Dealing with LDO overheating for Arduino and LCD board

[dferyance]

I have a few different Arduino knockoff boards that I am trying to interface together. I have a SainSmart Mega2560 with an ElecFreaks LCD TFT01 shield with their matching 7in TFT LCD. When do I do this, I am able to get the LCD to work but after a short period of time it goes into a reset cycle. The voltage regulator is very hot (to the point of even smelling hot) so my guess is thermal shutdown.

From looking at the board, it is using an AMS1117 1A LDO regulator. It is rated for up to 15v and I am using 7.5 so I expect that should be fine.

While trying to troubleshoot it, I connected my multimeter along the 7.5v wall-wart that I am using to power it. If I measure it at the 10A range, I get a reading of about 0.6 amps. But when I read it that way, the voltage regulator only gets warm rather than overheating. Could it be that the impedance of measuring the current prevents that from happening? I tried it on the milliamp range and my meter beeps at me for going over range.

I know this combination of boards is a little odd from a power perspective. Sending in 7.5 which is regulated down to 5v which is sent into a boost converter to 12v for the LED backlight.

The boards I am using are designed to work together so I suspect that something is wrong with one of the components. Anyone have any ideas on what is going on and how I can deal with it, or maybe other things to troubleshoot?

[mariush]

(7.5v - 5v) x 0.6A = 2.5 x 0.6  = 1.5w

Either your arduino+lcd is drawing more than 1A making the regulator go into internal limits, or you're actually making the regulator dissipate more energy than the pcb can safely dissipate.

The regulator is advertised as capable of up to 1A, but it doesn't mean it can do 1A in any condition. It can only do 1A if it's cooled properly and the input voltage is at least 5v plus the dropout voltage (which is maximum 1.25v at 1A, lower at lower currents). If it can't cool itself through the pcb then it may run intermittently. 

The TO-252/DPAK packages should handle 1-2 watts with just the pcb radiating the heat, if the regulator is another package (sot-223, soic etc) then it shouldn't be able to do more than 1w for a very long time.

Note that some wallwarts that say 7.5v don't really output 7.5v at any current... when the current drawn is low the adapter may output more than 7.5v which again could cause a lot of power dissipated in the regulator.

[cosmicray]

From looking at the board, it is using an AMS1117 1A LDO regulator. It is rated for up to 15v and I am using 7.5 so I expect that should be fine.

One observation ... looking at the sainsmart product page shows a picture of the board, with zoomed images of 4 of the chips. One of them looks like the regulator. The board image shows what appears to be an On-Semi MC33269 in a DPAK case 369C (or at least that is the closest I can relate the chip markings to the datasheet). The DPAK has a Thermal Resistance of 6.0 °C/W. The AMS1117 is either a SOT-223 or a TO-252. The datasheet for that part says  Thermal Resistance (all packages) of 15 °C/W. It sounds like they switched parts (for whatever reason) and the AMS part may be somewhat less able to shed heat.

The sainsmart product page actually links back to the arduino reference design schematic, which lists the MC33269 part.

http://www.sainsmart.com/sainsmart-mega2560-development-board-for-arduino.html

[dferyance]

The replies have been helpful. I figured there was no way it should be drawing 1 amp so something must be faulty. However even when it isn't drawing an amp, it still may have to put out a lot of energy as heat. I have a few 5v switch-mode power supply modules so I am trying those and skipping the linear regulator. So far this seems to be working pretty well. I just started playing with this board so I may end up using PWM to control the backlight brightness but it is important to be stable with full-on.