MAX3232: MAX232 overheat problem

This appears to me definitely a latch up issue when the device at slave end receive VDD even when the RX TX line is seeing activity from Master end. 

FYI, a 20E resistance in series with the VDD line, reduces heating of the device- but the latch up condition still recurs. 

A thread that arrived at this conclusion was inexplicably closed by TI. Could TI be kind enough to release an errata/ work around?

Hello Pawandeep,

It looks to me as if a root cause was not found when this issue came up previously, and the thread was later locked automatically (this now occurs on the e2e forum so that follow-up questions like yours are more easily visible to our engineers). I haven't heard of this issue before, but I'd like to work with you on understanding it better.

Could you please let us know what is used to provide the VCC rail for this device, its nominal output voltage, and its maximum rated output current? Would you also please see what voltage is measured at the VCC pin of the device when it is heating up?

I ask these things because my first suspicion is that this may be related to a high-current condition occuring at start-up. The charge pump of this IC will operate by using the VCC rail and switching it across a network of capacitances to generate the higher-voltage V+ and V- rails. During steady-state conditions, it is designed to switch only when these rails drop to a minimum thresholds, and the switching action serves to dump more charge on the V+/V- filter capacitors so that the voltage increases. When VCC is low, though, the charge pump may try to switch continuously in order to push the V+/V- voltages above their minimum required threshold level. This continuous switching can result in higher supply current than during normal operation.

What I am concerned may be happening here is that the power supply ramps up, reaches a point where the charge pump starts to switch continuously and sink a lot of current, but then the supply is unable to source more current and becomes stuck a lower voltage. The system remains at this point, dissipating lots of power in the RS-232 transceiver indefinitely and unable to continue the start-up process since more current cannot be provided.

If you are able to check the items I mentioned below we can see if this theory is reinforced (in which case we can further look into corrective action) or if we need to look down another path.

Regards,
Max

Dear Max

Thanks for the update. 

Our analysis of the problem was that the latch up like condition does occur at start-ups only. So in theory what you are suggesting may be right. Our Data terminal derives power from the same connector as the Rs232 connection. Is there a solution to the probem? We did add a 20E resistance in series as many forums suggested- same result.

Yes, we did solve the problem- albeit by changing the part to an alternate make where no such problem was witnessed. 

Bests

Pawandeep

Hi Pawandeep,

Thanks for the update, and I'm glad to hear you've reached a solution.

If the root cause of the issue is higher current at start-up, it seems to me like the only solution would be to use a power rail capable of fully sourcing the start-up current. That way, the supply voltage can ramp up out of the high-current region and into the normal operating region where supply currents are lower. It may be possible as well to adjust the charge pump capacitances to reduce the current required by the charge pump or reduce the frequency of the charging cycles, but we have not yet experimented with this and I would expect the efficacy of this kind of solution to vary with power regulator design.

Max