Author Topic: ACS711 Current Sensor Fault Line Output Voltage???............  (Read 324 times)

0 Members and 1 Guest are viewing this topic.

Online SmokeyTopic starter

  • Super Contributor
  • ***
  • Posts: 3105
  • Country: us
  • Not An Expert
ACS711 Current Sensor Fault Line Output Voltage???............
« on: January 20, 2025, 08:18:57 am »
well this is frustrating. 

tl:dr = With no pull-up resistor on the fault line, and in normal non-faulted state, the ACS711 is outputting 4.4V on the unloaded fault line which is supposed to be an open-drain with an internal 240K pulldown.

History:
I've successfully used these ACS711 series hall effect current sensors in a bunch of designs for a number of years, both personal projects and production stuff.   They have their issues (like easily picking up adjacent fields) but do great for general current sensing.  And they are cheap!  I've even used the Fault line before (but always like a normal person, with a pull up). 

Situation:
ACS711KEXLT-15AB-T is the version at issue here (but all from family probably apply).  Powered from 5V.  Confirmed the supply is good.  Chip looks happy, with actual current output sitting nominally at about 2.5V and moving correctly with changes in actual current.  Legit parts sourced from Digikey.
https://www.allegromicro.com/~/media/Files/Datasheets/ACS711-Datasheet.ashx
I'm not sure what the deal is here.  The datasheet clearly shows the fault output as an open drain with internal 240K pulldown. 


When I measure the unloaded fault line voltage, it shows a solid 4.4V.  Scope shows no appreciable ripple. 
Shorting the fault line to ground results in the voltage collapsing but still sourcing about 200uA.
Another strange data point is that with a 7K load, it can still source about 150uA and the voltage holds up to about 1V.  That sounds like a lot for some sort of internal coupling or other parasitic. 

If I force the ACS711 into an actual over-current fault condition then the fault line pulls hard to 0V as specified.  No issue there.

I cut the trace so the fault output line is disconnected from the rest of the board, so it's not something backfeeding it.  I also replaced the chip (twice) to make sure I didn't just break something.  Again, no pullup is installed here.  There isn't anything on the board going on close by either (nothing switching, or high current).  Total supply draw for this whole board is 13mA during all this testing.

Does this make any sense to anyone?
« Last Edit: January 20, 2025, 09:01:38 am by Smokey »
 

Online SmokeyTopic starter

  • Super Contributor
  • ***
  • Posts: 3105
  • Country: us
  • Not An Expert
Re: ACS711 Current Sensor Fault Line Output Voltage???............
« Reply #1 on: January 20, 2025, 11:48:46 pm »
Afternoon crowd bump.  I'd be surprised if I'm the first person to actually measure that line without a pullup.
 

Offline thm_w

  • Super Contributor
  • ***
  • Posts: 7663
  • Country: ca
  • Non-expert
Re: ACS711 Current Sensor Fault Line Output Voltage???............
« Reply #2 on: January 21, 2025, 12:43:30 am »
No, doesn't make sense, but, is there a reason you aren't using a pullup on the line?

They seem to have a variety:
- ACS710 is actively driven: https://www.allegromicro.com/~/media/Files/Datasheets/ACS710-Datasheet.ashx
- ACS71240 is shown as no pulldown: https://www.allegromicro.com/-/media/files/datasheets/acs71240-datasheet.pdf

What is the resistance from Fault to Gnd and Fault to Vcc on a bare IC?
Profile -> Modify profile -> Look and Layout ->  Don't show users' signatures
 

Online SmokeyTopic starter

  • Super Contributor
  • ***
  • Posts: 3105
  • Country: us
  • Not An Expert
Re: ACS711 Current Sensor Fault Line Output Voltage???............
« Reply #3 on: January 21, 2025, 01:31:16 am »
I'm using the ACS711 fault output to pull down the enable on another part in case of overcurrent.  That other part has it's own enable source voltage that needs to be able to independently enable/disable the system.  So if I had a pull up on the ACS711 fault line it would keep the other chip alive when it's enable went low, which would be bad (which is also what's happening now that the fault line is making it's own voltage).

I've already added a series diode on the ACS711 fault line for the next revision, which will keep the fault line mystery voltage from keeping the other part on.  But based on the ACS711 part spec I shouldn't need to do that, and it's sort of a pain to hack it into the existing layout. 

Fault to ground = 200K when the part is off.
Fault to Vcc = 173K when the part is off.
« Last Edit: January 21, 2025, 01:35:26 am by Smokey »
 
The following users thanked this post: thm_w

Online SmokeyTopic starter

  • Super Contributor
  • ***
  • Posts: 3105
  • Country: us
  • Not An Expert
Re: ACS711 Current Sensor Fault Line Output Voltage???............
« Reply #4 on: January 22, 2025, 12:55:46 am »
well I guess I've leave this as a PSA...

...if you want to use the fault output line on the ACS711 then be heads up that the internal block diagram isn't accurate to what the device is actually doing, and expect it to source essentially Vcc with about a 22K output impedance on that output pin instead of being an open drain with a pull down as shown in the datasheet.
« Last Edit: January 22, 2025, 01:03:24 am by Smokey »
 

Offline fourtytwo42

  • Super Contributor
  • ***
  • Posts: 1222
  • Country: gb
  • Interested in all things green/ECO NOT political
Re: ACS711 Current Sensor Fault Line Output Voltage???............
« Reply #5 on: January 23, 2025, 10:56:41 am »
Well I dont think those sort of internal block diagrams are intended to form part of the specification, they are an indication of  function, no more.

If you read the written spec, fault VOH & VOL are specified with an external pullup of 1K. There is no leakage current specification for that pin. So if you use it outside of the specification anything can happen especially between batches of chip.

If you are designing a product beware to use parts within specification.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf