. I think the problem occurs because the charge-pump circuit sometimes does not start with 1 uF capacitors.

    Regards,

    Kmz

  • Kmz,

    Thanks for the information. What type of 1uF capacitors were used, tantalum, ceramic, or other?

    Regards,
    Ron Michallick

  • Ron,

    They are ceramic capacitors.

    Regards,

    Kmz

     

  • was the problem solved? I have the same problem. I have MAX232ID and all capacitors installed as described in datasheet. But sometimes (not very often) MAX232 overheats a lot!!! Power cycle usually helps.

    It happens when RS232 side is disconnected, but TTL part is connected to MCU.

    What can be the cause of the problem?

  • I've three Maxim MAX232CPE IC having symptoms as described by you in this thread, mainly gets hot on power-up.

    I've prototype circuit conforming to reference schematics, but TTL ports connected to uC.

    The circuit operates correctly when Ti MAX232, or ST 232BN  IC is used instead.

  • Hello,

    I have exactly the same overheating problem and I explained the problem in detail at the below link.

    http://e2e.ti.com/support/interface/etc_interface/f/392/t/233847.aspx

    David suggests connect a 20 ohm resistor to VCC and I applied this. I'm not sure but this solution seems solved the overheating problem.

  • Hi,

    I might be able to add some more clues to the this issue.

    I am having the overheating problem with MAX232EPE's. The circuit they are in has the correct 1uF capacitor orientations with no other connection errors or problems. I know this because substituting the MAX232EPE with a MAX232N works perfectly with no overheating,

    I have tried several different MAX232EPE's in the same circuit. Some work fine while no external circuit is connected but as soon as data is applied to Tin1 they overheat. Another worked fine as long as Tout1 was not connected. The circuit only has Tin1/Tout1 and Rin1/Rout1 connected and through all of the testing I have done yesterday and today I've only put data through Tin1/Tout1.

    I have a 22 ohm resistor in series with the 5V supply to pin 16 (as suggested in a previous post) but this only prevents the MAX232EPE's from heating quickly - they still overheat but more slowly.

    As i have a socket for the MAX232 I am able to swap them in and out of a known good circuit (good with the MAX232N installed that is) so I am definitely comparing apples with apples.

    It does seem as though there is something peculiar to the MAX232EPE's which is causing the problem, in my case anyway.

    I think I will just stick with using the MAX232N's until somone comes up with a solution for this.

    Cheers,

    Neil.

  • Neil,

    Is TIN2 unconnected?
    Logic inputs must always be connected to either a valid low and high signal.
    Connecting TIN2 to the ground or VCC pin is acceptable, but a 1k to 10k series resistor is preferred.

  • Hello everyone.

    I've just solved my problem with overheat for my case.

    I have a breadboard with only Atmega16 and MAX232 installed. All connections were made according to the datasheets and were triple-checked. My power supply was a usb-port. When I tried to do something with rs-232 my MAX232 didn't do anything but overheating.

    The solution was to draw power from another source (a laboratory power supply in my case). First I thought that USB does not provide sufficient power, but then I measured voltage between Vcc and GND with the oscilloscope - it was around 4.7 V. No ripple was noticed either.

    So, generally, if you don't know what else to do, try to find a way to increase the "quality" of your power supply.

    Hope this will help someone, because this post is #1 Google result for "max232 overheat" :)

    Best regards,

    Alexander. 

  • Alexander,

    Do you have both T1IN and T2IN connected to valid logic levels?

  • T1In was connected to ATMega16, T2In was not connected to anything.

    Have you tried to power your MAX232 from USB? Just interesting.

  • Hi Ron.

    I have the very same problem of overheating... The conections are checked and I´ve made the same conections that the datasheet...

    Is there any chance that if some of the signals entering in Din1 o Din2 rises before Vcc is macking some latching in the MAX3232 that I´m using, and then gets overheat?

    We have a complete reel with almost 2500 units... and I have tryed almost 20 of it....

    Is there same way solve this?

    PLEASE, LET ME KNOW...

  • Hi again Ron...

    Here I´ve got the lot written in the reel:

    LOT:5397198WCW.

    Maybe that helps a little...

  • and here´s the TOP MARKING of my MAX:

    92AY3FM

    MAX3232C G4