STM32F4 Discovery Board

toeztuerk · ‎2013-01-30

Posted on January 30, 2013 at 12:49

Hello all,

at the moment I have some issues with the voltages on the STM32F4 Discovery Board. The two diodes D1 and D3 (BAT60JFILM) causes a voltage drop around 0.32V, so that the 5V only delivers 4,68V and not 5V. What is about the 3V Pin ?! I measure 2.95V, but I guess it must be 3.3V or ?! Secondly I want to know if I can power the the Board through the 5V Pins with a external 5V power supply and at connect the USB cable to programm/debug the board at the same time ?! I just need a Vref of 3.3V because a ADC implementation in my project. I want to sample the signal of an MEMS accelerometer.

Thank you.

best regards,

opcode

raptorhal2 · ‎2013-01-30

Posted on January 30, 2013 at 16:26

The Discovery board Vdd is 3V (approximately), not 3.3V.

The Discovery user's manual has a schematic you can use to see if the two diodes will isolate an external 5V in from the USB connector.

Other options to consider:

1. If the accelerometer won't be subject to maximum accelerations, the signal might always be less than 3V.

2. If it is a rescalable accelerometer, set it to a higher range if possible to reduce signal voltage.

3. If only the accelerometer can be powered, use the Discovery board Vdd.

For all options other than 3, correct the result for the ratio of ADC reference to accelerometer reference.

Cheers, Hal

toeztuerk · ‎2013-01-30

Posted on January 30, 2013 at 19:22

Hello Baird Hal,

thank you for your quick reply.

I am using the ADXL001 (Datasheet can be found here: http://www.analog.com/static/imported-files/data_sheets/ADXL001.pdf ). One can obtain from the datasheet, that the minimum supply voltage is 3.1V. Secondly the sensitivity values in the dataheet are only specified for 3.3V and 5V supply voltage. So far, I can power the STM32F4 Discovery board with external 5V, so that the diode D1 is not in function, and I do not have a voltage drop to 4.68V and can power up the accelerometer with the other 5V pin with true 5V. But then I have the problem, that at maximum g-level (70g) the accelerometer will have an output voltage of about 4.3 V (zero-g offset + 70g*sensitivity). A dirty trick would be to place the sensor 180 degree oppositve of the measurement direction so that I have zero-g offset - 70g*sensitivity and I would not expose the maximum analog voltage of the ADC pins. The sensor will surely fixed placed, so I can gurantee myself that measurement will be 180 degree opposite of measurement direction. But it would be good, when I can equal the ranges from the ADC with the one of the accelerometer to have full scale range. Any Idea what can I do ?!

Do the STM32F407VGT6 provide a inbuild amplification for the ADC inputs ?! I could not find anything in the datasheet about this, but maybe I have just miss it.

Thank you !

raptorhal2 · ‎2013-01-30

Posted on January 30, 2013 at 20:39

Do the STM32F407VGT6 provide a inbuild amplification for the ADC inputs ?

No.

The ADC mux has a resistance and capacitance as defined in the data sheet. The accelerometer output amplifier is very low power (100 uamp) and will have a tough time charging up the mux sample and hold capacitor.

I believe the best answer is to use an op amp between the accelerometer and ADC to rescale the accelerometer output and provide current needed for accurate conversion.

Cheers, Hal

Tesla DeLorean · ‎2013-01-30

Posted on January 30, 2013 at 20:40

You need to run both parts from 3.3V

The ADC inputs are not going to tolerate 5V, or frankly anything exceeding VDDA,VREF.

Replace the 3V regulator with a 3.3V one.

The ADC has a sample/hold, it does not have an amplifier, if you want one add it externally.

Tips, buy me a coffee, or three.. PayPal Venmo Up vote any posts that you find helpful, it shows what's working..

toeztuerk · ‎2013-01-30

Posted on January 30, 2013 at 21:58

Hello ! Thank you very much for your rapid answers.

Can you give me an example or a good start point for opamp rescale solution ?! how do I have to dimension the external parts like resistors. I am realy sorry, but I do not worked so much with opamps yet.

Thanks for your help.

raptorhal2 · ‎2013-01-30

Posted on January 30, 2013 at 23:10

Connect the accelerometer signal to two resistors in series, first a 47K ohm, then a 56K ohm, and connect the 56K to ground. Connect the output of the 47K also to the plus terminal of the op amp. Connect the op amp output to its minus terminal. Use 3V (or 3.3V if you follow Clive1's advice) for the op amp reference. You can use 5V, but startup transients may overdrive the ADC. Connect the op amp to the ADC channel input.

This will scale the accelerometer signal by a factor of 56/(56+47) = 0.5437. Pick other resistor combinations if convenient, but if they total less than 50K, more than 100 uamps will be needed. If very much larger values are used, noise pickup can be a problem.

Cheers, Hal