Measure 2 voltages without common ground?

I have a cheap solar charge controller that seems to be common positive. I want to measure the PV and battery voltage using the WC8 analog ins. The WC8 runs off the battery side of the charge controller so the battery ground and WC8 ground are common. This works fine for collecting accurate battery voltage. However I also want to monitor the Solar panel voltage on the input side of the charge controller which seems to have different ground reference. Using the second analog in does not read properly because of the ground reference issue. Probably because the controller is switching the negative side between PV and battery to control charge.  Any simple suggestions to collect the PV voltage that has a different ground plane?  
 
Will need an A/D that can be optically coupled to the WC since solar has a different ground reference.

/tom
 
Machineman said:
  Any simple suggestions to collect the PV voltage that has a different ground plane?  
 
This can be achieved quite simply with a differential amplifier. (Simplest case would be 4 resistors and a single opamp, or you could adapt one of my terminal and amplifier boards to do it)
Keep impedances high enough that you are not confusing the charge controller or causing unnecessary leakage (thats easy).
Power the diff amp from a supply that shares the same ground as the webcontrol.
Careful selection of your divider and gain resistors and you'll get results that are certainly within a few percent. If you have a decent meter, you can tweek it for better accuracy.
 
As an after-thought, there's another way of doing it even more simply.
 
Measure the ground-side of the PV input to another analog in.
Then, if your charge-controller has common-positive, you can determine PV voltage by:
 
Measure the output voltage (ground to CC output (call this Vbat)
Measure the PV negative. (call this V)
Subtract V from Vbat to get Vpv   (ie, Vpv=(Vbat-V)
This assumes PV voltage is less than system voltage, which is unreasonable, you will likely need an inverting amplifier to achieve it, but the principle is sound and can be used for other non-ground-referenced signals.
 
Yep, was playing around with a similar idea.
 
PV ground to Bat Ground is <0VDC when the PV voltage is higher than the battery, and >0VDC when the PV voltage is below the battery. So an inverter would only work if the PV was higher than the battery.  But really all I need to know is if the PV isn't charging the battery anymore and to shut off my water pump.  So just solved this by reading A/I, if its 0 (negative input) then the PV is charging the battery. If positive then that's how much lower the PV voltage is below the battery, not charging.
 
Off course this would all be much easier if that cheapy charge controller was common ground,  but looks like most lower end <10A capacity units aren't.
 
Still need to run thru all the scenarios to make sure it works 100%
 
Machineman said:
Yep, was playing around with a similar idea.
 
PV ground to Bat Ground is <0VDC when the PV voltage is higher than the battery, and >0VDC when the PV voltage is below the battery. So an inverter would only work if the PV was higher than the battery.  But really all I need to know is if the PV isn't charging the battery anymore and to shut off my water pump.  So just solved this by reading A/I, if its 0 (negative input) then the PV is charging the battery. If positive then that's how much lower the PV voltage is below the battery, not charging.
 
Off course this would all be much easier if that cheapy charge controller was common ground,  but looks like most lower end <10A capacity units aren't.
 
Still need to run thru all the scenarios to make sure it works 100%
 
Another option would be a shunt in the battery ground, and an amp (opamp) to give you a voltage proportional to the actual current. You could read that with the webcontrol analog input. (I do this here for my six PV arrays, each nominally 100V/6A).
You could also use a simple hall current sensor. They usually have an output at 1/2 rail, and go up or down depending on the direction and magnitude of current flow, and are completely isolated from the line you're measuring.
 
Yep, have to scrap the plan of using the WC to measure A/I's on different grounds. Even the differential ground PV-Bat calculation. Doing this makes the A/I's even on the same ground have some inaccuracy's. Maybe ground loops, noise, etc... To complex to figure it out.   Will just monitor the battery voltage and if at a certain low level assume that its not charging and its dark or cloudy.
 
Machineman said:
Yep, have to scrap the plan of using the WC to measure A/I's on different grounds. Even the differential ground PV-Bat calculation. Doing this makes the A/I's even on the same ground have some inaccuracy's. Maybe ground loops, noise, etc... To complex to figure it out.   Will just monitor the battery voltage and if at a certain low level assume that its not charging and its dark or cloudy.
 
Check http://www.circuitstoday.com/differential-amplifier
About 1/3 down the page "Fig 1" is a single op-amp differential amplifier circuit.
The common-mode noise/error in this arrangement is negligable and does exactly what you want.
 
what about measuring charge current?  this sensor can work very well
 
 
Code:
http://www.ebay.com/itm/131376978042
 
CAI_Support said:
what about measuring charge current?  this sensor can work very well

http://www.ebay.com/itm/131376978042
 
Yup, that what I was suggesting above in post #6 when I said "You could also use a simple hall current sensor. They usually have an output at 1/2 rail, and go up or down depending on the direction and magnitude of current flow, and are completely isolated from the line you're measuring."
 
They are a little noisy though but with filtering and/or oversampling they're not bad.
 
may be it is good to run the WC over an induction coil, having galvanic separation ;-)
No this is not nonsense as we see wireless charging appear, but we would need wireless data connection making that dream real.
 
Back
Top