Arco Zeus regulator - first impressions and a look inside

[BEngo]

40 minutes ago, cheesegas said:

48 minutes ago, BEngo said:

Expand  

Whilst it's indeed possible to control the charge/discharge FETs externally, this is the reverse of how other BMSs work. For example, RECbms/Pylontech will take control of all charge sources and tell them when to start and stop charging rather than being told to switch on and off. The contactor/FET in these BMSs is a last resort, which is how it should be in my opinion.

 

I agree that one should not rely oon the BMS  FETs for control. I would prefer to take the decision information from the BMS and then shut down the alternator  when the conditions are met.

That means a modified alternator brush box or regulator at the very least, but does not address the alternator heating issue.  A halfway house would be to extract the BMS info and shut down the alternator on that basis but use a BtoB  to limit the alternator output, possibly also  controlling the half-rate switch depending on conditions.

 

N

 

 

[IanD]

10 minutes ago, dmr said:

I feel that UARTs and RS485 etc etc are of little use to the average user (or even me and I know what they are 😀). CANbus is probably the best as it is slowly becoming a boat standard but is not available on most batteries.  A very simple "alow to charge" signal (a relay drive type thing) from the batteries that could be progammed (via bluetooth) to switch at 90% or whatever (well before the internal disconnect) would be a very useful solution for those who don't want a full CANbus or VEbus system.

 

Zeus and some Victron DC-DC converters etc are able to respond to a simple allow to charge signal.

 

That would be a good thing for this application, but I don't think the drop-in manufacturers are likely to do it, they seem to see their batteries as either standalone or at best telling you what they're doing so you can monitor them -- they're "in charge" of their own (dis-) connection, but not external gear. Adding CANBUS costs money (a lot more than Bluetooth which is essentially free) and allows careless users to damage the battery/interface, so they don't fit it.

 

In contrast the external BMS (like my REC-BMS) assume they're the "battery master" for the system, and signal to other charge sources (shoreline, generator, solar/MPPT, alternator) when to charge or shut off -- usually via CANBUS to an external controller like a Cerbo or combo, but sometimes also via an "allow to charge" wire like you described.

 

 

 

2 minutes ago, dmr said:

I feel that UARTs and RS485 etc etc are of little use to the average user (or even me and I know what they are 😀). CANbus is probably the best as it is slowly becoming a boat standard but is not available on most batteries.  A very simple "alow to charge" signal (a relay drive type thing) from the batteries that could be progammed (via bluetooth) to switch at 90% or whatever (well before the internal disconnect) would be a very useful solution for those who don't want a full CANbus or VEbus system.

 

Zeus and some Victron DC-DC converters etc are able to respond to a simple allow to charge signal.

 

That would be a good thing for this application, but I don't think the drop-in manufacturers are likely to do it, they seem to see their batteries as either standalone or at best telling you what they're doing so you can monitor them -- they're "in charge" of their own (dis-) connection, but not external gear. In contrast the external BMS (like my REC-BMS) assume they're the "battery master" for the system, and signal to other charge sources (shoreline, generator, solar/MPPT, alternator) when to charge or shut off -- usually via CANBUS to an external controller like a Cerbo or combo.

 

Edited February 22 by IanD

[Jon57]

3 hours ago, cheesegas said:

Whilst it's indeed possible to control the charge/discharge FETs externally, this is the reverse of how other BMSs work. For example, RECbms/Pylontech will take control of all charge sources and tell them when to start and stop charging rather than being told to switch on and off. The contactor/FET in these BMSs is a last resort, which is how it should be in my opinion.

 

However, having an external battery gauge such as the Victron BMV along with a JBD BMS is a good interim solution though; setting the BMVs relay to close at 100% becomes the command to stop charging, and opening it at 95% or whatever is the allow to charge. This will however result in short cycling if it's being charged with a heavy load on the battery which is less than the current the charger can supply

I largely agree with this, but there are more and more drop ins appearing on the market with CANbus. Renogy's newest offering is fully Victron compatible for example, and there's a few more coming to market, can't remember the brands offhand though!

 

There are of course Pylontechs etc which have an internal BMS and will speak to a Cerbo happily to take control over charge sources, but they are designed for domestic installations and I would not recommend using them on a boat for a few reasons.

I largely agree with this, but there are more and more drop ins appearing on the market with CANbus. Renogy's newest offering is fully Victron compatible for example, and there's a few more coming to market, can't remember the brands offhand though

Fogstar drift pro are victron compatible I believe. 

[nicknorman]

3 hours ago, dmr said:

 CANbus is probably the best as it is slowly becoming a boat standard but is not available on most batteries. 

CANBUS is very good, highly robust and easy to implement. However it is just a hardware specification that moves bytes of data around. What the data means, is up to the system designer. So just because 2 devices have CANBUS in no way means they can necessarily talk to each other meaningfully. Apart from NMEA2000 I’m not sure that there is a standard protocol for boats or Li batteries. Maybe there is? Dunno.

[IanD]

15 minutes ago, nicknorman said:

CANBUS is very good, highly robust and easy to implement. However it is just a hardware specification that moves bytes of data around. What the data means, is up to the system designer. So just because 2 devices have CANBUS in no way means they can necessarily talk to each other meaningfully. Apart from NMEA2000 I’m not sure that there is a standard protocol for boats or Li batteries. Maybe there is? Dunno.

 

AFAIK there isn't, as you say this is the next level up in the CANBUS software stack -- the meaning of data/messages, not just data protocol. I believe there are some "standard" messages, but there's no requirement to support them, manufacturers can do what they want. It's why only certain BMS/CANBUS batteries are on the Victron approved list, the only way to be sure they will be interoperable with the controller is to try them.

 

CANBUS guarantees that you can plug the hardware together and the data (e.g. the letters of a message) will get there OK -- but if it's being sent in English and the receiver only understands French that doesn't help... 😉

[dmr]

32 minutes ago, nicknorman said:

CANBUS is very good, highly robust and easy to implement. However it is just a hardware specification that moves bytes of data around. What the data means, is up to the system designer. So just because 2 devices have CANBUS in no way means they can necessarily talk to each other meaningfully. Apart from NMEA2000 I’m not sure that there is a standard protocol for boats or Li batteries. Maybe there is? Dunno.

 

This is the impression that I get. I'm very new to CANbus (but have lots of experience of old stuff like GPIB) but it does appear to be just an electrical and "bit level" standard. So here is a chance for you to make a name for yourself, after all JSON was pretty much a one man effort (only joking here) 😀.

 

A problem with standards is that various manufacturers and other organisations have to get together and it sometimes all takes years to sort out. Maybe if somebody like Victron published a set of default messages that they would read then it could all happen quickly. "Compatible with Victron" does appear to be the current gold standard. Another issue is that most batteries are using various Chinese BMS systems that are very poorly documented, with little chance of finding out what they do on CANbus.

[PeterF]

15 hours ago, Jon57 said:

I largely agree with this, but there are more and more drop ins appearing on the market with CANbus. Renogy's newest offering is fully Victron compatible for example, and there's a few more coming to market, can't remember the brands offhand though

Fogstar drift pro are victron compatible I believe. 

The problem with a lot of these drop ins with CAN Bus output is if you put several batteries in parallel, they still appear as two batteries, they do not aggregate themselves and present as a single battery to the control device which is a problem. The storage battery systems like Pylontech parallel up and appear as a single large battery. The Fogstsar Drift Pro says for connection to Victron Cerbo GX that you can connect one battery to the BMS Bus and one to the VE CAN Bus, so you have 2 individual batteries that can be monitored but only one can control the charging devices unless you add custom code to aggregate them as a single virtual battery.

[IanD]

15 hours ago, dmr said:

 

This is the impression that I get. I'm very new to CANbus (but have lots of experience of old stuff like GPIB) but it does appear to be just an electrical and "bit level" standard. So here is a chance for you to make a name for yourself, after all JSON was pretty much a one man effort (only joking here) 😀.

 

A problem with standards is that various manufacturers and other organisations have to get together and it sometimes all takes years to sort out. Maybe if somebody like Victron published a set of default messages that they would read then it could all happen quickly. "Compatible with Victron" does appear to be the current gold standard. Another issue is that most batteries are using various Chinese BMS systems that are very poorly documented, with little chance of finding out what they do on CANbus.

 

The CANBUS situation is very similar to GPIB -- different instruments can all connect to the same bus but they can't talk to each other and interact, unless (sometimes) they're from the same manufacturer -- even different instruments of the same type don't always have the same "command set", certainly not from different manufacturers. It's the nature of the beast because different hardware has different features, even if there some common commands you need a device-specific driver to do anything more than the basics.

58 minutes ago, PeterF said:

The problem with a lot of these drop ins with CAN Bus output is if you put several batteries in parallel, they still appear as two batteries, they do not aggregate themselves and present as a single battery to the control device which is a problem. The storage battery systems like Pylontech parallel up and appear as a single large battery. The Fogstsar Drift Pro says for connection to Victron Cerbo GX that you can connect one battery to the BMS Bus and one to the VE CAN Bus, so you have 2 individual batteries that can be monitored but only one can control the charging devices unless you add custom code to aggregate them as a single virtual battery.

 

It's because "drop-in" batteries -- especially the cheap ones -- are really intended to be dumb standalone batteries each in charge of (geddit?) their own fate, not for multiples to be connected and used as one co-ordinated battery -- which then means one internal BMS has to be the master and the others have to be slaves (you can only have one "boss"), but usually the master also has to read data from the slaves (e.g. cell voltages and temperatures) and act on it, and talk to the outside world (and the others don't any more) -- or the system has to be made multi-master with all the batteries talking to the outside world and each other, which is even harder.

 

It's all a lot of effort (and cost) to design/debug so they mostly don't bother, because the vast majority of their customers don't want/need these features, it would just increase the cost (which they don't want) for no benefit.

 

The bigger storage/off-grid suppliers like Pylontech and BYD (and now Fogstar?) do make all this work, because their main application is stacked/paralleled systems -- as do suppliers of external BMS. Fogstar are also introducing similar systems at good prices, but these are very different to their "drop-in" batteries.

 

If people want a large LFP bank that works as one and talks properly to external controllers/alternators/chargers/MPPT/generators, then they should use batteries which are designed and intended for this purpose -- either ready-made ones form the likes of Pylontech/BMS (and now Fogstar?), or a DIY LFP build using cells and a proper BMS controller (e.g. REC-BMS, but there are many others, or even this can be DIY) designed to control and manage a large stack of cells, together with all the appropriate switching/contactors/sensors.

[dmr]

30 minutes ago, IanD said:

 

The CANBUS situation is very similar to GPIB -- different instruments can all connect to the same bus but they can't talk to each other and interact, unless (sometimes) they're from the same manufacturer -- even different instruments of the same type don't always have the same "command set", certainly not from different manufacturers. It's the nature of the beast because different hardware has different features, even if there some common commands you need a device-specific driver to do anything more than the basics.

 

I'm busy today, doing wiring upgrades and then hopefully starting to install the Zeus.

CANbus was electronic geeky stuff but is now becoming something that some boaters (and many car owners) are going to have to get to grips with, especially if the hybrid-lithium scheme falls out of favour. Rather than diverting this thread I propose that we start a new CANbus thread, I will do this this evening if nobody else get their first.

[dmr]

I'm installing the Zeus now, but while I have access am doing some wiring upgrades and tidy up first.

 

I needed to contact Arco yesterday via email. The good news is that they responded quickly.

 

They pointed out that the battery ground and battery negative sense must go to the exact same location (identical voltage). It is Not ok to connect the ground to a busbar and the sense to the battery terminal as any voltage difference can damage the Zeus. The installation instructions are ambiguous on this and do sort of infer that its ok.

[cheesegas]

8 hours ago, PeterF said:

The storage battery systems like Pylontech parallel up and appear as a single large battery. 

Pylontech and similar are actually a bit different - if you were to chain up say 4 of them, the Cerbo talks to the first one only via CANbus - this automatically assigns it the Master status. This master battery then talks to the other three via a proprietary protocol and aggregates the data and any cell overvoltage etc error. That's why, for example with Pytontechs, each battery has a CAN and RS485 port (connecting anything to either port will automatically make it the master) and two Link ports to connect battery to battery. Most of them are sensitive to which Link port you use too, as some also assign master status based on the first one in the chain.

8 hours ago, IanD said:

...either ready-made ones form the likes of Pylontech/BMS (and now Fogstar?), or a DIY LFP build using cells and a proper BMS controller (e.g. REC-BMS, but there are many others, or even this can be DIY) designed to control and manage a large stack of cells, together with all the appropriate switching/contactors/sensors.

I'd strongly advise not putting Pylontechs for one on a boat - they're not designed for things that are damp and move. No conformal coating on the boards, mechanically things aren't very well fixed down inside compared to something like a Victron inverter. Seen a good few fail within the first year, and one install in the engine bay of a large Dutch barge didn't even last a month.

[David Mack]

3 hours ago, dmr said:

They pointed out that the battery ground and battery negative sense must go to the exact same location (identical voltage). It is Not ok to connect the ground to a busbar and the sense to the battery terminal as any voltage difference can damage the Zeus. The installation instructions are ambiguous on this and do sort of infer that its ok.

So why are there two connections then? Why not link them internally in the unit and have a single connection to the battery negative/ground?

Edited February 23 by David Mack

[cheesegas]

2 minutes ago, David Mack said:

So why are there two connections then? Why not link them internally in the unit and have a single connection to the battery negative/ground?

I expect it's because there's around 10 amps flowing in the supply cable, which will induce a small voltage across it, making it unreliable to use to sense battery voltage. The sense line will pass a matter of milliamps which will mean the voltage at the battery end of the cable is the same as the voltage at the regulator end of it.

 

It does seem odd that it'll cause damage if there's a 1v or so difference between sense and supply cables though, and I don't think it's mentioned in the manual.

[dmr]

3 minutes ago, David Mack said:

So why are there two connections then? Why not link them internally in the unit and have a single connection to the battery negative/ground?

 

This two wire approach is normal, Adverc do it and I think Sterling do it too. The ground wire carries any current that the Zeus uses and the cable and any connections could have a small volt drop. The sense wire measures the voltage at the battery and will have almost no current flowing so is not subject to a volt drop. Think about the separate sense connections on a shunt as an extreme case.    The Arco appears to use so little current that a separate semse wire might not be essential but its still a good idea.

[David Mack]

So if the sense wire and battery wire have different voltage drops, then the unit will see different voltages on each. Yet the post I was responding to said any such voltage difference would damage the unit. So which is it?

[cheesegas]

3 minutes ago, dmr said:

The Arco appears to use so little current that a separate semse wire might not be essential but its still a good idea.

The reg itself may use very little but it'll still be powering the alternator's field which I think for the Zeus is up to 10 amps, so there's a good bit of current flowing in that cable which explains the sense cable.

Just now, David Mack said:

So if the sense wire and battery wire have different voltage drops, then the unit will see different voltages on each. Yet the post I was responding to said any such voltage difference would damage the unit. So which is it?

Yep, that's why I mentioned it was odd - perhaps a large difference will damage it, but a 1-2v difference won't. Would be nice to know from them.

[dmr]

16 minutes ago, cheesegas said:

The reg itself may use very little but it'll still be powering the alternator's field which I think for the Zeus is up to 10 amps, so there's a good bit of current flowing in that cable which explains the sense cable.

Yep, that's why I mentioned it was odd - perhaps a large difference will damage it, but a 1-2v difference won't. Would be nice to know from them.

 

The Zeus has two power supplies, it takes the high current alternator drive from the alternator output post, and the general electronics drive from either the battery directly or via the ignition switch. It must be low current consumption as it has an option to be permanently on to drive its own amp-hour conter (though there is no obvious sign of this in the app).

 

Arco say that even a difference of millivolts between the ground and the negative sense wire can damage the Zeus. Lots of electronics does like input signals below the negative supply but this is stiil something they should have protected against.

[IanD]

1 hour ago, cheesegas said:

Pylontech and similar are actually a bit different - if you were to chain up say 4 of them, the Cerbo talks to the first one only via CANbus - this automatically assigns it the Master status. This master battery then talks to the other three via a proprietary protocol and aggregates the data and any cell overvoltage etc error. That's why, for example with Pytontechs, each battery has a CAN and RS485 port (connecting anything to either port will automatically make it the master) and two Link ports to connect battery to battery. Most of them are sensitive to which Link port you use too, as some also assign master status based on the first one in the chain.

I'd strongly advise not putting Pylontechs for one on a boat - they're not designed for things that are damp and move. No conformal coating on the boards, mechanically things aren't very well fixed down inside compared to something like a Victron inverter. Seen a good few fail within the first year, and one install in the engine bay of a large Dutch barge didn't even last a month.

 

The master/slave protocol is what I described, it's easier than a multi-master approach -- it's similar to what Victron do with parallel inverters/combos.

 

I wasn't specifically recommending Pylontech, I was recommending the class of batteries that are designed to be connected in series/parallel/cascade to make one big battery bank, usually for off-grid applications because this is a *far* bigger market than boats. Whether a particular brand/model is suitable for boat use is another kettle of fish entirely -- I was looking at BYD, but it's quite difficult to find out the kind of things you mention (whether they're suitable for possibly damp/vibration applications).

 

Though it has to be said that this applies to LFP cells as well, there's a big difference between a canal boat and a yacht that might well be thumping into huge seas for hour after hour with huge shock loads -- this is why in-the-know-LFP-in-boat suppliers (almost all lumpy water specialists) often recommend 100Ah or at most 200Ah cells because they're mechanically more robust, but the big ones (700Ah Winstons in my case) are absolutely fine in a narrowboat... 🙂

 

45 minutes ago, dmr said:

 

The Zeus has two power supplies, it takes the high current alternator drive from the alternator output post, and the general electronics drive from either the battery directly or via the ignition switch. It must be low current consumption as it has an option to be permanently on to drive its own amp-hour conter (though there is no obvious sign of this in the app).

 

Arco say that even a difference of millivolts between the ground and the negative sense wire can damage the Zeus. Lots of electronics does like input signals below the negative supply but this is stiil something they should have protected against.

 

That sounds like bad design to me; the whole point of a Kelvin connection (force/sense) is that it is insensitive to ground voltage differences caused by currents in cables etc. Not that you'd expect it to cope with several volts, but millivolts sounds far too restrictive to me, normally there are back-to-back protection diodes between the grounds which you don't want to turn on so a typical spec would be <0.5V absolute voltage difference between them in operation -- the circuit should measure the inputs (voltage sense and GND sense) differentially.

Edited February 23 by IanD

[dmr]

20 minutes ago, IanD said:

 

 

That sounds like bad design to me; the whole point of a Kelvin connection (force/sense) is that it is insensitive to ground voltage differences caused by currents in cables etc. Not that you'd expect it to cope with several volts, but millivolts sounds far too restrictive to me, normally there are back-to-back protection diodes between the grounds which you don't want to turn on so a typical spec would be <0.5V absolute voltage difference between them in operation -- the circuit should measure the inputs (voltage sense and GND sense) differentially.

I agree with you 😀 If we wanted to be really pedantic about shunt wiring we could say that the Zeus ground should go to the "loads" side of the shunt, but the sense has to go to the battery post, so the sense Will go negative wrt ground by the shunt voltage, plus any cable and connection drops.

[nicknorman]

1 hour ago, dmr said:

 

The Zeus has two power supplies, it takes the high current alternator drive from the alternator output post, and the general electronics drive from either the battery directly or via the ignition switch. It must be low current consumption as it has an option to be permanently on to drive its own amp-hour conter (though there is no obvious sign of this in the app).

 

Arco say that even a difference of millivolts between the ground and the negative sense wire can damage the Zeus. Lots of electronics does like input signals below the negative supply but this is stiil something they should have protected against.

Presumably only one of the positive or negative wires is taking a lot of current, the return current path is via the alternator. Most alternators found on boats have a max field current of 4 to 5 amps. I think these regulators only have 10-12A max current in case you want to parallel 2 alternators from the one regulator.

 

But anyway, why is it important to eliminate voltage drop so as to have an accurate measure of battery voltage by having separate near-zero current sensor wires? Li battery charge voltage is very flat and so the set voltage is not really of much relevance for 98% of the charging time, since the system is a current source and current limited. It only becomes relevant right at the end, or possibly during the float phase, both of which are periods when the alternator current and hence field current have fallen right off.

 

I made a huge mistake when I first wired up my alternator controller (which sends a positive supply to the alternator rotor, the other brush being grounded) because I connected the wire which was the combined regulator power feed and voltage sense, to the battery + directly (via a fuse). Whereas the alternator output went via the Mastershunt (which has a 500A fuse) and the emergency disconnect relay before reaching the battery +.

 

It all worked fine of course, but consider what would happen if the alternator B+ beinf connected to the boat supply systems, with the sense wire, measuring battery + voltage, became disconnected from each other eg due to the emergency disconnect relay opening or the 500A fuse blowing (or some other bad connection). The sense voltage would decrease below the regulated voltage which would cause the regulator to output maximum field current. Consequently there would be massive voltage coming out of the alternator catastrophically damaging the 12v appliances.

 

Fortunately I realised my error before the catastrophe happened, and I moved the sense wire to the engine stud so that such a disconnect between alternator output and voltage sense, wasn’t possible.

 

So now I sense more or less alternator output voltage, not battery voltage. But so what, highly accurate battery voltage is not that important when there is a lot of current flowing, because charging voltage will always be way below regulated voltage until the last few minutes after the current has fallen right off. And anyway, even at 150A charge the voltage difference is only about 0.1v. Inconsequential.

Too much LA mindset going on!

[dmr]

Right, I've "wired" myself into a corner here. I've finaly replaced the less than ideal alternator output wire with a nice new one, 35mm crimped and soldered 😀 . This has significantly increased the alternator output current so its now its running much too hot even with the Adverc unplugged. So I really need to push on and get the Zeus running tomorow so that I can charge at a controlled current. (or mess about with a short term long alternator wire).

 

Here is a photo of the alternator harness. Nicely made with free cable ends for me to shorten, terminate or extend as I see fit. However the thermistor is a sealed up unit with a very long cable leading to a proper waterproof connector. This cable is much too long for me. I can probably "loose" the excess behind a wiring panel but otherwise will have to cut and join the twin cable, no great effort for people with an electronics background but might be a pain to some users.

 

The thermistor is potted in an 8mm copper tube terminal (just like Victron do but they use 10mm?). The 8mm hole will bit most batteries but those with big 10mm Lithiums will have a problem. However its not going to fit onto the alternator very nicely. Alternators often use 10 or 12mm fixings. I can probably do some careful profiling and get it behind one of the alternator body screws but this is a potential issue.

 

Crap photos, table is full of tools and wires so took these on the dogs sofa blanket.

8 minutes ago, nicknorman said:

Presumably only one of the positive or negative wires is taking a lot of current, the return current path is via the alternator. Most alternators found on boats have a max field current of 4 to 5 amps. I think these regulators only have 10-12A max current in case you want to parallel 2 alternators from the one regulator.

 

But anyway, why is it important to eliminate voltage drop so as to have an accurate measure of battery voltage by having separate near-zero current sensor wires? Li battery charge voltage is very flat and so the set voltage is not really of much relevance for 98% of the charging time, since the system is a current source and current limited. It only becomes relevant right at the end, or possibly during the float phase, both of which are periods when the alternator current and hence field current have fallen right off.

 

I made a huge mistake when I first wired up my alternator controller (which sends a positive supply to the alternator rotor, the other brush being grounded) because I connected the wire which was the combined regulator power feed and voltage sense, to the battery + directly (via a fuse). Whereas the alternator output went via the Mastershunt (which has a 500A fuse) and the emergency disconnect relay before reaching the battery +.

 

It all worked fine of course, but consider what would happen if the alternator B+ beinf connected to the boat supply systems, with the sense wire, measuring battery + voltage, became disconnected from each other eg due to the emergency disconnect relay opening or the 500A fuse blowing (or some other bad connection). The sense voltage would decrease below the regulated voltage which would cause the regulator to output maximum field current. Consequently there would be massive voltage coming out of the alternator catastrophically damaging the 12v appliances.

 

Fortunately I realised my error before the catastrophe happened, and I moved the sense wire to the engine stud so that such a disconnect between alternator output and voltage sense, wasn’t possible.

 

So now I sense more or less alternator output voltage, not battery voltage. But so what, highly accurate battery voltage is not that important when there is a lot of current flowing, because charging voltage will always be way below regulated voltage until the last few minutes after the current has fallen right off. And anyway, even at 150A charge the voltage difference is only about 0.1v. Inconsequential.

Too much LA mindset going on!

 

The Zeus has separate supplies and grounds for the alternator drive and electronics power.

It does support Lead Acid as well as Lithium and many people believe that LA's need an accurate absorption voltage.

 

Thinking about losing the positive battery sense, this is slightly related to the issue of a lithium BMS disconnect, but while that only gives a short duration inductive peak the loss of the sense could give a long duration over voltage from the alternator. I hope that the software within the Zeus is clever enough to spot this and shut down the field drive.

 

It would be good to have a lot more technical detail about what the Zeus does, I hope this is going to be available before too long. Some American companies are a bit prone to compulsive secrecy so I hope Arco are not one of these.,

[nicknorman]

44 minutes ago, dmr said:

   

Thinking about losing the positive battery sense, this is slightly related to the issue of a lithium BMS disconnect, but while that only gives a short duration inductive peak the loss of the sense could give a long duration over voltage from the alternator. I hope that the software within the Zeus is clever enough to spot this and shut down the field drive.

 

I don’t see how it could be clever enough to distinguish between a sense disconnect vs a heavy load coming from the boat (inverter etc). Max field current and (apparent) low voltage = maintain max field current.

 

One can do something like diode OR the 2 voltage sources (sense wire and alternator output) such that a sense wire disconnect only creates a mild over-voltage of a diode drop’s worth. But whether the Zeus does that is probably unknown.

 

Anyway my system uses same wire (not particularly thick) for controller supply/voltage sense/field current supply, without any ill effects.

Edited February 23 by nicknorman

[dmr]

38 minutes ago, nicknorman said:

I don’t see how it could be clever enough to distinguish between a sense disconnect vs a heavy load coming from the boat (inverter etc). Max field current and (apparent) low voltage = maintain max field current.

 

One can do something like diode OR the 2 voltage sources (sense wire and alternator output) such that a sense wire disconnect only creates a mild over-voltage of a diode drop’s worth. But whether the Zeus does that is probably unknown.

 

Anyway my system uses same wire (not particularly thick) for controller supply/voltage sense/field current supply, without any ill effects.

 

A lot could be done. I think a common scheme is to simply look for significant differences between Alternator and Battery volts. Even with a huge current anything more than a volt or so (or even less) should indicate a fault. Zeus could monitor its supply and sense wires, though it probably doesn't, there should only be a very small difference. If an alternator shunt is fitted the potential for correctly handling volt drops is increased.

 

I suspect the problem with all these things is avoiding false alarms. We've just got a new (new to us) car with all sorts of electronics including an anti-collision system. Trouble is it goes off all the time, even a big puddle upsets it, and all the beeping and flashing lights are very distracting when there is already unusual road conditions.

 

Going off at a slight tangent, as this forum does, assuming the Zeus mico controller is the bluetooth module itself then that looks like a very interesting device, a decent micro-controller with built in Bluetooth and very low power consumption. I've done a quick www search and suspect that using it might have quite a learning curve. The development board is electrically and physically compatible with the Arduino, but the software is not.

[nicknorman]

26 minutes ago, dmr said:

 

Going off at a slight tangent, as this forum does, assuming the Zeus mico controller is the bluetooth module itself then that looks like a very interesting device, a decent micro-controller with built in Bluetooth and very low power consumption. I've done a quick www search and suspect that using it might have quite a learning curve. The development board is electrically and physically compatible with the Arduino, but the software is not.

I recently have been messing about with Bluetooth BLE (Bluetooth Low Energy) modules. Whereas cooking Bluetooth just creates a single and permanent serial link, BLE is a much more intermittent, sending a “data available” advertisement which, if acknowledged, results in the data being sent in a burst. The rest of the time the transmitter is deactivated, hence the low energy bit.
 

Anyway the BLE module in question is the HM-10 which uses the Ti CC2541 chip. As well as having the Bluetooth transceiver, this chip includes an 8051 microcontroller and exposes most of the port pins, and is loaded with peripherals including UARTs, SPI and I2C interfaces, counter/timers, A/D  etc.

 

I guess the thinking is that you need a microcontroller to manage the BLE protocol stack, but it is not processor intensive so you might as well expose the micro so the end application can make use of the spare processor resources.

 

These modules are about £3 from China, and for that you get the BLE interface plus enough spare processor power to run an alternator controller. Remind me how the Zeus can cost £800+!?

[Ken X]

9 hours ago, dmr said:

 

 We've just got a new (new to us) car with all sorts of electronics including an anti-collision system. Trouble is it goes off all the time, even a big puddle upsets it, and all the beeping and flashing lights are very distracting when there is already unusual road conditions.

 

We had the same issue. Delving into the handbook revealed there was a way to turn the sensitivity up or, more usefully, down.  Problem solved.