When is a shunt needed exactly?

Hi,

I was reading up in my Outback FM-80 user's manual (preparing for the install!) and on page 13 is pretty much the same setup as I will be using and the schematic shows a shunt being used. This is an off-grid system and here are the specs:

-6 Kyocera KD185GX-LPU wired in series/parallell (3 strings of 2 panels).
-Outback FM-80 charge controller.
-448 A/h Concorde Sun Xtender AGM 12 volt battery bank.
**(Eventually, an automatic-starting propane powered generator will "back-up" the cabin if this would be necessary, as in recharging batteries if system is down, or if battery bank voltage is too low, etc..)


I would like to know more about what a shunt does exactly and how to select the proper one for my needs.

Thank you!

Comments

  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: When is a shunt needed exactly?
    I would like to know more about what a shunt does exactly and how to select the proper one for my needs.

    A shunt is a very low resistance resister. When current runs through it, a small voltage develops across it (ohm's law). A voltmeter can measure that voltage and you can then calculate the current flow. Most battery monitors use a shunt to make their calculations about battery state of charge.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • CATraveler
    CATraveler Solar Expert Posts: 98 ✭✭✭✭
    Re: When is a shunt needed exactly?

    A shunt is a precision resistor that accurately measures the voltage drop across it and hence current flow through it. It needs to be sized for your maximum current draw. My monitor choices were 100A and 500A and I could possibly draw up to 200A so the choice was obvious. The voltage drop across the shunt is low and in my case it's 0.050V (50 mV) for the 500A shunt (with a 500A draw).

    A 50A draw would be 5mV drop across the shunt which is a small loss and good tradeoff to provide the monitor capabilities.
  • Mr.Cockburn
    Mr.Cockburn Solar Expert Posts: 34
    Re: When is a shunt needed exactly?

    Thank you both for your replies,

    My "theoretical" max load would be 210 amps, as the combined breaker ratings of my 2 breaker boxes is 210 amps. (All DC loads)

    -What gauge of wiring is required for the battery/shunt connection and the shunt/charge controller connection sides?

    -Is my FM-80 "pre-programmed" to understand the readings from this shunt, or do I need something else for the readings to be understood by my controller?

    -Considering my max. theoretical load of 210 amps, would this shunt do the trick? http://www.solar-electric.com/mkb-500-50.html

    Also, I noticed CATraveler mentionning that he has 2 different shunts, in what particular case 2 different shunts would be needed?
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: When is a shunt needed exactly?
    vtmaps wrote: »
    A shunt is a very low resistance resister. When current runs through it, a small voltage develops across it (ohm's law). A voltmeter can measure that voltage and you can then calculate the current flow. Most battery monitors use a shunt to make their calculations about battery state of charge.

    --vtMaps

    The other alternative to a shunt would be a non-contact device such as a Hall Effect transducer which calculates the current from the magnetic field. Although such devices will handle DC with good precision, there is a problem with zero drift and position sensitivity which makes them unsuitable for this sort of monitoring.

    Devices like inverters or CCs which have a current output display have an internal shunt, but its output is not readily available for use by other devices like battery condition meters.
    SMA SB 3000, old BP panels.
  • john p
    john p Solar Expert Posts: 814 ✭✭✭
    Re: When is a shunt needed exactly?

    The shunt you have noted in link would be correct for your application. All you need to connect accross the shunt is a millivolt meter.. Using that shunt 1mv will equal 10a. Its claim it can be used down to .1a is just about impossible. As that would give a reading on the millivolt meter of .001mv . No reasonable priced millivolt meter can reads anything like that low. Its praxctixcal useis from 10a. But thats ok for you.
    Its never a bad idea to use shunts between everythng. IE between panels and battery/ies between battery/ies and inverter.
    As others on this forum will know I believe the shunt is the most valuable item to show you how your system is performing and after a while you will get to know what readings are correct ansd if some loss is occuring in that part of the system.
    They also act as fuses for gross overloads. BUT some will tell you that is not what you should use. But I really recommend them for that but will leave that up to you and others .
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: When is a shunt needed exactly?

    The FM80 does not need nor make use of a shunt.
    Outback's Flexnet DC system monitor does. http://www.solar-electric.com/oupoflsymo.html

    Sure you haven't got the two confused? Or possibly looking at a diagram that includes both?

    The 210 Amps you list as your total DC current requirements is from batteries to loads, and as such does not involve the FM80. It is also a lot of current. If this is being handled on one set of wires they will need to be very large indeed: 3/0 at least.

    I think we need to get the component terminology straight first.
  • Mr.Cockburn
    Mr.Cockburn Solar Expert Posts: 34
    Re: When is a shunt needed exactly?
    The FM80 does not need nor make use of a shunt.
    Outback's Flexnet DC system monitor does.

    Sure you haven't got the two confused? Or possibly looking at a diagram that includes both?

    LOOK at page 18, what's that I see? A SHUNT http://www.outbackpower.com/docman/0307121010258900-0009-01-00RevB.pdf
    The 210 Amps you list as your total DC current requirements is from batteries to loads, and as such does not involve the FM80. It is also a lot of current. If this is being handled on one set of wires they will need to be very large indeed: 3/0 at least.

    I have 2 breaker boxes, 1 is 150 amps MAX, the other is 60 amps MAX.
    -The 150 amp breaker has 2 AWG wire for a maximum of 10 feet (one way) from battery to breaker box.
    -The 60 amp breaker box has 4 AWG wire for a maximum of 10 feet (one way) from battery to breaker box. **These are "theoretical" loads, actual loads are much, much less.**

    From the Blue Sea systems site:
    2 AWG wire max ampacity: 210 amps
    4 AWG wire max ampacity: 160 amps
    I think we need to get the component terminology straight first.

    There is absolutely NO NEED to be condescending, I am polite with people and expect NOTHING LESS from them. I'm just trying to LEARN (just like everybody on this forum) and will never pretend to know everything about anything. Thank you.
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    Re: When is a shunt needed exactly?
    LOOK at page 18, what's that I see? A SHUNT

    That shunt is not doing anything in that drawing since it is not hooked up to a battery monitor such as a FNDC. The FM-80 and inverter will not make use of a shunt without a battery monitor.
    There is absolutely NO NEED to be condescending, I am polite with people and expect NOTHING LESS from them. I'm just trying to LEARN (just like everybody on this forum) and will never pretend to know everything about anything. Thank you.


    ??? :confused: I didn't see anything remotely condescending in 'coots post. Chill.
  • Mr.Cockburn
    Mr.Cockburn Solar Expert Posts: 34
    Re: When is a shunt needed exactly?
    mtdoc wrote: »
    That shunt is not doing anything in that drawing since it is not hooked up to a battery monitor such as a FNDC. The FM-80 and inverter will not make use of a shunt without a battery monitor.

    I guess Outback's reasoning for putting it in the drawing is for future expansion (FNDC+MATE add-ons)... I will put it in for the same reasons (future expansion).

    ***The shunt I posted previously would be ok for my use, right? What kind of gauge wire is needed from shunt to charge controller?? (About 10 feet apart).
    mtdoc wrote: »
    ??? :confused: I didn't see any remotely condescending in 'coots post. Chill.

    I'm sorry if I misinterpreted the way Cariboocoot wrote his reply, it's just that the way I read it was like: "go do your homework, you're wrong on just about everything"!! I will make an effort not look into comments like that from now on. Thanks.
  • john p
    john p Solar Expert Posts: 814 ✭✭✭
    Re: When is a shunt needed exactly?

    ***The shunt I posted previously would be ok for my use, right?
    Read post #6
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: When is a shunt needed exactly?
    I guess Outback's reasoning for putting it in the drawing is for future expansion (FNDC+MATE add-ons)... I will put it in for the same reasons (future expansion).

    ***The shunt I posted previously would be ok for my use, right? What kind of gauge wire is needed from shunt to charge controller?? (About 10 feet apart).



    I'm sorry if I misinterpreted the way Cariboocoot wrote his reply, it's just that the way I read it was like: "go do your homework, you're wrong on just about everything"!! I will make an effort not look into comments like that from now on. Thanks.

    It was more of a "if we are using the same words to describe two different things, we will be talking past each other a lot" IMHO.

    The wiring to and from the shunt (current) will be exactly the same as the wiring you would have used without the shunt there at all. (For the same total wire length, that is.)
    The wiring from the voltage contacts of the shunt to the measuring device, whatever it is, will just be small gauge wire, preferably twisted pair to keep the noise down.
    SMA SB 3000, old BP panels.
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    Re: When is a shunt needed exactly?
    I guess Outback's reasoning for putting it in the drawing is for future expansion (FNDC+MATE add-ons)... I will put it in for the same reasons (future expansion).

    Yep, I think that's why it's there. Good idea to put one in. A FNDC is definitely worth adding. The sooner the better.

    I'm sorry if I misinterpreted the way Cariboocoot wrote his reply, it's just that the way I read it was like: "go do your homework, you're wrong on just about everything"!! I will make an effort not look into comments like that from now on. Thanks.

    Ok cool.:cool:
  • CATraveler
    CATraveler Solar Expert Posts: 98 ✭✭✭✭
    Re: When is a shunt needed exactly?

    Without a battery monitor then just connect the wires to the negative bus bar. Apparently the doc assumes you'll have a monitor which can be very useful. Perhaps the doc should be updated to clarify the shunt.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: When is a shunt needed exactly?

    Wiring diagrams for this stuff can be very confusing. Take the cited example from Page 18: it shows a shunt, and absolutely no component that makes use of it. But it does show an FM80. Result: confusing the OP into thinking the shunt has something to do with the charge controller.

    There are lots and lots of diagrams for Outback equipment. Some of them include the Flexnet monitor and shunt, and an FM80 (or two). It's just too easy to look at this stuff and either think or write the wrong thing.
  • Mr.Cockburn
    Mr.Cockburn Solar Expert Posts: 34
    Re: When is a shunt needed exactly?
    john p wrote: »
    ***The shunt I posted previously would be ok for my use, right?
    Read post #6

    Thanks john p, your explanation was very clear, I just asked again (since this was another person) to get a second opinion, if you will.

    Also, thank you all for informing me on the uses of a shunt. I will be including one to my system for the future addition of a FNDC+Mate as soon as possible.
  • nsaspook
    nsaspook Solar Expert Posts: 396 ✭✭✭
    Re: When is a shunt needed exactly?
    inetdog wrote: »
    The other alternative to a shunt would be a non-contact device such as a Hall Effect transducer which calculates the current from the magnetic field. Although such devices will handle DC with good precision, there is a problem with zero drift and position sensitivity which makes them unsuitable for this sort of monitoring.

    A small point, most of the problems you describe are properties of open-loop hall sensors, closed-loop hall/mr sensors can be as good as most non-lab shunts (Linearity better than 0.1%).
    http://www.hallsensors.de/ClosedLoopCurrent.pdf
    http://www51.honeywell.com/aero/common/documents/myaerospacecatalog-documents/Defense_Brochures-documents/Magnetic__Literature_Application_notes-documents/AN209_Magnetic_Current_Sensing.pdf
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    Re: When is a shunt needed exactly?
    nsaspook wrote: »
    A small point, most of the problems you describe are properties of open-loop hall sensors, closed-loop hall sensors can be as good as most non-lab shunts.
    http://www.hallsensors.de/ClosedLoopCurrent.pdf

    Thanks for clarifying that.

    Somewhat off topic - but I've been thinking of getting a couple of THESE closed-loop hall effect sensor-transducers and an inexpensive DAQ board just to play around with measurement of some circuits and maybe eventually programming some control functions

    You can also get these sensors prewired with a voltage regulator HERE - more money but would simplify set up.
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: When is a shunt needed exactly?
    nsaspook wrote: »
    A small point, most of the problems you describe are properties of open-loop hall sensors, closed-loop hall/mr sensors can be as good as most non-lab shunts (Linearity better than 0.1%).
    http://www.hallsensors.de/ClosedLoopCurrent.pdf
    http://www51.honeywell.com/aero/common/documents/myaerospacecatalog-documents/Defense_Brochures-documents/Magnetic__Literature_Application_notes-documents/AN209_Magnetic_Current_Sensing.pdf

    Thanks for the reference. Roughly what are the costs for a closed loop sensor (normalized to the price of a good but not lab-quality shunt)?
    Even a closed loop sensor needs to be shielded from or otherwise ignore/balance out the earth's magnetic field, right? How is this typically done? I know that this is a reported problem for typical (open loop?) clamp-on DC meters.
    SMA SB 3000, old BP panels.
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: When is a shunt needed exactly?
    mtdoc wrote: »
    Thanks for clarifying that.

    Somewhat off topic - but I've been thinking of getting a couple of THESE closed-loop hall effect sensor-transducers and an inexpensive DAQ board just to play around with measurement of some circuits and maybe eventually programming some control functions

    You can also get these sensors prewired with a voltage regulator HERE - more money but would simplify set up.

    Your first link points to an open-loop sensor. The Honeywell CSN series (also available through Mouser) seem to range from about twice that price on up and of course need an external voltage/power source.
    The specifications on your second link, the integrated package, do not mention open versus closed loop, so I would assume that it is actually open loop unless I missed something or you have independent information. :-)
    SMA SB 3000, old BP panels.
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    Re: When is a shunt needed exactly?

    Oops. You're right! I was looking at the picture and assumed it was the closed loop type...:roll:

    The premounted unit in the second link uses the same sensor I believe.

    From looking at the data sheet and this application note from Honeywell it certainly looks like they would be reasonable choices for measuring dc circuits in an RE system or household AC circuits.

    How much of a problem would drift and position sensitivity really be in this kind of application? I don't need lab-grade accuracy.

    I assume the TED and similar systems use these lower cost open-loop type sensors.
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: When is a shunt needed exactly?
    mtdoc wrote: »
    Oops. You're right! I was looking at the picture and assumed it was the closed loop type...:roll:

    The premounted unit in the second link uses the same sensor I believe.

    From looking at the data sheet and this application note from Honeywell it certainly looks like they would be reasonable choices for measuring dc circuits in an RE system or household AC circuits.

    How much of a problem would drift and position sensitivity really be in this kind of application? I don't need lab-grade accuracy.

    I assume the TED and similar systems use these lower cost open-loop type sensors.

    When the sensor core is fixed rather than clamp-on and the sensor does not get moved relative to the earth's magnetic field, zero shift and drift will be less of a problem. Whether you could still get away with it for a battery meter or other integrating application is still doubtful. Unless you check and adjust the zero current output from time to time.
    SMA SB 3000, old BP panels.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: When is a shunt needed exactly?

    As far as I know, the TED and equivalent are using CT (current) transformers. So, no DC measurement capabilities needed when using CT's.

    I believe, the output of the CT would go to a precision shunt/resistor and the electronics would measure the voltage drop across the shunt (Ishunt=Vshunt/Rshunt).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • nsaspook
    nsaspook Solar Expert Posts: 396 ✭✭✭
    Re: When is a shunt needed exactly?
    inetdog wrote: »
    Thanks for the reference. Roughly what are the costs for a closed loop sensor (normalized to the price of a good but not lab-quality shunt)?
    Even a closed loop sensor needs to be shielded from or otherwise ignore/balance out the earth's magnetic field, right? How is this typically done? I know that this is a reported problem for typical (open loop?) clamp-on DC meters.

    I've used several of the open and closed loop models of the low cost Ampsense devices in my DIY Solar BMS. My software rechecks the PV input sensor zero (using a PV input disconnect relay) at the start of any charging cycle to check for zero drift and updates a internal drift correction table. The battery inverter sensor zero requires a manual switch disconnect to check but once it's done I find the device drift is very close to the published spec.
    Models and price list http://www.ampsense.com/
    http://www.ampsense.com/HANDBOOK%20web.pdf

    A fixed position open-loop sensor connected to a microcontroller 10 or 12bit ADC with a simple software zero offset and termistor for drift/gain correction keeps current measurements stable and accurate if the sensor has a flux concentration toroidal core (most do) to shield it from external fields. Closed loop designs use a flux feedback core so they normally don't need external shielding from low level fields.
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    Re: When is a shunt needed exactly?
    BB. wrote: »
    As far as I know, the TED and equivalent are using CT (current) transformers. So, no DC measurement capabilities needed when using CT's.

    Pardon my ignorance, but I guess I thought CTs were a type of Hall effect current sensor. :blush:
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    Re: When is a shunt needed exactly?
    nsaspook wrote: »
    A fixed position open-loop sensor connected to a microcontroller 10 or 12bit ADC with a simple software zero offset and termistor for drift/gain correction keeps current measurements stable and accurate if the sensor has a flux concentration toroidal core (most do) to shield it from external fields. Closed loop designs use a flux feedback core so they normally don't need external shielding from low level fields.

    Good information. Thanks.

    Well, I've just purchased this NI USB DAQ device. I've got an old version of LabView around somewhere so I'll probably play around with a current sensor sometime soon. Years ago when I ran a neurobiology teaching lab I did quite a bit of programming with LabView based data acquisition and processing. Hopefully it will come back to me... a fun winter project in any case.

    Looks like I have some learning to do about drift/gain correction with these current sensors.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: When is a shunt needed exactly?

    A CT is a "current transformer"... For a single run of wire through the donut, a second set of windings pick up the magnetic field with the Current equal to:

    CT current = wire current * 1/N turns of sense windings

    Note that these are "true" current transformers. If you do not have a load on the output of the CT transformer, it can develop hundreds to thousands of volts because of the "infinite resistance" of the open circuit--shocking the user and/or ruining the CT assembly :

    http://en.wikipedia.org/wiki/Current_transformer

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • nsaspook
    nsaspook Solar Expert Posts: 396 ✭✭✭
    Re: When is a shunt needed exactly?
    mtdoc wrote: »
    Good information. Thanks.

    Well, I've just purchased this NI USB DAQ device.
    Looks like I have some learning to do about drift/gain correction with these current sensors.

    That card with Labview should make a very nice system. I haven't used Labview in ages but the NI DAQ hardware is top notch.
    I wrote the driver for the NI DAQCard-700 pcmcia card in the current Linux kernel and built a simple Linux based solar monitor interface for it.
    https://raw.github.com/nsaspook/nidaq700/master/solar_daq_interface.jpg
    https://github.com/nsaspook/nidaq700/blob/master/8030740020_90f46dfb77_c.jpg
    https://github.com/nsaspook/nidaq700/blob/master/8030739924_1e2ae346c7_c.jpg

    http://tomoyo.sourceforge.jp/cgi-bin/lxr/source/drivers/staging/comedi/drivers/ni_daq_700.c