Using hall effeft sensors instead of shunts for SBMS

[Chris R8]

Hi Dacian,

Can I use hall effect sensors instead of shunts to measure current?

Shunts create voltage drop and introduce heat into installation, both features you actually try to avoid also buy running bigger cables...

I figured during stress testing my newly top balanced bank that actually have the problem that my 700A shunt is allready getting quite hot at 300A and introducing a lot of heat into the plus busbar of the bank that connect the first 4p cells in the bank and transfer that to terminal of cell.

No chance and space to replace it with a 1000A victron shunt or put 2x 500A parallel. But a 500A hall sensor would easily fit and solve the issue. 

Chris

[Dacian Todea (electrodacus)]

Chris,

300A continues is at the limit what a 700A shunt will be capable of.

All shunts are rated to handle continues 66% of the rating so limit for the 700A shunt will be 466A

Not sure what shunt is that but a 50mV shunt will create less heat than a 75mV shunt and it will also be smaller because of that.

If your shunt is 75mV 700A then 32mV will drop on it at 300A and that means 9.6W lost as heat and it should be no problem if the shunt was connected with a 5 or 10cm of wire not directly to cell terminal.

But if the same 700A shunt was 50mV then you will have just 6.4W lost as heat.

Shunts will be orders of magnitude more accurate than a hall effect sensor. They usually have their own built in current amplifier that outputs 0 to 5V instead of what the SBMS0 needs 0 to 90mV

Of course it can be adapted but the accuracy of those is so bad that it will affect the SOC calculation fairly significantly.

As for energy lost as heat due to current shunt that is more than insignificant.

Say you have a 12V battery (worst case) and the shunt is 75mV so loss of 9.6W as heat at 300A vs 12V * 300A = 3600W as load that is an efficiency of 99.73% 

When you take loss on wires the battery connections at terminals and maybe inverter your overall efficiency will be at best 90% so 10% lost as heat vs just 0.26% loss on the shunt that becomes insignificant.

[Chris R8]

Well if finally an integration of SBMS into victron cerbo GX would be possible I could put a 1000A shunt on the negative busbar and via the DVCC feature the victron shunt is the source for all currents+SOC and charge source in the system. SoC Victron is equal to SBMS in measuring.

Still surprised how hot a 700A quality shunt gets at 300A load=42% load as you said 466A=66% is the limit continuous, it heats up the busbar to 55 degrees Celsius over time of 20min...not good at all.

It's 50mV 700A rideon shunt which is directly connected to the positive terminal of the bank which is a 30mmx7mm silver plated Cooper busbar for the fist cell pack of 12V 4S4p bank. Output of shunt is with a 10cm long 7mmx30mm silver plated cooper busbar connected to victron 500A 50mv PV shunt and a NH3 500A main load fuse. The luck is that all (fuses, shunts) besides the victron shunt have the same height of the connector, victron shunt just got a 4mm PVC board underneath to bring it to same height.

No chance at all to get a 1000A shunt installed, no space as it's longer and as also much higher. That's all optimised to keep length and number of contact points where up the total of 500A continues runs is as less and short as possible. 10cm Wire need to be a 250qm cable with crimped lugs for 500A cont is a nogo...it will also transfer the heat being that short...

Typically I am at 300A during cooking when induction stove and oven runs at the same time. At lunch 100A from solar keeps it lower at 200A load but dinner the full 300A kick in.

The silver plated busbar 7mmx30mm (with 99.8% cooper core) are from Eaton for grid stations and certified for 1500A, cannot get better busbars...victron lynx has 5mmx25mm tin plated standard cooper busbar and is rated for 1000A.

I have all loads on main shunt and all charge sources via PV shunt. Only exception is multiplus which is connected to main load side as its shorepower charger is very very rarely used at all. Main charge source is solar and backup 2/4x 80A alternator charge (2x115A alternator per engine, two engines) in 90% one engine running eg during motorsailing on 1400rpm consuming 1-1.5l per hour diesel for a long time for propulsion reason and I just harvest the extra energy as engines are also heavily oversized (instead 2x20hp I have 2x50hp) the 160A charge is peanuts for them.

[Dacian Todea (electrodacus)]

You have a 4S4P battery so ideal connection will be 4 short wires can be 15 or 20cm long if needed and each wire will see only 125A (500A total load) so wires could be thinner and more flexible.

So you will just need 4x #2 silicone insulated wires (the super flexible type) say 20cm long each connected to the 700A 50mV shunt.

This not only prevents the heat from the shunt getting to battery terminals but also properly splits the current since my guess is that your terminal to one of the cells gets hot because shunt is connected to just one terminal instead of all 4.

I only have around 100A max both for charge and discharge and I do that with #4 silicone wire  (it is a bit better than #4 as it is from China so it is 25mm^2 a bit more than 21mm^2 for a true #4 awg).

But #2 has 33m^2 likely 35mm^2 if you get flexible silicone insulated wire from China and that has 0.5mOhm/m so a 20cm piece will only have 0.1mOhm and the heat loss on that will be 1.5W so all 4 about 6W the same as the shunt but on a larger area and length from connections. The big advantage is the uniform current sharing to 4 cell terminals instead of just one. Also vibrations will be isolated from battery terminals by the 20cm long flexible #2 (35mm^2) wires.

The size of the engine is not important and that extra energy you extract from alternator to put in battery will cost you extra fuel.  Diesel can be fairly efficient maybe 20% and a liter of diesel is around 10kWh thus 2kWh per liter is available as mechanical power but alternators are as low as 50% efficient so only 1kWh per liter is typically available to battery.

So if you take say 80A from an alternator at 12V that is about 1kW and over the course of one hour 1kWh will cost you 1 liter of diesel. If you use both engines that is 2 liter to charge 2kWh in to the battery.

[Chris Sailor]

4 wires, crimps, lugs adds tons of resistance, thats noGo for 500A continious loads. Also stacking 4 lugs on the connection of the shunt noGo...

Also not allowed if i have a 500A main fuse and that one fuse also acts as the main disconnection switch as you can simply pull the NH fuses even under full load with just 12V. that is also requested by ISO installation laws in boats and RVs, most like house too but i don't know them.

thats the reason you use massive busbars with silver coating for minimal connection resistance and highest rating pure cooper above 300A loads, so a 210mm2 busbar can deliver 1500A safely. ANd the reason Eaton is producing them like this as nothing else gets you above 300A to 1500A rating which the installation laws for such high current environment request for certifying. thats the reason all that publicly available installtion material ends at 300A and you can hardly find anything above. that 7mmx30mm silver coated busbar from Eaton is 500Euro per m and only avaliable to buy by registered certified high current electrians or grid engineers.

Ever put 1000A cont on the VIctron Lynx dsitributor?? my buddy a grid engineer did that and it got so hot at 700A that the plastic casing starts to melt...its just 5mmx30mm tinned cooper standard busbar inside....its 500A cont and 1000A peak...he also supplied me the Eaton certified stuff, left overs of a big solar park project he just finished. he thinks i have a faulty shunt.

oh yes it has a lot to do because its the actual load thats on the engine and with 1400RPM motorsailing the actual load is like idling for a 4 cylinder 50hp which has excess power that would have been wasted otherwise but not a 20hp 3 cylinder one which will run at 1600RPM for less additional speed. my flow meter in the fuel line gives me the exact add on which is around 100-300ml per hour (depend on current of water, sea state and sailing angle) for full 160A charge. then around 30A with one alternator charging lead. my buddy has the 30hp 3 cylinder (same engine then 20 hp that just revs higher to acheive the 10 extra hp)) with exact same alternator and it needs with 0.8-1l significantly more. thats again theory versus real life where a lot factors play with.

[Chris Sailor]

the fuses and shunts are part of the battery case and screwed down so nothing put load, stress or vibration on the cell terminals. done that on purpose like this. the 6x70sqmm2 cables connected to the output of th 6xNH2 200A fuses (that have a 3cmx3cmx20cm ceramic body screwed down too.

2x70sqmm2 per multi, one must have loads and one all other loads. PV all charge sources are connected so PV shunt to NH2 250AH main charge fuse (screed down to with a 3cmx3cmx20cm ceramic body and via 70sqmm2 cable to main charge bus.

minus main busbar all 7x70sqmm2 cable connected, thats also screwed down.

so basically 7x positive and 7x negative 70sqmm2 cable leaving the battery case connected to the massive NH fuses bodies that are screwed down. The battery case consits out of 20mm thick PVC boards so even an external  impact won't penetrate the case and nothing will twist here even not 1/100mm.

its a world circum navigation cat and i already did sail in 8m ocean waves so that needs different protection levels then a stationary off grid installatiion.

[Dacian Todea (electrodacus)]

If your connection goes to a single cell positive terminal it is irrelevant that you have an oversized 210mm^2 busbar.

The other 3 cells will need to be supplied with current and they will not be getting equal current leading to imbalance and faster degradation of that first cell where connection is made.

What sort of connection do you have between the 4P cells ? as that first connection will nee to handle close to 75% of the current  and I'm fairly certain that is nowhere near 210mm^2

Since I do heating I do have connection points with over 300A for hours and for that I use busbars capable of 570A Bussmann PDBFS377 It has 12 connection points on one side so I can connect the negative common of my 36 x 255W panel PV array as panels are connected in groups of 3 in parallel.

It is not just the amount of copper section but also the surface area for cooling that is important.

There will be overall significantly less heat at all points using 4x 35mm^2      20cm long wires to each cell terminal than that single 210mm^2 bussbar connected to a single cell positive terminal.

I never used any Victron equipment but looking at the Victron Lynx spec https://www.victronenergy.com/media/pg/Lynx_Distributor/en/technical-specifications-lynx-distributor.html

I see that it is rated 1000A continues and they say the buss material is 8x30mm. I'm personally not a fan on crimped connectors and the steel screw terminals that they have inside the Lynx but if connections are done properly torqued to spec I will expect that will handle 1000A for minutes and likely around 750A continues in 35C ambient with bar not exceeding 90C and that should not create any problems with the plastic.

Now that I think about I did had a Victron BP65 but was just for the purpose of testing and tear down to see how it is constructed internally and I was not that impressed but the price is fair for that.

There is no such thing as excess power (not in the sense you probably think about). Yes there is extra power available if needed but it will come at the cost of more fuel. Anytime you extract 1kWh from the alternator and put in to a battery you can expect a penalty of about 1 liter of extra fuel consumption compared to not have that alternator connected to battery.

If you are OK to pay for 1kWh of electrical energy whatever it cost you to buy 1 liter of diesel then that is fine but the 1kWh you take from alternator will never be free and it always comes with increase in fuel consumption.

It seems you like to oversize the wires way above what is needed :) and that is OK if you can afford the cost but there are not many benefits do doing that.

If 500A is your peak continues current (excluding up to 2x so 1000A surge of a few seconds) 4x 35mm^2 is all that you will have needed. The 7 x 70mm^2 seems quite a bit above that. I think 4x 50mm^2 will be a reasonable oversize for 500A continues

The best protection is always redundancy so two batteries maybe in different locations and slightly differently build will be better than just one large battery.

The safest way is not to find yourself in 8m high waves :) Thus the reason I chose to build a house in the middle of the continent very far from that.

[Chris Sailor]

it doesn't go to a single cell, its one 40cm long 210mm2 busbar where all 4 cells and the shunt is connected. This has also more surface area and better heat exchange then 4x35cm cable. 35sqmm2 to run 500A is forbidden, again by ISO and each connector must alone be able to handle the full load in case the other gets defective or cut off...need to check but i think it min. was 70sqmm2 min for 500A.

Again i would need to fuse all 4 cables, i would add massive resistance and stacking 4 lugs is forbidden. additional where is then my single load switch off requested by ISO too.

reagrding the 7x70sqmm2: yes a part could have been 50sqmm but to simplify parts (lug nuts, fuses...) and also replaement parts i need to carry I only have 16sqmm2 and 70sqmm in my total boat install. Victron requests 2x50 or 70sqm per Multi 3000, have 2 so thats 4x70sqmm2, must have and new loads go from battery bank to navstation where DC distribution is. thats a 10m run one way, so 20m and for voltage drop 70sqmm2 just fits. charge bus 250A so 70sqm is just right too. all victron MPPT, charger...connect with max thats fit into its connector and thats 16sqmm2. fuses NH2 with 200A and 250A for 70sqmm and NH00 63A for 16sqmm2 cable. watermaker with 40A draw with 70sqmm too as 6m cable run one way and here every mV loss counts as that runs a lot, yes 35sqmm2 would have done it. more only do the fridge/freezers which run basically 24/7/365 each 4A till up to 8 A wired with 16mm2 cable.

 only from distribition panel then 2.5,, 4 and 10mm2 to the single loads.

well i start my starboard engines from house so i need min 840AH otherwise startup current rush of starter will damage the cells, i replace 100kg of starter lead batteries like that and have a less complex system. but yes i will add a 2nd 304AH battery to bank with 304AH EVE cells and a JK BMS to get even more capacity that can act as backup bank. inside my battery case i can always regroup the cells or eg take single defective or 4 cells out and eg make a 3p4S bank, the 4th cell each pack is a working spare.

engines my diesel flow meter show me consumption, so 100-300ml more but not 2l for 2x80A charge (2 alternator per engine). can switch them on/off with the sense wire. same with 1 or 2 engines, 1 engine pure motoring 1800RPM 3l/per h, doing 7kn,  both engines 1800RPM are 6l per hour but i am only 2knots faster as resitance of water goes exponentinally up but its 2 gearboxes and 2 propellers that creating resistance and losses. but 1 engine needs 2300RPM and only 4,5l to reach the 9kn too. Advantage of being overpowered i only run 1 engine while the standard with 2x20hp would need to run both at even higher RPM making more noise and need more fuel as higher losses in heat.

well the cost difference is 67000Euro and will never amortisate, thats why this is not offered offically. but with 2x20hp the cat would have been washed into the cliffs during that lee strom with 8m waves like 4 other sail boats did while i could fight it with my 2x50hp running them at 2500RPM both.

[Dacian Todea (electrodacus)]

It is not 35mm^2 for 500A but for about 125A. Current will split fairly equally trough those 4 flexible copper cables.  I'm not sure what the legislation is but I can understand one of the 4 wires failing in some way (bad connection so less current trough that) Still 3x 35mm^2 cable can still handle that.

I was not aware that you had one of those thick buss bars connecting the 4 cells in parallel. That will likely be prohibited by the cell manufacturers if cells are used in high vibration environments. And even just thermal expansion will be a problem excluding the vibrations.

What type of cells are used in this install ? Maybe you mentioned but I do not remember if they are EVE style or Winston.

In any case the best solution here to both isolate mechanically (for vibrations) and to reduce the electrical contact resistance and to move the heat source away from cells will be to move this 40cm long buss bar on witch I guess the shunt is directly connected 10cm - 15cm away from terminals and use 4 flexible busbars to connect to cells. This way since cells are in parallel trough this buss bars the cell voltage will be the detection method to know if one or more of this 4 connections have failed or even one of the cells as you are using them fairly close to the limit.

I will try to take a thermal image while I operate at or above 100A for over one hour and you will see all cables and connections are below 50C in 25C ambient temperature except maybe for the circuit breaker that is rated at 125A and when it operates close to that typical 107 to 110A it gets close to +60C as it is fairly warm to touch (not the best quality circuit breaker).  It is especially true if I have a 100A charge and 80A discharge trough it.

There is a single 25mm^2 cable going to the 500A 100mV shunt (there is now also a 200A 75mV shunt installed for PV it was not in this photo after initial install).

The connections between cells are all flexible and can handle high vibration environment (not the case for me).

You can notice that I have left two connection points one for each cell but since this is fairly low current 125A peak it is not necessary as both cells and connections will be cool no more than 10C above ambient.

But if I went for 250A so double then I will have used two wires one on each cell terminal and probably went up to 35mm^2 for each.

I'm referring to this type of flexible buss bars as what I use to parallel the group of two cells (mine are just the thinner ones 35mm^2) but you find options up to 75mm^2 https://www.aliexpress.com/item/1005003515631469.html

They are 90mm long so enough to put a distance from the shunt and do an excellent job at vibration insulation.

If you use this same type of cells then this will be the same size as your battery if is 12V 4s4p. Will love to see a photo if you have one as then I may have a better idea on how it is build.

[Dacian Todea (electrodacus)]

The flow meter is likely very inaccurate. 300ml of fuel contain just 3kWh of energy and no diesel engine is more than 20% typical best case 30% and when you add the typical efficiency of an alternator of just around 50 to 60% efficient then you are at 10 to 15% overall efficiency in converting fuel to electricity.

2x 80A * 13V = 2.1kW so in one hour 2.1kWh and that at 10% overall efficiency requires 2 liters of Diesel.

There is just no way around that. If you have best engine available and it is run at the peak of the efficiency curve you may get away with just 1.5 liter of diesel for 2kWh extracted from alternators but that is extremely unlikely to be the case.

Assuming 20% efficiency at 3l per hour that will be about 6kW of mechanical power (that is about 8hp) So you could almost run electric at low speed.

How much fuel you normally take with you max ?

On Monday, November 6, 2023 at 12:53:35 AM UTC-6 captain...@gmail.com wrote:

[Chris R8]

Dacian,

To install in boats and RV to be safe and get it insured in Europe its ISO Norm and in US its AYBC install guidelines be followed.

And they require for parallel cables that each of them must be able to carry the max current it will reach in total so 500A here, under a huge voltage drop but one cable must physically survive and for 500A if I remember correctly the minimum was 70sqmm2 for cables not running together in a closed conduit and not in engine rooms. Additionally stacking lug nuts is only allowed up to 6 till 300A and above 3 in Europe and 4 in US. So not possible to do like you suggested.

Also the flexible busbars like you have are rated 300A, not 500A so I can't use them.

Yes cells are 272AH Lishen which are EVE type, compression with 8mm threaded rods.

6W cannot create that amount of heat, something is wrong here. Question is just what? the 500A Victron PV shunt stays much cooler with similar load of 200A.

[Dacian Todea (electrodacus)]

The larger of the flexible bussbar's in that link was 75mm^2 and it is absolutely ridiculous to consider that 3 out of 4 bussbar's will be disconnected. Even if that happens you will see the significant voltage increase on that single cell that is still connected so the SBMS0 will disconnect the load anyway at 500A with a single cell.

There will be no stacked lugs as you will have 4 of those flexible 75mm^2 buss bars one connected to each of the 4 positive terminals then the other end of that flexible buss bar will be connected to the 210mm^2 buss bar that you have now witch is connected directly to the current shunt. So up to this point there is no stacking at all each has his own separate connection.

The flexible buss bars that I have are only 35mm^2 and in the link I provided above there are 3 types the 35mm^2 I have the 50mm^2 and the massive 75mm^2

Best guess is a bad connection that is the reason for excessive heat not the shunt but the shunt will also add up to the bad connection.  You can just set the multi meter on mV DC and check voltage drop across all connection points to identify the reason for excessive heat.

But with solid buss bar connection in a both vibration and corrosive environment the connections will always get bad over time.  That is the reason I insist on decoupling the battery from the shunt using flexible bussbar.

So what I'm saying is just extend each cell terminal by about 90mm using those 75mm^2 flexible buss bars (90 x 30 x 5.5mm). There is no other modification to your current setup other than moving that solid busbar a few cm away from the cell terminals by extending those with 4 flexible buss bars.  So cell sense will be connected on the 210mm solid bar and thus you will be able to notice the large voltage fluctuation if one and especially two of those flexible buss bars has a bad connection either at the cell or at the solid bussbar. And if only one remains (basically impossible for such a thing to happen ) at 500A (even at much less that that 300A) the single Lishen cell that remains connected will be incapable to deal with this without very large voltage drop with will trigger the SBMS0 under voltage or over voltage lock so you have a sensing method to know you have multiple bad connections.

[Chris Sailor]

Hi Dacian,

so what i could do is cut the terminal busbar a bit shorter and using 3 of the 75mm flexible busbars stacked per terminal negative and positive to connect terminal busbar with the shunt. that would give me much more surface area and will solve another problem i have. As there is no interface into Victron Cerbo GX of the SBMS i need to integrate the Victron shunt of the BV712 (to get battery data into cerbo GX, sadly not the  cell voltages)  into negative terminal but height of shunt is too large to connect it. the flexible busbars would solve this problem...mhh. the cerbo is integrated then in the boats systems like eg via chartplotter at the helm i get a full overview about the whole electric system.

well if you wanna be insured and the insurance cannot bail out you have to keep to the standards, my cat is german registered so ISO applies. my install due t DIY battery bank is checked and approved by a surveyor and therfor explcitly insured. Yes some stuff of ISO is ridicolous but well it is what it is. The stagged lugs rule makes a lot sense, first you need special stack lugs as standard are not flat enough and a big part of surface won't connect if you use standard ones. then you have a lot contact points in series where resistance can happen and eg if the 2nd lowest lug has a connection problem a very high current going through it and will heat it up massivly. saw some dripping insulation from stack lug cables where that did happen in a big car stereo install...you just torque down the top nut but how the torque is distributed inbetween the 6 stacked lugs you simply have no control of. again 6 stcked is actually quite high, my experience all above 3 create problems (not in in the till 200A range but above yes)

i did car Hifi competition systems with up to 1500A so i know how to deal with that much current, whats important and whats no gos...

all above 300A is tricky, hardware hard to get and only smallest differences create troubles.Thats why all brand inverter stop at 3kw@12V ( that creates 250-280A cont) and Victron is the only one who has a real 5kw 12V inverter...

with 500A all cells need to be idientically connected, only small differences will cause imbalances and tons of issue. thats the reason you eg use silver plated high spec cooper busbars and not standard tinned ones as connection with all lot of material are much better regarding resistance then with tin. silver connects much better with alu, brass or cooper then tin and doesn#t create any contact issues with eg oxidation that creates an isolation layer like eg alu.

Also only use class T or NH fuses above 300A continious, as all other fuse type speced for 70% load of rating continious and use the fuseholder and cables as cooling while NH are constructed to run at max rating 7/24/365.

i prefer NH as cheaper, have 100kA short curcuit protection and more massive plus all connector silver plated too. my 500A NH3 fuse is not getting more then handwarm when you run 500A continously over 7/24/365 it and can be used as cut off switch in till 48V environment too by just pulling the fuse with the removal tool under full load if necessary. save 2 contact points, resistance, voltage drop and has 2 points of failures less plus its cheaper as 500A cut off switches cost a liver...