Title: RTC Battery quickly discharged
Post by: WD8BXS on 26. December 2018, 13:50:59
How long is your battery lasting in your McHF?
Mine only lasted about 2 months.
Chuck |
Title: Re:RTC Battery
Post by: DF8OE on 26. December 2018, 14:19:08
Mine is about 9 months under use and does have same voltage as brand new one.
But there is known issue with F7 processor that leads to too high current when there is no 4.4V power supplied. se take a measure of battery idle - it must be < 1uA when no 3.3V is supplied. The issue does not impact all STMs. There is a workaround (for F7/H7) which may work for F4, too.
vy 73
Andreas |
Title: Re:RTC Battery
Post by: DL8EBD on 26. December 2018, 14:24:04
Die Lösung ist sicherlich nicht der Weisheit letzter Schluss. Aber bei mir funktioniert sie auf jeden Fall zuverlässig:
Parallel zu C94 wird ein Widerstand 47K gelötet.
Jetzt bricht die Spannung an den Vcc-Pins nach dem Ausschalten bis auf 0V zusammen und aus der Batterie wird kein messbarer Strom mehr gezogen. Die Lösung hat schon den Wärme/Kälte-Test sowie eine sich in der Spannung verändernde Batterie bestanden.
vy 73
Andreas
|
|
Quelle:
https://www.amateurfunk-sulingen.de/forum/index.php?board=18;action=display;threadid=850;start=15 (https://www.amateurfunk-sulingen.de/forum/index.php?board=18;action=display;threadid=850;start=15)
in wie weit man das auf den mcHF übertragen kann weiß ich jetzt nicht.
Man müsste sich die Schaltbilder vornehmen und C94 (in der I40 UI) auf der mcHF UI identifizieren |
Title: Re:RTC Battery
Post by: WD8BXS on 26. December 2018, 22:09:57
I am using the STM32F437VIT7 Processor.
There is no equivalent to C94 on the version 6 UI Board.
The NRST pin from the processor just goes to Pin 2 on the P8 header.
Should I put a capacitor with resistor in parallel top ground from this pin?
Chuck |
Title: Re:RTC Battery
Post by: DL8EBD on 26. December 2018, 23:42:11
Yes Chuk, try a 47k to GND and measure the battery current if mcHF is switched off
|
Title: Re:RTC Battery
Post by: WD8BXS on 27. December 2018, 02:19:48
Would the dead battery cause the "Start RTC" selection in the debug menu to go away??
Chuck |
Title: Re:RTC Battery
Post by: DF8OE on 27. December 2018, 07:12:29
Hi Chuck,
leaving NRST pin floating without resistor to Vcc and capacitor to GND is a hardware design issue. It is a wonder this works - no matter if RTC mod os present or not...
Dead battery leads to no identification of RTC mod.
Please follow the link Thomas provided - use google translate. I never have had done investigations on F4 MCU but it seems that F4 is also impacted as F7/H7.
vy 73
Andreas |
Title: Re:RTC Battery
Post by: DB4PLE on 27. December 2018, 11:35:33
Hi,
re NRST: a check in the Reference Manual for STM32F4 RM0090, page 151 reveals that the NRST is internal pulled up to Vdd/Vdda aka Vcc. An external resistor is thus not necessary. There is also a schmitt trigger AND a filter internally, anything less than 20uS pulse will not trigger a reset. So in many cases, no external circuit is necessary. Having said that, long wires to pin 2 or other picked up signal in a transmitter may theoretically reset the processor and in this case a simple RC filter close to the pin may make sense.
In any case: grounding this pin will result in a permanently stopped processor. Energy efficient but not really useful state I'd say.
re RTC Start: Appears if rtc detection code thinks there is an battery connected. The way we do this is not the most reliable way, and it often appears to misdetect the presence. Why: We check if the backup ram remains powered between reboots by reading a value from this memory we wrote there earlier. This is normally only possible if there is an battery present which keeps the backup ram from loosing its contents. However, it appears that the mcHF UI keeps enough energy floating into the backup ram for quite some time after power down and thus our code can be fooled when powering on before the backup ram gets turned off.
73
Danilo
|
Title: Re:RTC Battery quickly discharged
Post by: DL8EBD on 27. December 2018, 13:08:08
Einspruch Danilo!
hart an GND legen ist natürlich klar, aber mittels 47k-56k an GND funktioniert der workaround ufb wie wir ja an unserer OV I40 UI feststellen konnten.
Natürlich ist das laut Referenzmanual nicht korrekt, aber es funktioniert.
für den STM32F76x:
https://www.st.com/content/ccc/resource/technical/document/reference_manual/group0/96/8b/0d/ec/16/22/43/71/DM00224583/files/DM00224583.pdf/jcr:content/translations/en.DM00224583.pdf (https://www.st.com/content/ccc/resource/technical/document/reference_manual/group0/96/8b/0d/ec/16/22/43/71/DM00224583/files/DM00224583.pdf/jcr:content/translations/en.DM00224583.pdf)
siehe Seite 152 gleiche Innenschaltung wie in dem von Dir benannten Manual für den 32F42x (dort Seite 151)
STM32F42x:
https://www.st.com/content/ccc/resource/technical/document/reference_manual/3d/6d/5a/66/b4/99/40/d4/DM00031020.pdf/files/DM00031020.pdf/jcr:content/translations/en.DM00031020.pdf (https://www.st.com/content/ccc/resource/technical/document/reference_manual/3d/6d/5a/66/b4/99/40/d4/DM00031020.pdf/files/DM00031020.pdf/jcr:content/translations/en.DM00031020.pdf) |
Title: Re:RTC Battery quickly discharged
Post by: DB4PLE on 28. December 2018, 10:42:11
Hallo Thomas,
möglicherweise würde ein Widerstand auch von direkt von 3.3 V nach GND ähnliche Wirkung entfalten. Im Prinzip sorgt der Widerstand ja lediglich für eine zügige Entladung des Kondensators und hält den Prozessor dann im Reset auch wenn noch etwas Vdd vorhanden ist.
Hier wird https://community.st.com/s/question/0D50X00009XkgBTSAZ/vbat-current-draw-high-just-after-vdd-removed (https://community.st.com/s/question/0D50X00009XkgBTSAZ/vbat-current-draw-high-just-after-vdd-removed) ein ähnliches Problem diskutiert. Könnte es sein, das auch bei unserem Fall der hohe Stromverbrauch nach einiger Zeit nach dem Ausschalten sinkt?
73
Danilo
|
Title: Re:RTC Battery quickly discharged
Post by: DF8OE on 28. December 2018, 10:46:55
Hallo Danilo,
habe ich untersucht: bei uns bleibt der Stromverbrauch so lange hoch, bis die Batterie leer ist.
vy 73
Andreas |
Title: Re:RTC Battery quickly discharged
Post by: WD8BXS on 28. December 2018, 18:14:35
So Andreas, are you saying that it is better to just forget about the RTC?
Chuck |
Title: Re:RTC Battery quickly discharged
Post by: DF8OE on 28. December 2018, 18:41:28
No. For OVI40 we have investigated a workaround. NRST pin is sonected to a capacitor to GND and a "discharge resistor" of 47K...68K (board dependent) to GND. You can test following:
- Insert new battery
- set clock
- switch off mcHF
- test if Vcc pins of MCU do have voltage (>0.3V) even after few minutes
If so: F4 has same issue as F7 and H7. This voltage comes from battery and feeds MCU!!!
Now add a resistor from NRST to GND - start with 68K and decrease value until voltage disappears quickly. Now you have found correct value for parallel resistor.
vy 73
Andreas |
Title: Re:RTC Battery quickly discharged
Post by: WD8BXS on 28. December 2018, 21:23:40
Will do!
I am waiting on a new battery holder first.
de Chuck WD8BXS |
Diskussions- und Newsboard des DARC-Ortsverbandes I40 | Powered by YaBB SE
© 2001-2003, YaBB SE Dev Team. All Rights Reserved.
|