Author Topic: Current Ranger giving different readings as compared to uCurrent Gold  (Read 855 times)

shivams

  • Newbie
  • *
  • Posts: 7
We recently received our Current Ranger. When trying to measure the sleep state of Particle Electron module, it is giving different readings as compared to μCurrent Gold.

While the μCurrent Gold gives around 130 μA during sleep (which is also the officially stated metric), Current Ranger is giving 14-16 mA.

This seems to be a surprisingly big discrepancy.

Current Ranger seems to be giving correct values when measuring current across a given voltage and resistance.

Is it that the shunt resistor is somehow causing the Particle to increase its consumption? OR is it giving wrong values?

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5850
  • Country: us
    • LowPowerLab
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #1 on: January 25, 2019, 09:02:17 AM »
Is this with autoranging or manual mode?
Are you using a DMM to measure output?
How is everything connected?

shivams

  • Newbie
  • *
  • Posts: 7
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #2 on: January 25, 2019, 10:30:15 PM »
@Felix

  • This happens in both manual and autoranging mode.
  • I was not using a DMM to measure output. Simply the OLED screen.
  • DUT was being powered by a LiPo battery (at 4V) and wire from the high-side was cut to connect to the Current Ranger.

I will perhaps make a small video of the whole setup and demonstrating the discrepancy. I am out right now. Will share as soon as I get back to my lab.

shivams

  • Newbie
  • *
  • Posts: 7
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #3 on: January 26, 2019, 08:04:12 AM »
@Felix

Here is a video demonstrating the issue: (Pay attention to the Particle LED's state. ON means NORMAL state and OFF means SLEEP state. It's not very clear in the video)


In this video, we used Current Ranger and μCurrent Gold for evaluating the sleep state of Particle Electron module.

Official metrics for Particle Electron are:
  • Normal State ⇒ 38 mA
  • Sleep State ⇒ 130 μA

However, both Current Ranger and μCurrent Gold show these metrics:
  • Normal State ⇒ 55-150 mA
  • Sleep State ⇒ 14-16 mA

However, when using μCurrent Gold, if we bring the shunt-selection switch to μA range for a little while, when it is in Normal State, it starts showing the official metrics from there on wards. Perhaps, there is some kind of bias that gets removed by this switching.

Current Ranger seems to always show the same metrics, and never shows the official metrics.

Question then is: Are both of them reporting wrong metric? Or, somehow the current consumption of Particle increases because of an intervening circuit?

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5850
  • Country: us
    • LowPowerLab
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #4 on: January 26, 2019, 08:21:34 AM »
Thanks for the detailed video.

What I see though is that the Particle does not just settle in sleep mode but keeps jumping around all the time even if you don't change the range at all.
Also, on the CR you switch to auto-ranging, and on the uC you switch manually back and forth which causes it to go to "0" for the first time, thats quite different than auto-ranging which can switch much faster and has no switching artifacts that the mechanical switches have (if you look on the scope its very different).

Any way you could send me your setup and code for the particle so I can look at this myself?
Not sure how else to really tell what's going on. We really have to establish if the Particle is really sleeping, if switching burden voltage has any effect on it (brown out) and if so - how to avoid that.
I mentioned in the guide adding a diode from IN+->IN- could help in this regard, to limit burden drop to diode fwd voltage.

shivams

  • Newbie
  • *
  • Posts: 7
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #5 on: January 26, 2019, 11:50:10 AM »
Here is the code for Particle Electron: https://gist.github.com/shivams/6212275db9a69cd2444e064715f4e9b6

If you have access to a Particle Electron module, please try it with your Current Ranger.

I am now thinking that this question perhaps should be posted on EEVBlog (μCurrent Gold creator) or Particle forums as well, since it doesn't really look like an issue with CR.

And thanks for the advice of adding a diode from IN+->IN-. I will try that.
« Last Edit: January 26, 2019, 11:53:33 AM by shivams »

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5850
  • Country: us
    • LowPowerLab
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #6 on: January 26, 2019, 07:07:27 PM »
To really tell if there is a problem, I will need your exact setup, with the board in question or CR shipped to me. Then I can quickly test the CR and tell if it's faulty or not.

Anyway, the code you posted is very basic.
I did similar tests many times with Moteinos during development of CR, and even more complex code, waking up, transmitting, going back to sleep, or just flipping state from wake to sleep.

When I ask a board to SLEEP, I expect it to go to 0mA, not jump around like a rollercoaster, that's what I see in your video, so that doesn't inspire confidence that the node is truly sleeping.

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5850
  • Country: us
    • LowPowerLab
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #7 on: January 28, 2019, 04:47:15 PM »
shivams,
Can you take a close look at the SMD side, for any SMD components that might have been dislodged during assembly.
Especially close to the Terminal mounting points. If any of these SMD components are dislodged or damaged, that could explain the behavior you are seeing.

shivams

  • Newbie
  • *
  • Posts: 7
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #8 on: January 31, 2019, 09:46:39 AM »
Hi Felix

I am away on a small vacation. I will be back next week and will update you then about the SMD issue you mentioned.

However, I'd like to mention that to me it seems like it's an issue with Particle Electron module itself. Somehow, because of the burden voltage, it does not go to sleep, as this behavior is also evident with μCurrent Gold. Further, this behaviour is repeatable across all our Particle Electron modules (we have 30 in our lab) and 2 μCurrent Gold modules and one Current Ranger.

So, perhaps the Current Ranger's manufacturing has no flaw. Either, we'd need to change the design of CR (and also μCurrent Gold) somewhat so that whatever issue keeps the Particle awake may go away, or else it's some flaw in Particle Electron's design.


Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5850
  • Country: us
    • LowPowerLab
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #9 on: January 31, 2019, 09:57:45 AM »
shivams,
If you're willing to send me a Particle module that has this behavior, I can check it on a CR and also uCGOLD that I have here in my lab.

The problem the other person that posted here (I separated the thread) ended up being because during installing the terminals, some essential SMD components near a terminal were knocked off the board and hence the CR was not working at all and giving very erratic readings.
I will need to try to move those parts a bit further away from terminals to avoid users accidentally damaging them. Always good to double check for such damage just in case, although in your case I dont think is a problem.

shivams

  • Newbie
  • *
  • Posts: 7
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #10 on: February 05, 2019, 07:50:19 AM »
Felix, actually the behaviour is same across all Particle Electron modules (we have 30 in our lab, and all show the same behavior). Also, it would be difficult to send a module from our stock, as they are for a research project in our university and occupied for some field testing.

I can actually order one for you personally :) Or else, if you could order one "Particle Electron 3G Asia", it'd be great.

Thanks Felix for your amazing support.

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5850
  • Country: us
    • LowPowerLab
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #11 on: February 07, 2019, 02:29:09 PM »
shivams,
Your feedback is appreciated  :)
Do you have any other simple device/arduino or low power DUT that can sleep in nA or uA range?
That would be a very simple benchmark to compare against the particle boards.
Ie set them to start awake and perhaps blink an LED, then go to deep sleep for a few seconds, then wake back up, and repeat.

shivams

  • Newbie
  • *
  • Posts: 7
Re: Current Ranger giving different readings as compared to uCurrent Gold
« Reply #12 on: February 09, 2019, 08:16:27 AM »
Okay Felix. Will try that and let you know.