Electrodacus BMS- no contactors?

[sailingharry]

This is the 500A double shunt I mentioned. I got mine NOS (New Old Stock) for under $40 delivered, this one is expensive. Set an eBay search for "Link 2000 shunt" and be patient.
https://www.ebay.com/itm/256427267024

[svloki]

Originally Posted by sailingharry

Your plan is a major problem! Dacian is very clear -- the shunt must be directly (and closely) connected to the battery. No fuse, no switch. Fuses and switches must be AFTER the shunt. This means you need two fuses and two contactors. The 7713 is 3.75 x 5.5, and over $200 -- and then you need another Class T fuse and all the holders/cables/etc to wire it all up. Agreed, it is a one-time cost and you get all the goodness Dacian programmed in.


I think that the fuses/contactors cause two problems. First, they add additional voltage drop which I think impacts the current measurement. But more importantly, I think that if either of them open, it can fry the 'Dacus.

Interesting. The cardinal rule for safety is that the fuse is always first (protect the downstream wiring). The only allowable exception is for very very short cables. If the shunts are right next to the B+ terminal it should be quite safe.

Regarding the voltage drop across a class-T 600a fuse. If that voltage drop is significant enough to effect the readings in any meaningful way, the installation has not been correctly executed. Correct?

I believe that a properly sized disconnect relay should also have very low voltage drop. Can we see your meter readings? Need the current at the time of test. Awesome if you have a scope / differential probe for this.

Has the manufacturer confirmed a positive side disconnect will damage his BMS? I would think those expensive Blue Seas units have built in flyback protection. Dangerous assumption.

[svloki]

Originally Posted by sailingharry

This is the 500A double shunt I mentioned. I got mine NOS (New Old Stock) for under $40 delivered, this one is expensive. Set an eBay search for "Link 2000 shunt" and be patient.
https://www.ebay.com/itm/256427267024

If I remember correctly, the old Heart devices used an off the shelf 50mv shunt. Can just cut a copper bar to jump two ends of those. Easy?

[CaptainRivet]

Originally Posted by svloki

Are you planning on two latching relay switches? They are pretty small really. My design is just one, main disconnect (eliminates the manual battery switch). Bolting mine to the side of the batt box. Pretty insignificant?

Nope that won't work.
Shunts of electrodacus go into positive and both shunts in series with a takeoff after the first load shunt.
Both shunts need to be connected with a short cable directly to the positive terminal of the bank.
After load shunt comes directly the battery fuse and after the PV shunt the charge fuse.
The BMV712 shunt goes into negative.

Stay away from mechanical relays like 7713 wherever you can. That's what electrodacus tries to avoid whenever and whereever possible. Use it's concept and philosophy as is and don't modify it otherwise things won't work properly. If you don't like that then better use a different BMS.
It's already a stretch using it on a vessel and with much more charge sources then in an offgrid land based installation but philosophy and concept is kept and install adapted accordingly.
and use high quality SSR like Victron battery protect or. Never buy these SSR at Amazon or ebay or alibaba where you get mostly fakes. I recommend the big electronic wholesaler mouser and digikey.
You can use SSR when current flows in only one direction which is the case in load and charge bus install. Additionally SSR cannot do surge like big inverters create due to their huge capacitors, if you can relaibly measure surge and find an SSR where it's rating is equal or smaller then surge current of the inverter you can use thus.

The charge bus can be switched with a smart BP100 or 220 , you can use several in parallel.
In my chargebus i have 4MPPT with one smart BP220 (as they can in a defect send full solar voltage to bank and overcharge it) and another for the rest.
The huge advantage of the smart battery protect you can via BT connect app switch on/off the bus independently from the BMS. That makes test and troubleshooting very easy and comfortable.
Same loadbus with a smart BP100 and a BP220.
The multiplus as load and charger have their own bus and connected to load shunt after main fuse without a seperate disconnect relay as its interal cut off is much more relaible then an external big relay i need to use and introduces a new weakest link. Load bus as I hardly ever use shorepower, so it's charger is not used in normal operation.

As shunts I highly recommend to use victron shunts as they are very precise and you can really run 500A over the 500A shunt. Shunts rating are 1/3rd to high as you normally can run 350A cont over a 500A shunt but victron took that into account and speced it lower. So the 500A victron shunt is in reality a 700A shunt and the 1000A a 1300A shunt from the physical size so it can handle the 500A rating as continuous. Victron are the 50mV shunts which is also best compromise for heat vs resolution.
Shunts get hot, mount them very close to battery but isolate as good as possible as heat is the enemy of LFP.
Means don't mount shunt directly on the battery case That's why you also don't connect the shunts direct to the busbar as introduces heat into busbars and then to cells. Use a short cable with high temp silicone isolation, better 2 exact same length so heat from shunt to busbar is reduced as much as possible.

[CaptainRivet]

Download electrodacus beginners guide and follow it.
https://groups.google.com/g/electrodacus/c/qorsBKh2HaQ

If you don't like concept or philosophy choose another BMS. It's different then all other BMS, accept this and use it the way it was designed and don't try to modify...
One big main point of its concept is to avoid big notoriously failing relays and one central point disconnects with the total systems current.
In offgrid you have huge charger/inverter that create a huge problem when you try to cut it at the battery as these huge relays (due to surge and current flows in both direction) are notoriously unreliable due to mechanics and the problem resulting in switch currents you use for welding. The cut off features inside a multiplus on the low side are much more reliable then your big relay that you introduce to do this job, means you make your system more unreliable by introducing these relays.
And Darcian lives high up north in nomandsland in Canada and is screwed if the inverters fail, next neighbor is far away...so he thought about a different solution and created electrodacus BMS

[svloki]

Originally Posted by CaptainRivet

Nope that won't work.
Shunts of electrodacus go into positive and both shunts in series with a takeoff after the first load shunt.
Both shunts need to be connected with a short cable directly to the positive terminal of the bank.
After load shunt comes directly the battery fuse and after the PV shunt the charge fuse.
The BMV712 shunt goes into negative.

I re-drew things based on that. Still not 100% convinced that a single contactor is not viable. Is the only concern the possible introduction of a slight error in current flow measurement? Maybe I can live with it.

[sailingharry]

Originally Posted by svloki

Interesting. The cardinal rule for safety is that the fuse is always first (protect the downstream wiring). The only allowable exception is for very very short cables. If the shunts are right next to the B+ terminal it should be quite safe.

Regarding the voltage drop across a class-T 600a fuse. If that voltage drop is significant enough to effect the readings in any meaningful way, the installation has not been correctly executed. Correct?

I believe that a properly sized disconnect relay should also have very low voltage drop. Can we see your meter readings? Need the current at the time of test. Awesome if you have a scope / differential probe for this.

Has the manufacturer confirmed a positive side disconnect will damage his BMS? I would think those expensive Blue Seas units have built in flyback protection. Dangerous assumption.

Yes, the fuse and disconnect should be first. And a shunt, completely uninsulated and operating as a heat source (so can't be enclosed) is a risk. I'd rather the shunt be on the negative side (so I can use the one I arleady have there and won't be removing!), but I like the 'Dacus more than other BMS that use low side shunts.


Voltage drop across fuses and relays is very low. How that compares to the very low drop across shunts, I don't know. I also don't know for fact that this is the concern. I do know that Dacian wants very short cables to the shunt, and not for ABYC reasons. I'm guessing here. And I have no data because my install is still months (or even a year?) away. My AGM haven't failed yet, so I still have time.


Dacian's documentation is best described as "slim." There is another document that helps a lot here https://diysolarforum.com/resources/...ctrodacus.174/ but it still lacks a lot. However, Dacian is extremely responsive on his forum here https://groups.google.com/g/electrodacus where I have been lurking for a year or so. He quickly answers almost any question with great detail, usually with backup explanations, and he has often commented on how opening the connection between the battery Plus and the shunt will fry the SMBS0. Both the manual and the document I linked show the contactor after the shunt (but don't actually say it).


I would much prefer a fuse and contactor before the shunt, both for safety and for a neat and less expensive installation -- but it's not an option for this BMS. The extra cost for needing 2 vice 1 (probably north of $300 for contactor, fuse, fuse holder cables, etc) should be added into any comparison between the 'Dacus and other BMS (hmmm.... i need to do that.....).

[svloki]

two BP220 is about the same cost as one 7713 so it's not a bad change to keep Electrodacus happy. I believe ABYC recommendations are still met by the use of dual SCR's to get full disconnect. One source drives them in my design, the 712.

[svloki]

Originally Posted by sailingharry

He quickly answers almost any question with great detail, usually with backup explanations, and he has often commented on how opening the connection between the battery Plus and the shunt will fry the SMBS0. Both the manual and the document I linked show the contactor after the shunt (but don't actually say it).

Hmm I am not getting the fundamental difference between a disconnect before the shunts and after in terms of damage to the BMS. If it's back EMF, that would be the same it seems. Must be something else? I will try asking master Dacian.

[sailingharry]

This post is one of several that you can find on the group:
https://groups.google.com/g/electrod...m/gEbyoH5iAQAJ


It has to do with voltages. He says it can be "hundreds" which makes little sense (flyback issues aren't relevant in most of our boat wiring systems). But regardless, he believes an opening disconnect/fuse before the shunt can blow the BMS electronics. At the very least, the BMS will be running at 12V and the shunt will be at 0V, which may be beyond the tolerances for the ADC input values.

[CaptainRivet]

Originally Posted by svloki

Hmm I am not getting the fundamental difference between a disconnect before the shunts and after in terms of damage to the BMS. If it's back EMF, that would be the same it seems. Must be something else? I will try asking master Dacian.

Because the main load shunt must always be connected to the current measurements chip of the BMS. If under load the main shunt gets disconnected you fry the current measurement mosfet which then can cause damaging the BMS chip or the board due to heat. A shunt can never create a short or anything dangerous so it can also be before the fuse. Also in a desaster shut off the electrodacus way the BMS is still full operational.
There a dozens of interdependencies you only get if you operate the system over years and installed several of them. As said it's different then to all other BMS.
Having it the electrodacus way and have the battery always online has a lot advantages, even more with alternators as you can never get a diconnect under full load.
My surveyor had first a problem with this too till he got familiar with the different concept and realized it's in the end saver then other solutions...its already closed to six sigma but a 1000% solution we will never have doesn't matter which concept or BMS you use. And by trying to make it safer too much complexity is introduced that actually in the end make it unsafer.
Dacian also has his own way for solar with his own steering based on panel voltage where it cannot happen that a 50V panel can deliver the full 50V directly to battery as this can happen with MPPT even if you shut it or cannot shut it off via the remote. That's why the MPPT shut off via remote in normal operation but mostly a battery protect is used by the community for a desaster shut off that cuts the MPPT on the output side.

[CaptainRivet]

Originally Posted by svloki

Hmm I am not getting the fundamental difference between a disconnect before the shunts and after in terms of damage to the BMS. If it's back EMF, that would be the same it seems. Must be something else? I will try asking master Dacian.

Because the main load shunt must always be connected to the current measurements chip of the BMS. If under load the main shunt gets disconnected you fry the current measurement mosfet which then can cause damaging the BMS chip or the board due to heat. A shunt can never create a short or anything dangerous so it can also be before the fuse. ElectrodacusBMS has no fuses as the thin twisted pair cables used act also as fuse but cannot cause a fire..I know I managed that once (i am a human and mistakes happen), they basically disappear into dust right away. BMS still worked afterwards.
There a dozens of interdependencies you only get if you operate the system over years and installed several of them. As said it's different then to all other BMS.
The alternator is the best example that in can always stay connected with a proper speced fuse. In all cases switching the field or sense wire off is sufficent so the alternator doesn't produce any power anymore. In an internal defect like stator short out, field coil burns through...the fuse will disconnect the defect alternator.
But cutting an alternator on the output will running under load is what actually the root cause of all the problems around an alternator.
Having it the electrodacus way and have the battery always online has a lot advantages, even more with alternators as you can never get a diconnect under full load, one problem less to solve.
My surveyor had first a problem with this too till he got familiar with the different concept and realized it's in the end saver then other solutions...its already closed to six sigma but a 1000% solution we will never have doesn't matter which concept or BMS you use. And by trying to make it safer too much complexity is introduced that actually in the end make it unsafer.
Dacian also has his own way for solar with his own steering based on panel voltage where it cannot happen that a 50V panel can deliver the full 50V directly to battery as this can happen with MPPT even if you shut it or cannot shut it off via the remote. That's why the MPPT shut off via remote in normal operation but mostly a battery protect is used by the community for a desaster shut off that cuts the MPPT on the output side. Decision is done device by device what's the best and safest way and in 90% it's cutoff/steering via remote. This allows the Electrodacus to just shut off a device that would cause a desaster shutdown with other BMS.

[svloki]

Originally Posted by CaptainRivet

Because the main load shunt must always be connected to the current measurements chip of the BMS. If under load the main shunt gets disconnected you fry the current measurement mosfet which then can cause damaging the BMS chip or the board due to heat. A shunt can never create a short or anything dangerous so it can also be before the fuse. ElectrodacusBMS has no fuses as the thin twisted pair cables used act also as fuse but cannot cause a fire..I know I managed that once (i am a human and mistakes happen), they basically disappear into dust right away. BMS still worked afterwards.
There a dozens of interdependencies you only get if you operate the system over years and installed several of them. As said it's different then to all other BMS.
The alternator is the best example that in can always stay connected with a proper speced fuse. In all cases switching the field or sense wire off is sufficent so the alternator doesn't produce any power anymore. In an internal defect like stator short out, field coil burns through...the fuse will disconnect the defect alternator.
But cutting an alternator on the output will running under load is what actually the root cause of all the problems around an alternator.
Having it the electrodacus way and have the battery always online has a lot advantages, even more with alternators as you can never get a diconnect under full load, one problem less to solve.
My surveyor had first a problem with this too till he got familiar with the different concept and realized it's in the end saver then other solutions...its already closed to six sigma but a 1000% solution we will never have doesn't matter which concept or BMS you use. And by trying to make it safer too much complexity is introduced that actually in the end make it unsafer.
Dacian also has his own way for solar with his own steering based on panel voltage where it cannot happen that a 50V panel can deliver the full 50V directly to battery as this can happen with MPPT even if you shut it or cannot shut it off via the remote. That's why the MPPT shut off via remote in normal operation but mostly a battery protect is used by the community for a desaster shut off that cuts the MPPT on the output side. Decision is done device by device what's the best and safest way and in 90% it's cutoff/steering via remote. This allows the Electrodacus to just shut off a device that would cause a desaster shutdown with other BMS.

That is unfortunate that his design cannot tolerate the battery disconnect and fuses before the shunts. Likely a future revision will correct this.

There is a sound reason that fuses go first. Anything can accidentally get shorted to ground, including a shunt. If that was to happen, it would get bad very fast (fire).

I do like the fundamental approach that Dacian has implemented. I think with the compromise- fuses and SCR's immediately after the shunts to effectively disconnect the battery, we can both have ED doing it's job, yet (mostly) meet ABYC recommendations. If the shunts are very well-protected, risk of is short circuit is mitigated.

Many will agree with Dacian that the battery can be left on to power the BMS, low risk. My take is that it's sound engineering and a safer approach for yachts is to have a disconnect mechanism if all else fails.

[CaptainRivet]

Originally Posted by svloki

That is unfortunate that his design cannot tolerate the battery disconnect and fuses before the shunts. Likely a future revision will correct this.

There is a sound reason that fuses go first. Anything can accidentally get shorted to ground, including a shunt. If that was to happen, it would get bad very fast (fire).

I do like the fundamental approach that Dacian has implemented. I think with the compromise- fuses and SCR's immediately after the shunts to effectively disconnect the battery, we can both have ED doing it's job, yet (mostly) meet ABYC recommendations. If the shunts are very well-protected, risk of is short circuit is mitigated.

Many will agree with Dacian that the battery can be left on to power the BMS, low risk. My take is that it's sound engineering and a safer approach for yachts is to have a disconnect mechanism if all else fails.

No nothing will correct this, I got the curcuits plans from Darcian, the chip design needs that, will be never changed and he also has no desire to change as it works flawlessly like this since the beginning. Its only a problem in your head...
As I said Darcian is extremely constant and persistent on his way, even more then a Steve Jobs and this makes it for continuity.
It's the same then using apple during Steve Jobs, the apple way has a lot of advantages versus windows but also some limitations you have to accept. I converted late in 2014 to apple for Mac Air and whenever I do something the first time or forget how it was done in mircosoft it works, when I had the same thing done in Windows before it doesn't work and if I then Google I find out that apple is the simple intuative way and Microsoft the complex...electrodacus is apple here.

It's your job to make it impossible to short out the shunt, a simple separation of negativ and positive with a seperate cover achieves that. Never open both covers at the same time and nothing can ever happen.
Accept it or choose another BMS.
Always online are alternator and Multiplus, they have no battery protect to cut off only a fuse.
You cannot disconnect on DC side a multiplus or any big inverter with a battery protect or any SSR (unless its constant current is more then surge of inverter but thats only possible for small inverter) as the surge of the big capacitors will kill it, is also stated in the manual.
As explained you would need a big relay which is introducing a weaker shutdown then via internal shutdown mechanism of the victron which is extremely well tested by millions of user. You need to check with your inverter.
Alternator stays always online with only a fuse as cutting it actually creates the big problem.
Darcian has no idea and doesn't care about alternators (other then they are a very inefficient energy production and need to be avioded) as in offgrid not present.
You can shut off the alternator via DC2DC charger shutting down via remote as done in most RVs but DC2DC has his own drawbacks...eg you need a lead battery which I don't have anymore and never want one back.
And adding a new device is simple, you just decide which way to disconnect is the safest and implement that for that device and only have to care for this device to correctly spec it. You don't have to think or check or forget to check the whole path downstream of it like with all the other BMs. Eg. if you are still within the specs of your main general disconnect relay or the new device needs too much current so you have to go with a bigger relay or disconector. The bigger then needs more power, can your steering curcuit deliver that or do you need a different one because bigger works only with pulse... Or if you even allowed to cut the hot side under load or if that actually creates the bigger problem....many many things you have to think and check and if forget you are in trouble.
not so with electrodacus. Only thing I need to check downstream is if the new total draw or charge with this new device is within the main battery or charge fuse but that I check before buying it as I need to obviously power it or battery can accept this new total charge.

[sailingharry]

Originally Posted by svloki

There is a sound reason that fuses go first. Anything can accidentally get shorted to ground, including a shunt. If that was to happen, it would get bad very fast (fire).

Philosophically, you are of course correct. Practically, it's not a big deal. I'm not even sure that you would find a prohibition in a critical, word-by-word reading of ABYC.


Here's the thing. In a standard install, there is a short battery cable. It goes 6" to a fuse holder input, then to the fuse, then to the fuse holder output, then "onward" (battery switch, etc). The risk exposure is the battery terminal and the entire fuse and holder assembly.


In the plan as envisioned in my head (and parts are coming together), the install is a short battery cable to the shunt. The shunt "out" bolt matches my Class-T fuse. The other end of the Class-T will sit on the contactor. So, the exposure is the battery terminal, the shunt (about 2" square of exposed metal), and the inlet side of the Class-T. A simple perforated plastic shield could be mounted over the shut/fuse assembly.


Reading the above two paragraphs, the increased risk by installing as required by the 'Dacus is the shunt -- and it is immediately in contact with the fuse, so it is basically the same risk as the fuse/fuse holder, just perhaps a few additional square inches of exposure. It is an increased risk, but the delta is small.


I have attached a pic of the shunt I'll be using with 2 Class-T fuses for reference. The shunt itself is the only additional risk over a standard install. The battery cable will land on the bottom right bolt.