You don’t get it Bob.
You can only measure between ground and 3.3 volts. Put simply the LEM scheme means that negative currents are measured at ground to 3.3/2 and the positive currents between 3.3/2 and 3.3 volts.
1 Like
I do get it. That is how all these Hall effect bi-directional sensors work.
I was only providing a simple solution to turning negative numbers positive which is what I thought you were offering to do. No more to say really.
Cheers Bob
1 Like
With the LEM sensors you need to measure the reference voltage and the input voltage (both positive) and use these to calculate the current. Of course you need to know whether you are charging or discharging batteries so you respect negative values.
Regards
John
1 Like
John. I CAN see where you are coming from and I thoroughly agree that passing the supply cable through a device when dealing with high currents is a good idea. At these sort of currents minimising the connectors, board tracks and soldered joints gets rid of a number of potential disastrous problems.
Getting back to basics though, this post started out as a placeholder for a product stocked by Core. Bob enquired about the use of this product in his caravan set up. I don’t know anything about his exact set up but I imagine he may have a modest solar system supplementing his vehicle and 240V mains charging arrangements. I would think his worst case discharging scenario would be if he had a 240VAC compressor type fridge driven with a 12VDC to 240VAC inverter. When the compressor starts up the inrush current could be quite high and may approach 50A.
The subsequent replies have revolved around this sensor or similar and I think at these current levels the connection system would be OK IF DONE PROPERLY. I have tried to keep things simple for him hence my suggestion to use the 50A unit as it has a very convenient 40mV per Amp output which would simplify the number crunching to get a direct current read out. The LEM type sensor would be mandatory at high current levels and indeed as you say there are lower current devices available which would do the job.
But. Let us get back to the original. Core letting the community know this product is available and publishing some details.
Cheers Bob
3 Likes
It’s not that simple. The 5v supply for the Arduino can vary by up to 10%, which means that if it is used as the ADC reference then allowance has to be made for that variation. That can be done by calculating a new voltage range for the ADC range, a new midpoint (offset) and a new mV per ADC unit. But the Arduino cannot measure its own supply voltage without an external reference. It could use the sensor at zero current to calibrate, but that wouldn’t be applicable in this situation. An alternative is to supply a different reference voltage for the ADC. A zener reference could be used for a 4.7V reference (it will work with very low headroom), which means it is possible to calculate the correct ADC range, midpoint and mV from that reference, but the calculation is still required. So the idea of a ‘convenient’ 125mv or 40mV unit does not exist. The best way to consider the issue of resolution is to simply consider the counts available from the ADC. Whatever implementation is eventually implemented will benefit from any improvement in that count.
1 Like
Hi Jeff
What if the same Arduino 5V was used for the sensor V input or use the same sensor V input (5V) for the Arduino reference (in case the Arduino has not enough grunt but I believe it would). That way even if this varied a bit at zero A the sensor output should be Vin/2 and the ADC should be half maximum whatever that may be. That should pretty much get rid of the problem of two different references.
The next question is how much resolution and absolute accuracy do you want or need. Do you want or need something super accurate (laboratory ???) or just to monitor the health of the power system in a Caravan. Which is where this thread started.
With this statement in mind I have tried in my replies to keep things simple (the KISS principle) for Bob to give him some chance at success. I thought that for this application a resolution of 125mA and an accuracy of ± the same would be adequate but worst case. I agree that a couple of more bits in the ADC would be better but I have no idea what he intends to do about this ADC. If using an RPI he will have to provide this so it can be chosen to suit. I know very little of RPi so have been basing my replies on Arduino to be converted at will.
Cheers Bob
1 Like
[quote=“Robert93820, post:26, topic:11249”]
What if the same Arduino 5V was used for the sensor V input or use the same sensor V input (5V) for the Arduino reference[/quote]
That makes no difference. You have an ADC returning values from 0 to 1023 representing an unknown voltage. You know that 511 is half that voltage. If you assume that the sensor and the ADC reference are the same voltage (and that might not be easy to achieve) then you know that a reading of 511 should be zero current. But since you don’t know the absolute voltage represented by 1 step in the ADC then you can’t convert a divergence in ADC reading from that mid point into a voltage and that means you can’t convert it into a current. I would guess that you have never constructed one of these devices - it is complex.
Does it matter? I don’t know, but the point is that by attempting to explain how an ADC count becomes an amperage value you have not simplified things, but have created a significant complication that is probably irrelevant the the OP and which does not justify the assertion that my original statement was incorrect. That’s the point that you seem unwilling to see.
1 Like
If they are from the same source I don’t see why they wouldn’t be the same.
No I haven’t. But I would guess that over the years working with some pretty smart people in a company that used to design the things they made I would expect to find a parallel of sorts somewhere.
Actually I intend to construct a current/voltage measuring device in the near future as I can see a use for such a thing. If I can keep it simple.
Can be as complex as you want to make it.
I AM willing to see any point. What point. NO ONE has come up with is a specification (resolution, accuracy etc) yet and the whole thread has been going in circles getting ever more complex when there is no goal to work toward. Maybe WHEN someone DOES provide a specification it may be found the original subject device will not meet requirements. Then go no further. It will not be able to be done with that device.
I just have one more thing to say here. Think about the end result you need. Look at the weakest link in the chain. If that won’t meet the criteria you will have trouble meeting the end goal. Go back and think again.
I feel I have nothing more to add here so on that note I will respectfully bow out and wish all success.
Cheers Bob
1 Like
Hi Liam
I’m a solar engineer so to give you a bit of an idea. I know this thread has been about caravan systems, but as for residential solar systems here are some numbers that are typical for Australian systems. On the DC side, nothing will go above 600V in Australia (that limit is set by the Building Code, so to go above that the property cannot be classed as residential). As for current, these days a decent residential solar system will often consist of only a single string per inverter input (residential inverters have one or two inputs) but will occasionally have two strings in parallel per input. It is very rare to have more than 2 parallel strings as it requires reverse current protection across the array, as the built in reverse polarity rating of the modules is often around 20A. Most 60-cell PV modules coming into the Australian market have MPP currents around the 9-11A mark, with short circuit (maximum) current around 11-12A. So for DC sensing, you need up to 24A current rating on your sensor per input string. And up to 600V for your voltage sensor.
I mentioned 60-cell above. A lot of new modules coming into the market are 72-cell (typical for commercial applications), or 96 cell or even 120 cell half cut modules. Which typically have higher current values.
On the AC side, a significant number of the inverters we install around Newcastle are 5kW, so 22A is what you would normally see out of them. That being said there are a lot more 8-10kW systems at the moment, either single or three phase (so 36.5-45A for single phase or 13-15A per phase for 3 phase).
I know none of this is relevant to the original question, but it may be relevant to future projects for a number of other users. Possibly including me 
3 Likes
Hey Luke,
Yikers, thats some serious juice. I was way off from <50W mark I was thinking about! And and engineered solution would be much more robust than a maker or hobbyist one.
How have you been going @Bob147385 although the topic has moved quite a bit we’re still all ears and would be keen to see your implementation! 
2 Likes
I have done a bit of research, most of which goes over my head, and ordered a few parts that are yet to arrive, to do a few experiments. There seems to be quite a few differences of opinion about it all. I will let you know what I come up with.
Bob
3 Likes
