Switching regulator getting hot

[jdhar]

I have a plug-in power module by Artesyn, the PTH12050. It is supposed to regulate from 12V input to 5V. This module is getting quite hot, even though it's regulating fine. I am drawing very little current from it. Coudl this be the reason why? I have schematics located here:

**broken link removed**

Check page 3 of the schematics for the regulator connection.

Any suggestions?

 

[out to lunch]

are you sure that the load is small - have you tried to verify this? Perhaps there is some low impedance fault somewhere that is loading the module down - not enough to cause an overcurrent condition but enough to run it hot.

I would suggest trying to isolate the module on the board such that most or all of the load on the 5V rail is removed - perhaps the easiest way to do this would be to remove the module from the board and then power it up from 12V (you can air-wire in resistors, etc to set the output to 5V). If the module is getting hot in this condition, then there is something wrong with it and you should return it. If the module is regulating without getting warm in this condition, then you have more load on your board than you know about and you need to re-assess your 5V loading requirements.

If you have a bench power supply that can put out 5V, you can connect that to your board with the Atresyn module removed. Tie a low valued power resistor in line with the 5V from the power supply and measure the drop to determine the load current.

 

[jdhar]

are you sure that the load is small - have you tried to verify this? Perhaps there is some low impedance fault somewhere that is loading the module down - not enough to cause an overcurrent condition but enough to run it hot.

I would suggest trying to isolate the module on the board such that most or all of the load on the 5V rail is removed - perhaps the easiest way to do this would be to remove the module from the board and then power it up from 12V (you can air-wire in resistors, etc to set the output to 5V). If the module is getting hot in this condition, then there is something wrong with it and you should return it. If the module is regulating without getting warm in this condition, then you have more load on your board than you know about and you need to re-assess your 5V loading requirements.

If you have a bench power supply that can put out 5V, you can connect that to your board with the Atresyn module removed. Tie a low valued power resistor in line with the 5V from the power supply and measure the drop to determine the load current.

Thanks for the reply. Yes, I am sure the load is small. I did this by doing the very same thing you suggested - when I started populating the board, the first thing I did was to put the power components on, and nothing else. This is when I checked it, and it was very hot. However, I failed to mention something else. I have used this regulator before, and it also got hot on the last board. I thought this was because I temporarily shorted out one of the pins, but now that it's happening again, with a brand new power module, I am starting to wonder why. As for my 5V loading requirements, a PIC and 3 LED's hardly adds up to the 1 or 2A that this module can put out I was just wondering if too-little load can cause this, since I thought I remembered hearing once that switchers work better with higher loads.

 

[out to lunch]

I still suggest that you remove the module and apply a 5V supply with an inline resistor/shunt so the load current can be measured. There may be an error in the board layout, there may be poor solder connections, there may be a number of different reasons why what you would expect to be a small load is actually rather large. Once you measure the load current and verify that it is truly very low, then you can turn your attention to the module. The best way to troubleshoot something is to isolate it and verify what you expect is what you are getting.

Switching regulators don't necessarily work better or worse with higher loads - but they are more efficient with load applied than without any load or a very light load. This has to do with the power required just to run the regulator (losses due to just turning the FETs on or off are the biggest culprit) in relation to the amount of load being applied.

 

[jdhar]

Unfortunately, it's a 5-pin through hole component, with very large pins, so removing it isn't an option. And also being $20, I only have the one on the board. With only the module and caps placed on the board, the only way I could draw more current and heat it up is if there was a low-impedance path from the output to ground, correct? I have measured both before and after I put the module in (I always check for shorts on a new board), and it's open circuit both times. Also, it's nearly impossible to mess up a solder-job on a through hole component that's spaced as large as this one. I know it's not the best way to troubleshoot, ideally I could have the module removed, but I can't do that in this situation. Any other suggestions?

 

[House_Cat]

There is a TI spec sheet for the same module you are using. That data sheet (on page 4) shows a power dissipation curve at various loads for your module. Your module apparently dissipates about one watt at zero output current. With no heat sink or ventilation, that would be sufficient to give you about 40-50deg C temperature rise.

I suspect what you are seeing is normal behavior for the module you are using. You may want to give it a small heatsink or ventilation or you may have problems when you load it heavily.

The more detailed TI datasheet is at: https://focus.ti.com/lit/ds/symlink/pth12050l.pdf

 

[jdhar]

Damn I wish topic notification worked all the time... thanks for the reply House_cat. I was just surprised to see that much of a temp. rise for a switcher I guess. It gets hot to the point where it heats the board up (it's a smaller board - 2x3"), and I can't leave my fingero n the regulator. Ventilation will be ag ood optino, thanks aagain