Author Topic: Maximum input voltage limit? 12v power supply for DUT?  (Read 1968 times)

IOT_man

  • NewMember
  • *
  • Posts: 2
Maximum input voltage limit? 12v power supply for DUT?
« on: February 14, 2019, 05:29:33 PM »
I went to use my current ranger for the first time yesterday and must be using it incorrectly.

I connected my DUT in series with the positive lead of my power supply which was at 12V.  The current ranger was set to auto range and the biasing to 0V (the DUT does not have a battery so it is never charged).  I was just intending to read the current measurements from the LCD to get some rough ideas of current draw in various power states, so I did not have it hooked up to my scope in any way.  In fact, the scope has never been connected.  I did have the USB cable attached to a USB charger and was using a Lithium Polymer battery powering the current ranger.  The USB charger did not have a 3 prong outlet so I made the assumption that its ground was floating but perhaps I was wrong in that assumption.  The DUT only draws 30mA consuming its maximum power.  I later re-confirmed this with my fluke multimeter on this test setup.

When the device was powered on via an inline switch, the LCD flickered, powered down and smoke came from the case.

Upon inspection, it appears that the micro suffered some damage near pins 4-8.  I could not see any other components damaged.  I also inspected under a microscope and did not see any bad solder joints or obvious manufacturing defects.  Since the micro is damaged beyond repair, I decided to understand what went wrong.  I probed pins 4-8 of the micro while briefly turning on power to the setup again and noticed that pins 4 and 7 (PA03 and PB08) were roughly at the same voltage as the (source +) jack. 

I have not taken time to review the schematic in detail but I thought I would post this to the community to see what feedback there is.

Attached are pics of my hardware and a schematic of the setup.

In thinking about what I could have done differently or what might have happened, I had a few questions and ideas.

1)  Is there an absolute voltage limit for the Source / Load jacks?  Should I be safe when using 12V DC to power my DUT?
2)  Is it possible to get a ground loop between the USB cable and the DUT's power supply?  I assume the answer is yes and if so I could use a battery to power the DUT next time if 12V DC is allowable.
3)  Is there a problem with connecting the DUT on the low side of the circuit.  In other words, connect the CR between the ground of my power source and ground of the DUT?

Thanks in advance for any feedback.
« Last Edit: July 02, 2020, 08:13:51 AM by Felix »

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: CR setup with 12 power supply for DUT
« Reply #1 on: February 14, 2019, 09:32:30 PM »
Seeing the damage, I wonder what the baptism of fire has taken long with the SAMD21.

It's the first time I see this type of damage done, although it can be done easily if assumptions are made that input/output/USB are not connected in any way or share a common ground in any way. I would be interested to find out what was actually smoked other than the SAMD21 and see how fixable it is.

I know this is a huge bummer for you on your first use. I hope you had a chance to at least visit the guide, maybe it was not completely clear that such a use case will really damage the CR.
I can offer to try to fix it for you for a small fee to cover parts if you're willing to also return it and pay for shipping.
Otherwise I can help assist in doing the repair yourself if you got the tools to do it and SMD doesn't scare you.

The USB charger did not have a 3 prong outlet so I made the assumption that its ground was floating but perhaps I was wrong in that assumption.

It doesn't matter what voltage you power your DUT from. CR only cares about the current passed through the input terminals, meaning - the voltage drop across the internal shunts, that's it. It can measure up to ~3.3A unidirectionally (DC) and ~1.65V bidirectionally (AC).

The USB has a GROUND, which I bet was the same GROUND as your 12V PSU. Even if it wasn't, if there's a potential difference between the input terminal + (12V) and the GROUND of the USB, that's really bad. That's more than enough to smoke the CR if you hook pass 12V through it from a different GROUND wired to input or output. Or if you form a closed circuit through the CR in any way shape or form (ie 12V at some point at input/output terminals) and the GROUND through the USB or at input/output or some of the headers.

You should never have USB connected unless everything else is floating so current cannot form a closed circuit through the CR. The only reason for USB is charging the battery, or perhaps data logging, if your DUT is on battery and the output is not connected (or connected to a floating DMM).

It's all explained in the guide, I should perhaps make some graphical drawings with big green OK connections and some with big red NO-NO connections. (anyone with ninja skills in that deparment?).

« Last Edit: February 14, 2019, 11:27:38 PM by Felix »

IOT_man

  • NewMember
  • *
  • Posts: 2
Re: CR setup with 12 power supply for DUT
« Reply #2 on: February 15, 2019, 01:26:14 PM »
Felix,
Thank you for your reply and feedback.  In the short term, I have purchased another current ranger which will allow my development efforts to keep moving forward.  This is certainly a better option that buying 1000's of dollars of test equipment!  I have access to resources to place a new micro down and take any measurements you want but I am also willing to send the components back for you to analyze too, whichever you prefer.  If you are able send me a private message, we can figure that out.

After studying the schematic and taking more measurements, I agree that the ground loop is the likely issue.  Interestingly, in my test setup, the USB ground was floating until the connector was inserted and charging started, so it was hard to detect.

For the benefit of future users, I will try to summarize what I have learned through this experience.  Felix and others, please correct or improve upon this if I misstate something here.

There are three possible "grounds" in the system that can contribute to a ground loop:  The USB cable, the scope, and the DUT power source.  A maximum of 1 ground can be fixed and the rest must remain floating (as the safety portion of the instructions indicate).  Another way of stating this is two of those three grounds must be floating.  There are various ways to do this for each component. 

To insure your DUT power source is floating, use a battery for its power source or if a power supply is required, check the specs to make sure the ground is floating. 

To insure your scope is floating, there are a few options which are best summarized by this tektronix tech brief. 
https://www.tek.com/document/technical-brief/floating-oscilloscope-measurements-and-operator-protection
Or you can use a multimeter.

For the USB, the easiest solution is to just unplug it and run the CR off a battery.  If that will not meet your needs, I ran onto this item which might help (which I have not tested).
https://www.adafruit.com/product/2107

Thanks again

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: CR setup with 12 power supply for DUT
« Reply #3 on: February 15, 2019, 01:46:24 PM »
IOT_man,
Yes, you are correct, and I added a few more details in the Safety page to help make all this more clear.
Conceptually there are 3 connection points which do not share a common ground and are not galvanically isolated.

If the adafruit USB isolator can provide islated power (which it appears to be be doing through the DC:DC converter, then it might do the trick.
I would consider integrating a similar solution in the CR in a "pro" version perhaps - it would cost significantly more and would be more complex to make, larger board etc.
But that still leaves the problem of input-output ground differences, so if I go the isolated USB route, it may be worth researching how to achieve full input/output isolation as well - this might be tricky because I don't want to hinder the bandwidth of the device.

alexsh1

  • Jr. Member
  • **
  • Posts: 58
  • Country: gb
Re: CR setup with 12 power supply for DUT
« Reply #4 on: February 15, 2019, 05:13:42 PM »
IOT_man,

Your Adafruit USB isolator costs $34!!! That's 1/4 CR price!
Why just not running CR of the $8 battery? Why do you need to power the CR from USB?

I have tested many DUTs (DC only) with the CR and never had any issues. There are certain safety rules - both the CR and the DUT must be powered by separate batteries (floating ground). If you take this into account, there is no need to galvanically isolate anything IMHO. Isolation increases the level of complexity.