Helen Cussen

Important changes to forums and questions

All forums and questions are now archived. To start a new conversation or read the latest updates go to forums.mbed.com.

10 years ago.

LPC11U24 logger on Li-ion battery

I am trying to set up the LPC11U24 as a logger running on a 3.6V Li-ion battery. I plan to use a 5V step-up regulator and put the power on the VIN pin.

However, during testing with a bench power supply the write to file command does not seem to work at 5V. The code is attached. The file is created but there is no text within the files. If I increase VIN to 7V then all works as I expected.

Please help.

Helen /media/uploads/HelenC/mbed_logger_lpc11u24_-1-.bin

2 Answers

Doug .
10 years ago.

Since flash operations require more current than normal operation, you may be "running out of juice" at 5V. The spec is 4.5V minimum so if your 5V supply is not supplying enough current you may be dropping down low enough that the 3.3V regulator is dropping out. Have you tried running at 5,5V to see what happens? Also you may want to try monitoring the 3.3V regulator output with a scope while doing a flash write to see if the voltage stays in spec.

Good advice, but take into account the LPC11u24 has two 3.3V regulators, and the 3.3V output is a different one from the one used internally. Measuring directly on the correct regulator is probably the easiest solution (since it has relative large pins).

posted by Erik - 31 Mar 2014
Neil Thiessen
10 years ago.

I'm generally opposed to stepping up voltages just to knock them back down again, it's terribly inefficient. I'm working on a battery powered LPC11UXX project right now, and I'm running an ultra low dropout 3.3V regulator directly off the LiPo. Everybody will tell you that they're dead at 3.0V, but in my experience, by the time they hit 3.3V they're dropping rapidly so draining them all the way to 3.0V will only gain you seconds or maybe a few minutes of run time. And terminating early is better for longevity anyway, LiPos don't like being over discharged.

There is an adverse effect over discharging Li-Polimer batteries but generally relates to to high current discharge use such as the radio controlled model market who use these at currents in excess of 60 amps simply because you can. Taking them to 3 volts per cell is the limit. Li-Ion is a better option for MCU's as generally you're not going to need this power and they are far safer in case of wiring shorts. Also there are some very good inexpensive small USB charger PCB's available from Ebay this makes them easy to charge properly. Li-Ion chargers will not correctly charge Li-Po batteries. I use a cheep mobile phone with a 3.3v LDO regulator using a KL25Z MCU (32Khz clock crystal), running a temperature data logging to SD card, I will get over six months continuous running time.

posted by Paul Staron 31 Mar 2014

For my project, I used an MCP73832 Li+ charger IC. Extremely simple, just two capacitors and a programming resistor. Sparkfun has them on breakouts as well.

posted by Neil Thiessen 31 Mar 2014

Thanks Doug and Neil for your replies.

It has been a few weeks since I looked at this and I have now worked out what the problem is and it is not the voltage.

The problem is that when I switch off the power then the most recent file is not closed properly and hence has no data. When I was testing previously I did not give it enough time to write a full log file before switching off.

If I leave the logger running for long enough then I do get populated data files and only lose the last few readings. I can live with this but would be interested to know if there is a simple way to make sure the last readings are logged?

Thanks!

posted by Helen Cussen 24 Apr 2014