Paul is correct.
Really you just want to connect the IN- and IN+ to the battery + and - terminals for the most accurate voltage readings.
1 Like
Great. I’ll try it tonight. Here is my ultimate goal… To monitor the performance of a solar panel driving a load. Here’s my proposed final circuit.
(incorrect sketch removed to avoid future confusion)
That should work.
I did the same thing using the INA219 chip to measure a solar panels voltage and current.
Well, I connected it per my last diagram above, and I’ve got problems. The wire to the in- terminal got hot and melted. Here’s the data that was recorded (briefly) during this event.
| date | current_a | bus_voltage_v | shunt_voltage_v | buspower_w | comment |
|---|---|---|---|---|---|
| 2018-08-15 18:05:58 | 0.0044 | 0 | 0.000035 | 0 | with in+ and in- not connected |
| 2018-08-15 18:04:53 | 7.9948 | 0 | 0.063985 | 0.02 | with in+ and in- connected (melted wire) |
It seems like the board still works, but I have fundamental lack of understanding about how this breakout board should be connected, and I can’t seem to find any good documentation (in english) about how to properly wire this thing. Thanks so much for the help so far, but I’m still open to suggestions.
See if this helps any:
Ok. This tutorial is recommending a complete opposite approach. Everything I’v read has recommended placing the INA226 on the HIGH side of the circuit. This video has it placed on the LOW side. I’ll need to work on this one for a few days I think. Thanks for the tip.
Well, I tried rewiring to match the diagram above with the High-Side shunt. Now ANY function call to the INA.cpp library hangs the program. I’m guessing that one of the wiring iterations I went through damaged the board. Even if I disconnect the load power supply, the program just hangs at the function call such as ina.readShuntCurrent().
Sigh.
I’m off to order a replacement board. I’ll be back in a few days or weeks, depending on the speed of eBay
Been there and done that one quite a few times now 
1 Like
undergrounder996, the INA226 is an excellent choice. I use it in a portable medical instrument and get very stable, accurate power info. The diagram you’ve shown above has one missing connection: the INA226 does not read the bus voltage from the V+ pin… You must connect the VBS to source (+). In the above pic, just jumper the PCB terminal VBS to the V+ terminal.
1 Like
Well, I received my new eBay INA226 breakout board. I connected it up, and ran the Vbus connection on the INA226 to the high side of my battery. Everything worked fine with my 9V battery test circuit. Good readings.
Then, I added an input filtering circuit per the INA226 Datasheet page 14, figure 21. (a few resistors and a capacitor, because I planning to monitor a DC motor at some point)
Tested again with my 9V battery test circuit. Everything was working properly.
Then, I connected up to my “real” circuit’s 24VDC power supply. I hadn’t even connected the load (motor) yet, and the board went unresponsive. It seems fried, just like the previous one. Ugh.
At this point, I’m considering this problem unsolved and I’m moving on. My plan is to put in a Hall Effect Current Sensor (ACS712), and just live without voltage measurement for now.
Have you ever solved this problem? I just ordered two INA226 boards and am afraid to fry these straight away! I want to hook them up pretty identical to yours.
Any tips to avoid frying it and having a bad day?
From what I know, you may isolate both circuits, but I’m not 100% sure. I read that somewhere about the INA219 and I think the INA226 might be pretty similar.
That said, the circuit should look like this:
So yes, you would need to galvanically isolate your Microcontroller circuit from the load circuit. Why? I don’t know. Chances are it’s not needed (or is it?) Why did it fry the circuit in the previous post? Also, when you galvanically isolate it, you won’t have the GND reference anymore and it probably can’t measure volage anymore.
Funny thing is, if it proves to work, then the datasheet is likely wrong and rendered quite an amount of frustration and fried boards.
I’ll update as soon as I get them and hook them up. However, if I forget to update, consider it working successfully (or I’m just dead lol jk)
If I understood this wrong, please correct me and propose a better answer!
@RWB Do you have a solution?
No, I never solved this. If I were to try, I’d skip this board and try to find a hall effect sensor that works with the Particle. Seems simpler with built-in isolation.
I never used one of these current sensors although I do have one or two around here.
I always used the INA219 or the 3 input version and adjusted the shunt size for what ever current I needed up to like 15 amps.
Too late! Already ordered the boards... Well, wish me luck, I guess?
I ordered them too! But they're for at most 26V, I need 32V@3A.. Does decreasing the get me to that value?
So you need to measure a supply up to 32v and 3A max or what exactly?
Exactly. I have a bench power supply on which the LCDs have failed. I’m trying to hook up an Arduino with an INA2xx chip and a 144x32 LCD to measure and display the values.
The onboard shunt is basically a copper wire turned around a couple of times, and measures 00.2 on my multimeter (not even sure if it’s just the cable, my multimeter isn’t that good). I think it might be equal or under 0.1Ohms.
(H4 measures between it)
It also has a separate ±15VAC output just for the meters, it seems galvantically isolated. Not sure if needed, but I’ll mention it just in case.
The INA boards should arrive in a couple of days, but I’m really afraid to fry them, considering the replies here.
The INA boards are fine as long as you operate them within spec.
I’ve used one to measure around 10 amps of current successfully.
If you need to go over 26v then I don’t see why we can’t figure out how to get the Hall Effect Current sensors to work. From what I understand is that they provide a voltage output linear to the amount of current flowing through the shunt resistor.
1 Like