new board, high current consumption, Vcc pulled down

Sage Radachowsky

Other Parts Discussed in Thread: MSP430F5419, MSP430F5418, TPS60213

Hi friends,

I have a new circuit board using the MSP430F5419. I cannot program it and when I turn on the power, the Vcc gets pulls way low. The uC gets slightly warm. With power off, I measure resistance from ground to Vcore, and I see 1.2 Ohms. Why is that??? Also, the resistance from Vcc to Ground is also very low, under 1 Ohm.

I have similar board using the MSP430F5418. On that one, the resistance from ground Vcore is very high, 15 MegaOhms.

I have two identical boards with the same problem. I have checked all the solder joints with a microscope. I have double checked the pinout. I have checked an identical board with no components on it, and the Vcc and Ground are not shorted on the board. It is something in the connection to the uC.

Any leads? Any specific problems with MSP430F5419?

All that I have on the circuit board is the 3.3V regulator for Vcc, the uC, a 32KHz crystal on XT1 and XT2, the reset pull-up resistor & capacitor, and a low-power voltage reference connected to Vref+. I have not programmed the uC yet. It is off-the shelf.

As I said, I have two identical boards with identical problem. I have triple-checked the pinout of all ground and Vcc connections. Vcore is connected only to a capacitor to ground. I have examined all solder joints with a microscope to check for bridging or anything wrong, and it's all picture perfect. But still I have this problem.

Please, any suggestions and help on how to diagnose this.

over 15 years ago

0 Sage Radachowsky over 15 years ago

Prodigy 180 points

By the way, I am using Rev L of the chips.

0 memoryleak over 15 years ago

Intellectual 345 points

Given that the chip has already gotten warm, it's likely already been destroyed and you'll have to start with a fresh one :( But don't solder it on just yet!

-----------------------

A quick aside on the "ohms" function of multimeters: it's usually accomplished by forcing a voltage or current through whatever you're probing and calculating the resistance by measuring the resulting current/voltage (respectively). Most multimeters use different excitation signals for different resistance ranges, and a multimeter may present something pretty damaging to active circuits (such as an MCU). I have a cheapie at my desk that likes to use 5V and measure the current flowing, and I have an expensive one that sends somewhere between a few nanoamps and a few milliamps (depending on the resistance range) and measures the resulting voltage, and I think it has a +/- 12V supply backing up its current source. Needless to say, my expensive multimeter *will* force 1mA of current into a low power MCU, by simply cranking up its voltage until it gets to 1mA. (possibly damaging or destroying the MCU in the process).

Morale: Ohms mode on multimeters is for resistors and traces

-----------------------

Now on to why your MCU is unhappy. Good call checking all the solder joints first, and then checking for Vcc/Gnd shorts, that was going to be what I suggested first :) After you remove your likely-burned MCU, I'd start by turning the design on (if it's safe to do so without the MCU in place) and probing all the vacant MCU pads looking for voltages that aren't expected (higher than 3.6V, lower than ground, or hanging out awkwardly between VCC and ground). 9 times out of 10 for me it's a misbehaving powersupply, but even IO can contribute to blowing your MCU.

Speaking of which I just fried a massive FPGA through one of its IOs with a slipped scope probe and a nearby 12V rail =\ I feel your pain.

I hope any of that helps, good luck!

0 Sage Radachowsky over 15 years ago in reply to memoryleak

Prodigy 180 points

Hi, memoryleak -

Thanks for all the ideas and thoughts. :) I appreciate the time you took to write.

I'm using a good Fluke 115 multimeter and I've tested it in the past and seen only ever about 0.5V from it while Ohming. Also, this problem happened before I Ohmed it.

My power supply is a good steady 3.3V charge pump regulator (TPS60213) that I've used in dozens of MSP430-based boards in the past... and the same problem is showing up on both boards I have built up. I don't have any other rails, either... nothing higher than 3.3V on the board.

I've seen MSP430 chips get warm in the past and survive with no signs of damage, so I'm not ready to throw in the towel on the chip yet.

Something tells me it's a bug on the board but for the life of me, I've spent 8 hours now and I can't figure it out!! I've had *plenty* of MSP430s working fine in the past!

Any more ideas appreciated again...

Sage