Different value in "upis" (PiCO) vs serial @status call

Support for the The Peripherals I2C Control – PiCO – Interface
flyvert
Posts: 26
Joined: Sat Jul 19, 2014 3:22 pm

Different value in "upis" (PiCO) vs serial @status call

Post by flyvert » Wed Oct 01, 2014 9:56 pm

Hi

The upis.c program I downloaded from this site returns a slightly higher current than the @CUR command via the serial interface
Discrepancy in upis vs minicom status call.png
Discrepancy in upis vs minicom status call.png (11.3 KiB) Viewed 7906 times
Left: repeated @CUR commands Right: while [ true ]; do upis -a; sleep 1; done

The voltages are also a little bit skewed.

What can be the reason?

/f
User avatar
Pimaster
Site Admin
Posts: 1615
Joined: Fri Sep 14, 2012 9:50 am

Re: Different value in "upis" (PiCO) vs serial @status call

Post by Pimaster » Thu Oct 02, 2014 2:29 am

Hi,
Are you making measurements at the same time? as the differences are very small about 30-40 mV, could be part of small voltages changing.
Please make sure that you are doing measurements at the exact same time.

Warmest Regards
Pi Master
Warmest Regards
PiM
---
Designing with Mentor Graphics PADS - www.pads.com
Please read and follow the PiForum rules
http://www.forum.pimodules.com/viewtopic.php?f=13&t=196
---
flyvert
Posts: 26
Joined: Sat Jul 19, 2014 3:22 pm

Re: Different value in "upis" (PiCO) vs serial @status call

Post by flyvert » Thu Oct 02, 2014 8:54 am

Pimaster wrote:Hi,
Are you making measurements at the same time? as the differences are very small about 30-40 mV, could be part of small voltages changing.
Please make sure that you are doing measurements at the exact same time.

Warmest Regards
Pi Master
Exactly the same time is difficult to acchieve without writing an application that simultaneously queries RS232 and I2C/PiCO.

I therefore made a "shell loop" iterating the PiCO temp check each second while I simultaneously hacked "@CUR" into the minicom console.

I think its "simultaneous enough"...
- All but one PiCO read resulted in 413 or higher current
- All but one RS232 read resulted in 407 or lower current

Looks odd to me...

Have you tried yourself in some reference system at your office?
flyvert
Posts: 26
Joined: Sat Jul 19, 2014 3:22 pm

Re: Different value in "upis" (PiCO) vs serial @status call

Post by flyvert » Mon Oct 06, 2014 11:57 am

Hello Pimaster - have you had time to test the PiCO interface yourself?

In the picture below you can see the results from the PiCO read - as you can see I get spikes on many of the properties returned by UPiS PiCO interface: Vusb, Vbat and Itotaldrain.
UPiS - 1 day - minute values.png
1 minute values
UPiS - 1 day - minute values.png (81.14 KiB) Viewed 7889 times
I believe the spikes are incorrect values - rather due to some bug in the PiCO interface, etc.
Please comment.


By calculating averages the spikes vanish - below recordings from my UPiS's first week!
UPiS - 1 week - hourly averages.png
Hourly values
UPiS - 1 week - hourly averages.png (88.18 KiB) Viewed 7889 times
As you can see I replaced the old powersupply with a new 5.25 volt as per your recommendation (at least you said 4.6 volts is too low).
Please comment on the charging - is this how "trickle charging" works? I.e. to charge when Vbat reaches some threshold (4.1 volt)?

/f
User avatar
Pimaster
Site Admin
Posts: 1615
Joined: Fri Sep 14, 2012 9:50 am

Re: Different value in "upis" (PiCO) vs serial @status call

Post by Pimaster » Mon Oct 06, 2014 9:53 pm

Hi,
No, I will do it within next days. As we are closing next product production was not able to work on it. It need to to analyze in depth. But we will do it with next days.
Warmest regards
Pi Master
Warmest Regards
PiM
---
Designing with Mentor Graphics PADS - www.pads.com
Please read and follow the PiForum rules
http://www.forum.pimodules.com/viewtopic.php?f=13&t=196
---
flyvert
Posts: 26
Joined: Sat Jul 19, 2014 3:22 pm

Re: Different value in "upis" (PiCO) vs serial @status call

Post by flyvert » Mon Oct 06, 2014 10:19 pm

Pimaster wrote:Hi,
No, I will do it within next days. As we are closing next product production was not able to work on it. It need to to analyze in depth. But we will do it with next days.
Warmest regards
Pi Master
OK - let me know if you prefer plain text logs of the output of "upis" command or if you can interpret the trends I paint with a Javascript class (Flot).
User avatar
Pimaster
Site Admin
Posts: 1615
Joined: Fri Sep 14, 2012 9:50 am

Re: Different value in "upis" (PiCO) vs serial @status call

Post by Pimaster » Mon Oct 06, 2014 10:25 pm

Hi,

I will contact you soon, on the forum or directly. We will search on it, but due to new production releasing we are overload and not able to perform these tests, as a silence and no interruption is needed. We are a small company just few people.

My Warmest regards
Pi Master
Warmest Regards
PiM
---
Designing with Mentor Graphics PADS - www.pads.com
Please read and follow the PiForum rules
http://www.forum.pimodules.com/viewtopic.php?f=13&t=196
---
flyvert
Posts: 26
Joined: Sat Jul 19, 2014 3:22 pm

Re: Different value in "upis" (PiCO) vs serial @status call

Post by flyvert » Sat Oct 11, 2014 6:16 pm

Any news?

Here is another example - all measurments (Vusbfeed, Vbat, Vrpi and Irpi) occasionally show large spikes which I doubt to be correct.
image.jpg
image.jpg (123.03 KiB) Viewed 7823 times

I believe it should be fairly simple to repeat this in an arbitrary UPiS setup and a shell script loop, eg:

while [ true ]; do
upis -u
upis -r
...
sleep 10
done


/f
User avatar
Pimaster
Site Admin
Posts: 1615
Joined: Fri Sep 14, 2012 9:50 am

Re: Different value in "upis" (PiCO) vs serial @status call

Post by Pimaster » Sun Oct 12, 2014 10:25 pm

Hi,
We are still setting up the new production of the UPS PIco Basic

http://www.forum.pimodules.com/viewtopic.php?f=12&t=105

We will be able to test it more in depth by the end of this week. It needs to setup a special measuring workshop with 2 multimeter and DSO, but these devices are actually used in different stuff. Please be patient as it is not a very critical matter.

My Warmest Regards
Pi Master
Warmest Regards
PiM
---
Designing with Mentor Graphics PADS - www.pads.com
Please read and follow the PiForum rules
http://www.forum.pimodules.com/viewtopic.php?f=13&t=196
---
Kai
Posts: 2
Joined: Fri Oct 17, 2014 12:19 am

Re: Different value in "upis" (PiCO) vs serial @status call

Post by Kai » Fri Oct 17, 2014 3:33 am

Just for consideration:

a) Could that spikes be related to the data transfer of the I2C Interface itself. It uses pull-ups (but these are usually in the 1 to 10 K Ohm range)

b) I have also seen in other applications during the I2C-Ack cycle a pin-driver crash generating a high current spike due to wrong timing (usually the sink is stronger and thus the software works fine since it sees the Slave-Ack).
Post Reply