Differential voltage range across ADS1115

General project help for Adafruit customers

Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.
User avatar
gica
 
Posts: 6
Joined: Mon Oct 30, 2017 6:30 am

Re: Differential voltage range across ADS1115

Post by gica »

Problem solved, well not solved but identified

up to the ADC there is a chain from the mains, with an SCP-12-75(220V to [email protected]) SMPS <-> DC-DC buck (LM2596HV) <>RPi <> ADC1115

Well all power supply chain is non-isolated, so the battery output control pin form the SCP-12 has a "shortcut" to the ADC through the common ground, stupid mistake and I think this is the mother of all burned ADC in equivalent project
So I did manage to feed ADC with full voltage

User avatar
adafruit_support_mike
 
Posts: 67454
Joined: Thu Feb 11, 2010 2:51 pm

Re: Differential voltage range across ADS1115

Post by adafruit_support_mike »

That would do it.

Those kinds of problems are tricky because the GND connections aren’t always obvious (or specified).

There are ways to build DC isolation into a signal path if you need it though. If you’re measuring AC signals it’s fairly easy: just put a couple of large capacitors in series with the ADC inputs. The caps will block all DC current, and will level-shift the input signals to fit in the ADC’s safe voltage range.

You do need to provide a reference voltage for AC-coupled signals like that. For a single-supply ADC like the ADS1115, a1M-1M voltage divider between VCC and GND on each pin will give the signals a VCC/2 reference to work against.

If you need to measure slow signals or DC, it’s usually easiest to convert it to AC going through the isolation capacitors then convert it back to DC. Connect a switch between the input ends of the two isolation capacitors and open/close that at a frequency that’s high compared to the signal of interest.. 1kHz is good enough for DC work.

The switch turns the DC signal into a square wave that will go through the capacitors without much attenuation, but you’ll only see the square wave on the ADC if you take one reading when the switch is closed and another reading when the switch is open. If you skip the ‘switch closed’ readings and only read the ADC when the switch is open, you’ll see what looks like a DC or low-frequency signal.

User avatar
gica
 
Posts: 6
Joined: Mon Oct 30, 2017 6:30 am

Re: Differential voltage range across ADS1115

Post by gica »

using voltage divider , if by any means the path to GND come undone, I'll end up into the same problem. Feed up full voltage into the ADC.

regarding AC voltage, as far as I know a capacitor in series in a AC path will only limit the current by phase shifting not the voltage. Maybe you might be saying that lowering the current to a resonable value you can afford to drop it on a voltage divider, due to low power to dissipate. That might be working.

My source to be sensed is DC 10.6 - 14.7v. I've read that a max471 breakboard can do the job. It is said that the bkb it is powered by the source to be sensed so it seems to be fair enough. I some power path fail some junctions in max will change to block and I'll read null probably, but no smoke.

User avatar
adafruit_support_mike
 
Posts: 67454
Joined: Thu Feb 11, 2010 2:51 pm

Re: Differential voltage range across ADS1115

Post by adafruit_support_mike »

Like this:
chopper.jpg
chopper.jpg (16.29 KiB) Viewed 1360 times
The voltage divider on the 10v-15v signal drops the voltage to the correct range. The capacitors block any DC connection between that divider and the ADC. The voltage dividers on the ADC input pins hold both pins at VCC/2 with no other input. The mosfet turns the voltage across the lower resistor into a 1kHz square wave which will pass through the capacitors.

The signals that reach the ADC pins will be a pair of opposing square waves centered around VCC/2. If you only read the ADC while the mosfet is shut off, you'll only get measurements for the half where the voltage at A1 is lower than the voltage at A0, and it will look like you're measuring a DC value.

User avatar
SpaulDrummerNerd
 
Posts: 2
Joined: Sat Jan 06, 2018 1:42 pm

Re: Differential voltage range across ADS1115

Post by SpaulDrummerNerd »

This is somewhat similar to the initial posts. I am looking to make a battery voltage measurement using the differential mode/connection. I am concerned what would happen if the battery cell (0-4.2Vdc) is put into test fixture in a reversed orientation. Would this potentially damage it or would is simply just show up as a negative measurement because the connection has changed between the two differential connections. It would still be a positive battery voltage, so I'm thinking (hoping) that it would still be ok. Any help is much appreciated!!!

User avatar
adafruit_support_mike
 
Posts: 67454
Joined: Thu Feb 11, 2010 2:51 pm

Re: Differential voltage range across ADS1115

Post by adafruit_support_mike »

Reversing the battery won't damage anything if you're taking a differential reading. You'll just get output that says pin1 is N volts lower than pin2 rather than N volts higher than pin2.

User avatar
SpaulDrummerNerd
 
Posts: 2
Joined: Sat Jan 06, 2018 1:42 pm

Re: Differential voltage range across ADS1115

Post by SpaulDrummerNerd »

That was my thought, that even changing the connection of the battery, it will still be a positive battery voltage so to speak. Just concerned by the datasheet saying that anything below GND will damage the IC. This must be in Single Ended only??? Anyway, thanks for the reply/your time Mike!

User avatar
adafruit_support_mike
 
Posts: 67454
Joined: Thu Feb 11, 2010 2:51 pm

Re: Differential voltage range across ADS1115

Post by adafruit_support_mike »

It isn't only single-ended. Any signal that goes to a negative voltage and has a DC connection to an input on the ADC will probably damage the chip.

The circuit a few posts up makes it harder for an arbitrary signal to go to a range that can hurt the ADC because the capacitors block any DC voltage. Assuming equal 10k-10k voltage dividers on both inputs, the pins A0 and A1 will see signals centered around VCC/2, and those signals will only go negative if the voltage on the input side changes by more than VCC.

Locked
Please be positive and constructive with your questions and comments.

Return to “General Project help”