next topic »

[on:]

Disclaimer: I'm still learning about electronics so forgive me for anything that sounds completely stupid.

My project is pretty simple. I have a single 48V DC battery as the power source, connected to that is a 48V -> 12V DC converter that will power the Arduino. This DC converter is isolated (no continuity between in & out ground pins). However, I want the Arduino to monitor the voltage of the battery, so I have the 48V run through a voltage divider and wire the output of that to one of the analog input pins. As I've found just hooking up the positive to the analog input doesn't work and produced values all over the place. This makes sense as I've read the analog input pins measure with reference to ground, so I run a wire from the ground of the voltage divider to the ground pin on the Arduino.

However, isn't this effectively bypassing the ground isolation of the DC converter? And if so could this cause problems / fry something?
What if I only had one of the grounds connected to the Arduino? Will I get inaccurate values from the analog input because of different grounds?

Thanks in advance.

[Reply #1 on:]

However, isn't this effectively bypassing the ground isolation of the DC converter? And if so could this cause problems / fry something?

Yes and yes. It depends upon what you're doing with the 48V source. What is your need for isolation in the first place?

Will I get inaccurate values from the analog input because of different grounds?

Yes, unless your voltage divider measurement is made relative to the same reference point (voltage is always relative!) as the Arduino's, you will get "all over the place" values. The easiest way to do so, as you've noted, is to connect the grounds together. This isn't always a bad approach, it just depends on why you need the two systems to be isolated.

If you absolutely need isolation you can use an isolating linear optocoupler as in this circuit:

http://ruggedcircuits.com/html/circuit__24.html

--
The Rugged Motor Driver: two H-bridges, more power than an L298, fully protected

[Reply #2 on:]

IF and it's a big IF, if the output side is truly isolated from the input side of the 48/12 convertor then you should be able to form a common ground between the 48 volts, 12 volts and the arduino.  Similarly you should be able to monitor the 48 volts supply provided you use say a 15:1 voltage divider chain across the 48 volt system.  15:1 since you will drop 15/16 of 48 volts (45 volts) leaving 1/16 (3 volts) to measure on the arduino.  You will need a bit of leaway since a 48 volt charged battery will be measuring somewhere in the order of 56 volts so the 3 volts could rise as high as 4.6 volts.

jack

[Reply #3 on:]

Thanks for the quick replies.

Yes and yes. It depends upon what you're doing with the 48V source. What is your need for isolation in the first place?

The 48V battery mainly powers a motor. As far as I know I don't need isolation, it just that the $5 easily available and otherwise perfect converter I bought happens to be isolated.

Jack - I've already worked out the divider to be ~20:1 since we want to use the same setup for 60V systems and possibly even a 72, so plenty of leeway.

So both of you agree that connecting the 2 grounds shouldn't be an issue so long as I don't need the isolation feature of the converter. Do you think running a 1000W motor off that battery would be a reason to need isolation?

[Reply #4 on:]

Do you think running a 1000W motor off that battery would be a reason to need isolation?

As long as you can keep the ground return path of the motor completely separate from that of the Arduino it should be OK, as the motor itself is an isolated device (unless you're earthing it??)  It may not pass electrical codes

In other words, electrons going to and from the motor should never travel on the same wire as electrons going to and from the Arduino.

--
The Aussie Shield: breakout all 28 pins to quick-connect terminals

[Reply #5 on:]

Just to clarify for a beginner:-

In other words, electrons going to and from the motor should never travel on the same wire as electrons going to and from the Arduino.

This is true but it doesn't stop the wires themselves from being connected. It is just that the electrons should have no need to go to the arduino if they don't want to. And they don't because they will always take the shortest path round a circuit. Any curious electron could wander up the culdesac and take a peek but electrons by and large are stupid and lack imagination. (unlike protons)

[Reply #6 on:]

As long as you can keep the ground return path of the motor completely separate from that of the Arduino it should be OK, as the motor itself is an isolated device

Correct, the motor is run by a separate controller which I figured was isolated. Just to be sure I took the meter and did a continuity check between the ground wire going in to the motor from the controller and the ground wire coming off the battery - no continuity. So it should be good to go. Of course once I get everything hooked up I'll still check with the meter that I'm getting sane voltages before plugging in my $60 yellowjacket.

Thanks guys.

[Reply #7 on:]

What RuggedCircuits said was that the grounds had to be connected but the return path should be separate. A point I tried to make in my last post but you did not get it.

[Reply #8 on:]

No I got that. The return path is separate, the motor returns to the controller, the DC converter returns directly to the battery. Are you suggesting separate wires from the battery to the converter & arduino vs just tapping off the ones that run in to the controller? I was going to hook it up that way and I don't see why the electrons would venture back up the much smaller gauge wire to the converter & arduino when they can take the short path to the battery. Especially with the motor controller being isolated I don't even see how this is possible, but again see the disclaimer in the first post - I'm a newb.

[Reply #9 on:]

No I got that.

OK that's fine sorry.