Beware! T4x RTC power consumption

blahfoo said:
More on getting the standby (xtal RTC only) current down, and a question (anybody listening to this thread? :)
...
Click to expand...
I'm listening. It needs to be noted that the recommended batteries are 3V lithium coin cell, not a 3.6V lithium, different chemistry and voltage.
 
bicycleguy said:
I'm listening. It needs to be noted that the recommended batteries are 3V lithium coin cell, not a 3.6V lithium, different chemistry and voltage.
Click to expand...
Yep - just a thing to watch out for, and a solution. Rechargeable lithiums also go to 4V or so while charging, so that's something to regulate down for longest battery life. The big surprise is that the consumption goes UP when you turn the power on, if you supply at least close to 3.3V Batt, as the processor consumes more and the T4 board's diodes don't make a hard switch.
 
I'm listening but this is part of the discussion above my pay grade. Because of the current draw, I re-implemented an older circuit of mine for Push On - Hold Off, and am now using a separate RTC as noted earlier in this thread. I'd like to use the onOff and RTC of the Teensy, but the power draw was too much. Unless you find a solution!
 
bicycleguy said:
I'm listening. It needs to be noted that the recommended batteries are 3V lithium coin cell, not a 3.6V lithium, different chemistry and voltage.
Click to expand...
This may be the reason it won't start with that battery on?
That keeping the RTC part alive with 3V is good to help power on in some cases noted by PJRC. But that pin like others is 3.3V limited and having higher voltage may be blocking startup with 3.3V?

Try it with a 3V battery and see if same p#25? issue occurs.

The diode I suppose does some voltage drop - but 3.6V or higher alone will kill the Teensy. And higher than 3.3V supply may disorient the electrons from 3.3V supply or some technical thing.
 
I'm watching as well as I'm suffering this same issue. I power up my system once every week or four. CR2032 battery - a couple of months and I no longer have an accurate date and my data file naming goes bananas. :(
Just ordered an external RTC, but that'll drive a board spin and figuring out how to get it functioning...
If I've got 12V available, is there a super low current way to power the RTC on the Teensy so the vehicle battery could power it? Can I have the 2032 and the 12V reduced voltage in parallel (diode or'd) if I unplug the 12V, I don't have to reset the clock? Or cap that'll power it for a few hours if I disconnect the 12V?
 
I power the RTC using a super capacitor, like this one: 1.5F 3.5v cap. I use a schottky diode between the Teensy's 3.3v output and the RTC power input to charge the cap. With a Teensy 3.2 it powers the RTC for several days. I haven't experimented with a Teensy 4.1, but I expect it to last a few hours.
 
AndyMeyer said:
I'm watching as well as I'm suffering this same issue. I power up my system once every week or four. CR2032 battery - a couple of months and I no longer have an accurate date and my data file naming goes bananas. :(
Just ordered an external RTC, but that'll drive a board spin and figuring out how to get it functioning...
If I've got 12V available, is there a super low current way to power the RTC on the Teensy so the vehicle battery could power it? Can I have the 2032 and the 12V reduced voltage in parallel (diode or'd) if I unplug the 12V, I don't have to reset the clock? Or cap that'll power it for a few hours if I disconnect the 12V?
Click to expand...
From a couple of CR2032 battery data sheets, it looks like they run 3.0V or less most of their life, but for a bit at the beginning. Dropping voltage shouldn't be needed. You could measure the current consumption using a good multimeter in series with the battery. Maybe one diode drop might help, but I'm using a 1.2 AH 3.6V battery. The little .22 Ah coin cells should give you 9 months of power off time. Sure, you could use a 3.3V voltage regulator on that 12V supply and two diodes to combine with the battery for the RTC power. I'm considering a rechargable battery and charging circuit also.
 
blahfoo said:
From a couple of CR2032 battery data sheets, it looks like they run 3.0V or less most of their life, but for a bit at the beginning. Dropping voltage shouldn't be needed. You could measure the current consumption using a good multimeter in series with the battery. Maybe one diode drop might help, but I'm using a 1.2 AH 3.6V battery. The little .22 Ah coin cells should give you 9 months of power off time. Sure, you could use a 3.3V voltage regulator on that 12V supply and two diodes to combine with the battery for the RTC power. I'm considering a rechargable battery and charging circuit also.
Click to expand...
This issue is probably driving me to an external I2C RTC - RV-3028-C7 is what I'm planning. My device will be in the field where CR2032's last years... (Think similar to your computer BIOS...)

I'm assuming I can reuse D1 for D2 and only need one BAT54 and build that part of the circuit accordingly since I won't need any low power operation. I'm assuming that the 2 .22uF caps (C17 and C18) and the 10uF (C31) help with the power sequencing. Be great to get rid of the BAT54 and caps altogether, but I'm not that smart on microcontroller circuit design yet. Can I?

All this is moot if there's a fix for the RTC current drain.
 
defragster said:
This may be the reason it won't start with that battery on?
That keeping the RTC part alive with 3V is good to help power on in some cases noted by PJRC. But that pin like others is 3.3V limited and having higher voltage may be blocking startup with 3.3V?

Try it with a 3V battery and see if same p#25? issue occurs.

The diode I suppose does some voltage drop - but 3.6V or higher alone will kill the Teensy. And higher than 3.3V supply may disorient the electrons from 3.3V supply or some technical thing.
Click to expand...
Yes, the processor never sees 3.6 V, there's >0.5V through the diode on the Teensy board.
 
I stopped working on getting out of SNVS mode and settled for the 27 uA or so standby consumption. If I read the data BOOK correctly, the only way to wake up from SNVS is to use a timer. But I'd need that to be once or twice a second so as to not excessively delay power-on startup. Then first thing, a tiny bit of code puts it back to SNVS sleep if power is down. But I don't think the T4.1 will boot quickly enough to do that. This monster may have hardware support that does just that, but I don't know enough to find the way. Or perhaps wake up on pin voltage?
 
I have recently run into this problem too. What I have found is among the 5 teensy 4.0's I have, the power consumption for the rtc varies wildly. This is measured with no external components for 5 different teensy 4.0's: 2x 92uA, 84uA, 46uA, 27uA @ 3.0V. Measured before and after a blank sketch upload (aka firmware update). The firmware update nor sketch did not change the measurement. I guess this is a production/hardware issue...?
 
Are all pins either OUTPUT or INPUT_PULLUP as required to minimize pad power drain? Floating inputs can drain current.
 
You must log in or register to reply here.
Back
Top