Current Measurement?

[Dockhead]

I long ago retired my Victron battery monitor, which I was not able to wire up usefully after I combined my battery banks, and which was never much good anyway (I don't believe in the amps-counting method of battery monitoring).

I now have a Smartgauge which I will try to install this week.

But one thing I will miss about the Victron is measurement of current going in and out of the batts, so I think I will add some current measurement separately. WHat's the best way?

Why does the Victron use a shunt -- when hall effect sensors exist for DC? Could I use these?

There is no single point in my DC system where all the current goes in and out, so I will have to have TWO ammeters for the batts which I will have to add up in my mind -- UNLESS someone makes some kind of totalizer which will add up the figures for two different sensors? Anyone know of such a thing?

I think I would like to measure current in a few places, actually:

1. batts in and out

2. charger/inverter DC output/current draw

3. charger/inverter AC output (which will equal total AC power used since it all goes through there).

4. AC current drawn from shore power or generator.


This would give me a nice picture of what is going on in the electrical system.

[Paul Elliott]

I've purchased some cheap and fairly-accurate shunt/meter combos from Amazon (something like this: Amazon.com: Volt Meter/DC Digital LED Amp Volt Meter + Shunt /200V 200A /Blue: Automotive). This is a low-side sensor, so needs to be connected between the battery and ground.

As for Hall-effect sensors, I'm not so sure about the clamp-on style. My Fluke meter has a "zero" button and you really need to press that button to get any accuracy.

There are integrated Hall-effect current sensors that can connect to the high (or low) side -- any place in the loop actually. Here is the component: Allegro MicroSystems - ACS759 Thermally Enhanced, Fully Integrated, Hall Effect-Based Linear Current Sensor IC With 100 μΩ Current Conductor and Optimized Performance at 3.3 V. I don't know if anyone uses this in a packaged unit, so you might have to build your own. The specs are good.

[transmitterdan]

Dockhead,

There are available Hall effect current transducers. Most of these output 4-20mA current loop. Not easy to combine them to automatically sum their outputs as you want. But they are available. Also, there are 4-20mA panel meters that can use these Hall effect tranducers. But you won't find them in a marine chandlery. Check out http://files.redlion.net/filedepot_download/213/3824 for one example. This one sells on Digi-Key for about $350.00.

One thing to watch out for in Hall effect devices is not all of them will measure current in both directions. Some will and some will not.

A typical panel meter: http://www.murata-ps.com/data/meters/30pc_420.pdf

[Dockhead]

Thanks for all that.

Does anyone know whether two shunts wired in series will give the totalized current?

Surely there is some way to do this.

[Dockhead]

I wrote to this people:

Precision Current Shunts - Shunt | Ohm-Labs, Inc.

With the same question. I'll let you know what they say.

[transmitterdan]

Originally Posted by Dockhead

Thanks for all that.

Does anyone know whether two shunts wired in series will give the totalized current?

Surely there is some way to do this.

Dockhead,

It is possible to build a simple op-amp circuit that will sum two or more shunts. But you can't just wire the shunts in series. Basically the idea would be to take each shunt output and feed it to an instrumentation amplifier. Then take all the IA outputs and feed them to an op-amp summation circuit. This will create a single output that is the sum of all the current shunts. I suspect it is possible to buy something like this but haven't seen it. For audio applications these are widely available but probably they don't work at DC. It would be fairly trivial to home brew this for not much $$.

[undercutter]

More information required. What charger/inverter? What controller if any? How is shore/genset power fed to the inverter/charger? You say that there is no single location that all the current passes through. Does this mean you have more than one neg lead back to the batteries? I understood this to be poor practice from threads on this forum?

Many of the newer controllers will provide at least half of the info you want and the A/C side could be picked up in front of the main breakers for your panel depending on how A/C is presently fed from the inverter. My old heart system had one AC feed for the charger and a return feed from the Inverter. The new Mass Combi has 1 AC feed which is fed thru back to half the panel on a 2nd lead and a 3rd AC lead which is switched between the input AC and the inverter output. In my case I had to put 2 AC amp meters to see all of the draw. This controller also reads the charging current from my 200amp engine driven generator even though it doesn't control the charging output. Just add a bulkhead terminal and combine all neg leads back to the batteries and use this for the shunt installation. No expert here but I learn a lot this way.

[exMaggieDrum]

Originally Posted by Dockhead

I long ago retired my Victron battery monitor, which I was not able to wire up usefully after I combined my battery banks, and which was never much good anyway (I don't believe in the amps-counting method of battery monitoring).

I now have a Smartgauge which I will try to install this week.

But one thing I will miss about the Victron is measurement of current going in and out of the batts, so I think I will add some current measurement separately. WHat's the best way?

Why does the Victron use a shunt -- when hall effect sensors exist for DC? Could I use these?

There is no single point in my DC system where all the current goes in and out, so I will have to have TWO ammeters for the batts which I will have to add up in my mind -- UNLESS someone makes some kind of totalizer which will add up the figures for two different sensors? Anyone know of such a thing?

I think I would like to measure current in a few places, actually:

1. batts in and out

2. charger/inverter DC output/current draw

3. charger/inverter AC output (which will equal total AC power used since it all goes through there).

4. AC current drawn from shore power or generator.


This would give me a nice picture of what is going on in the electrical system.

I really, really have grown to love the wiring promulgated by ABYC where you have a shunt that is the last thing before your one and only one DC negative cable to your battery bank. A shunt there will give you the total "net" all all current in and out of the batteries. But it is also nice to have amps for your major battery charge sources (solar, alternator, charger, wind charger). Not required but nice. The shunts are not very expensive and don't involve any more wires than a Hall effect sensor. But you then need a meter for each one. You can easily wire one meter with a selector switch for those if you wanted to though.

I have seen many Hall effect sensors for AC circuits but have not run across them for DC for some reason. If they are accurate they should work too I suppose. There may be some reason that they are not widely used. I suspect there must be since installing a shunt requires big cable lugs to install and a good spot to put the shunt, etc.

One of the things I liked about some of the old alternator regulators is that they also had ammeter readings (requiring a shunt). My Outback solar regulator gives that too. Very handy.

I am pretty anal about KISS and your battery wiring sounds like it would give me heartburn. I am going to rewire mine this winter since the PO seemed to like complexity more than I do. But one of the aims is to be able to absolutely understand what is happening to each part of the system at all times.

[Rapanui]

A stand alone live data option would be to use DC Hall effect current sensors connected to either a PLC or something like a Rasperry Pi or Arduino. Then you sum the current into both banks or say totalise the amp.hours used by the fridge. What you can achieve is limited to where you are sensing and a bit of programming.
Another way of doing it would be to use the Hall sensors plugged into a PC data data acquisition unit. This will give you real time data and they will have software so programming is easy. If you do not need real time then a data logger which you periodically download to a PC usually with Excel compatible output so you can easily calculate whatever you want.
You can get the Hall sensors with virtually any type of output, accuracy and current capacity you want. A quick check on the web would suggest you could get something adequate for less than £10 each. PLCs, data acquisition or data loggers seem to be available from about £50.

[Dockhead]

Originally Posted by Rapanui

A stand alone live data option would be to use DC Hall effect current sensors connected to either a PLC or something like a Rasperry Pi or Arduino. Then you sum the current into both banks or say totalise the amp.hours used by the fridge. What you can achieve is limited to where you are sensing and a bit of programming.
Another way of doing it would be to use the Hall sensors plugged into a PC data data acquisition unit. This will give you real time data and they will have software so programming is easy. If you do not need real time then a data logger which you periodically download to a PC usually with Excel compatible output so you can easily calculate whatever you want.
You can get the Hall sensors with virtually any type of output, accuracy and current capacity you want. A quick check on the web would suggest you could get something adequate for less than £10 each. PLCs, data acquisition or data loggers seem to be available from about £50.

Thanks; really useful!

It would be quite a lot of trouble just for solving this one problem, but a small on-board computer handling this kind of data could be easily made into a broader function device which could have lots of uses. Feed it with the network data and you could make it a full fledged VDR, and you could also analyze historical wind and weather data. Small panel mount display and bluetooth keyboard/mouse. Sounds pretty cool actually. Hmmm.

[Dockhead]

Originally Posted by exMaggieDrum

I really, really have grown to love the wiring promulgated by ABYC where you have a shunt that is the last thing before your one and only one DC negative cable to your battery bank. A shunt there will give you the total "net" all all current in and out of the batteries. But it is also nice to have amps for your major battery charge sources (solar, alternator, charger, wind charger). Not required but nice. The shunts are not very expensive and don't involve any more wires than a Hall effect sensor. But you then need a meter for each one. You can easily wire one meter with a selector switch for those if you wanted to though.

I have seen many Hall effect sensors for AC circuits but have not run across them for DC for some reason. If they are accurate they should work too I suppose. There may be some reason that they are not widely used. I suspect there must be since installing a shunt requires big cable lugs to install and a good spot to put the shunt, etc.

One of the things I liked about some of the old alternator regulators is that they also had ammeter readings (requiring a shunt). My Outback solar regulator gives that too. Very handy.

I am pretty anal about KISS and your battery wiring sounds like it would give me heartburn. I am going to rewire mine this winter since the PO seemed to like complexity more than I do. But one of the aims is to be able to absolutely understand what is happening to each part of the system at all times.

KISS is a luxury sometimes. My boat was designed with separate battery banks for domestic power, and for heavy services like bow thruster, winches, and windlass. Four 12 volt batteries for each bank, in two battery boxes some distance from each other.

It was a correct decision to combine these two banks into one, but the tradeoff was this complexity in the architecture of the system.

[JayH]

Originally Posted by Dockhead

Thanks; really useful!

It would be quite a lot of trouble just for solving this one problem, but a small on-board computer handling this kind of data could be easily made into a broader function device which could have lots of uses. Feed it with the network data and you could make it a full fledged VDR, and you could also analyze historical wind and weather data. Small panel mount display and bluetooth keyboard/mouse. Sounds pretty cool actually. Hmmm.

You're getting pretty close to describing N2KView.

Unfortunately, I don't think they have a feature to perform custom calculations on the inputs. (i.e. adding amperages from multiple devices). I have long wished for it.

[Dockhead]

Originally Posted by JayH

You're getting pretty close to describing N2KView.

Unfortunately, I don't think they have a feature to perform custom calculations on the inputs. (i.e. adding amperages from multiple devices). I have long wished for it.

Indeed -- and I have looked lustfully at N2KView for some time, held back by the cost, but also because of several limitations in system, like the one you mention, and also lack of ability to deal with non-N2K data.

I think that for the cost of N2KView it could be worthwhile to cobble up a Rasberry PI -- but some programming would need to be done.

[smac999]

Ft nav vision allows you to display added values... but it'll cost you a few hundred grand.. had to commission one of those once. good times.



I think series shunts would work. but you'd have to test it. they are generally 0-50mv for 0-500a.


so say 100a through a shunt reads 10mv. if you have 2 shunts each with 100a at 2 different banks, you'd read 2 shunts in series at 20mv, which would say 200a on your gauge. but I guess if they are far away makes the twisted pair would be kinda weird as it's split up. and may get noise.


actually I don't think it would work. the ground would have to be perfect between the 2 banks. you are definatly going to have a bunch of MV drop on the cable between the 2 shunts so it would be wacky. you'd proably get a reading of like 100mv. 10 on each shunt and 80 on the wire between shunts.

[DeepFrz]

Two shunts in series will only record the same value. Current is the same through each shunt so you don't need to add the values. If the two shunts are wired in parallel of course you need to add the values if you want the total.

Does anyone know whether two shunts wired in series will give the totalized current?