Arco Zeus regulator - first impressions and a look inside

[dmr]

1 hour ago, IanD said:

 

Beer is *always* a good idea 🙂

 

Back on topic, it would be helpful if you posted both good and not-so-good things you find out about the Zeus, it does look like an excellent (if expensive!) solution but what it's really like to set up and use (and how well it works) is what matters -- especially if there are things that they haven't thought of which are less-than-ideal for a narrowboat application, which is presumably very low on their priority list.

 

Though like a lot of Victron kit, if the issue is software/firmware not the hardware this can often be fixed with an update, assuming the manufacturers are willing to listen to users and make changes -- and it's been built so that the firmware can be upgraded easily... 🙂

 

The installation is taking a bit of effort, retro fitting/upgrading into a limited space is difficult. The captive thermistor connector is a pain as it won't feed through the existing cable hole, so had to empty out the cupboard/wardrobe and drill a new hole in the bulkhead, then thought I would tidy up some other wires.  Need to get a couple of fuse boxes from Amazon/eBay but can maybe get an improvised install working tomoro.

Will then report back on how it goes.

The LED would have been better on the top rather than the bottom.

The alternator harness was just a couple of inches too short, a longer harness would not add much at all to their manufacturing costs.

I would like a lot more detail and background in the documentation? are feature ins and "allow to charge" active lo or active hi etc.

can the "allow to charge" be shared with other things like victron?

[dmr]

The Zeus is up and running and charging my batteries right now with a limited field duty cycle to keep the alternator cool.

I have not yet installed the alternator shunt or alternator thermistor and there are in-line fuses dangling on flying leads, but my Amazon fuse boxes have just arrived  so will tidy up tomorrow.

I am struggling a bit with the app but as my backgound is PC engineering software I find all mobile phone GUI's very non intuitive and frustrating.

It does look good and comprehensive and can even do a top off charge on Lead Acids, so the Trojans can be really happy for the last bit of their lives till I go lithium.

A firmware update was painless.

Much more detail another day.

[cheesegas]

16 hours ago, dmr said:

The Zeus is up and running and charging my batteries right now with a limited field duty cycle to keep the alternator cool.

I have not yet installed the alternator shunt or alternator thermistor and there are in-line fuses dangling on flying leads, but my Amazon fuse boxes have just arrived  so will tidy up tomorrow.

I am struggling a bit with the app but as my backgound is PC engineering software I find all mobile phone GUI's very non intuitive and frustrating.

It does look good and comprehensive and can even do a top off charge on Lead Acids, so the Trojans can be really happy for the last bit of their lives till I go lithium.

A firmware update was painless.

Much more detail another day.

Great! Do you have it talking to a Cerbo over canbus? DVCC doesn't yet work with the Zeus which is a shame, it would do away with the voltage sense cables.

[dmr]

28 minutes ago, cheesegas said:

Great! Do you have it talking to a Cerbo over canbus? DVCC doesn't yet work with the Zeus which is a shame, it would do away with the voltage sense cables.

No, we are a low tech sort of boat so no bus systems, though am likely to be going that way before too long. Currently have Trojans but plan is to get some sort of lithiums that can talk to the Zeus on CANbus. Another option is to use the Victron BMV or the Zeus amp-hour counter to control "allow to charge" but I think CANbus is the way to go.

[IanD]

22 minutes ago, dmr said:

No, we are a low tech sort of boat so no bus systems, though am likely to be going that way before too long. Currently have Trojans but plan is to get some sort of lithiums that can talk to the Zeus on CANbus. Another option is to use the Victron BMV or the Zeus amp-hour counter to control "allow to charge" but I think CANbus is the way to go.

I assume you need to control solar charging too (when you go LFP), also inverter/charger -- and do you have a generator?

 

At some point, rather than trying to get disparate gear to all work together and argue about who's boss, it becomes easier to have a centralised controller like a Cerbo (or RPi for cheap DIY, but you have to then support it) to tie everything together and make it all work nicely, where the battery BMS controls SoC and the controller makes everything else do its bidding and gives you control over everything via a single unified interface-- whether this is an LCD display or a phone/PC/tablet app.

 

This can also mean you're not constrained to be CANbus-only, most controllers (and software) also support Bluetooth and/or Wi-Fi, which makes it a lot easier to add new equipment -- for example batteries with inbuilt BMS only accessible via Bluetooth, or Ruuvi sensors for monitoring temperature/humidity.

Edited February 27 by IanD

[cheesegas]

Agree with Ian above. Also, the price of a Pi4 with case and power supply plus all the USB dongles for an MPPT, ve.bus for the inverter etc is very close to the price of a Cerbo S or even a GX if you have two MPPTs. You can make DIY ve.direct cables but these lack the galvanic isolation of the Victron ones.

 

With VenusOS Large on a Cerbo GX, you also have access to NodeRed which makes the programming possibilities almost limitless. You can add 3rd party Ethernet relay boards for unlimited outputs and so on.

[nicknorman]

1 hour ago, dmr said:

No, we are a low tech sort of boat so no bus systems, though am likely to be going that way before too long. Currently have Trojans but plan is to get some sort of lithiums that can talk to the Zeus on CANbus. Another option is to use the Victron BMV or the Zeus amp-hour counter to control "allow to charge" but I think CANbus is the way to go.

I'm sure I saw you on a bus the other day?

Rather than buying some "closed source" "drop in" Li batteries, I suggest it might be better to get some bare cells and add a commercial BMS such as JBD or Overkill. Then you have full control. Depending on your usage it might be better not to use FET based switching in the "emergency disconnect" switch, you can use eg a bistable relay like the Tyco BDS-A which obviously has a better ability to cope with large current surges than a FET based system - if you can find a BMS that drives it correctly (pulse on, pulse off).

[IanD]

8 minutes ago, nicknorman said:

I'm sure I saw you on a bus the other day?

Rather than buying some "closed source" "drop in" Li batteries, I suggest it might be better to get some bare cells and add a commercial BMS such as JBD or Overkill. Then you have full control. Depending on your usage it might be better not to use FET based switching in the "emergency disconnect" switch, you can use eg a bistable relay like the Tyco BDS-A which obviously has a better ability to cope with large current surges than a FET based system - if you can find a BMS that drives it correctly (pulse on, pulse off).

 

I'd agree, assuming you're up to speed with how to balance the cells initially and are happy with installing/connecting/configuring the BMS and all the other bits needed.

 

Most BMS like this are designed to control large battery banks and talk to external controllers. If you're going to use something like a Cerbo or other Victron gear (e.g. inverter/charger), it would save you an awful lot of grief (and get more support) if you use one of the 3rd-party BMS that are known to be compatible, see list here (mine is REC-BMS):

 

https://www.victronenergy.com/live/battery_compatibility:start

DIY / Self-build oriented BMSes

The information above refers to manufactured battery packs with a supplied BMS.

Besides those, there are also companies selling seperate BMS-es intended to be used with self built lithium batteries. These are typically used when manufactured batteries with integrated BMSes don't meet the requirement of the application, or to save costs.

SUPPORT WARNING: At Victron, we offer very limited support to systems running on these custom BMS batteries.

The following BMS systems are known to correctly follow the Victron battery specification (but are not tested by Victron!):

• 123\SmartBMS - Victron instructions
• REC BMS - https://www.rec-bms.com/ (various Victron documentation on their website)
• Boostech GmbH BMS - https://www.boostech.de/bms-konfigurator/

Common issues we have seen from other DIY / Self-build oriented BMSes

Batteries using a BMS that blocks charge, or discharge current, or sets CCL to 0 when full, can trigger a number of confusing or misleading inverter/charger warnings and alarms.

For example the Victron inverter/charger reports overload conditions with little load, resulting in a lot of time and effort troubleshooting the fault, only for the inverter to work perfectly when on a test bench, and then discovering the actual cause is a custom BMS with a DIY battery.

Another common problem as a battery becomes fully charged, Victron requires the BMS to change the Charge Voltage Limit (CVL) to restrict the power flowing to the battery. If, however, the BMS attempts to restrict power using the Charge Current Limit, without adjusting the Charge Voltage Limit, this can trigger Overcharge, Overvoltage or Overcurrent warnings.

[dmr]

1 hour ago, cheesegas said:

Agree with Ian above. Also, the price of a Pi4 with case and power supply plus all the USB dongles for an MPPT, ve.bus for the inverter etc is very close to the price of a Cerbo S or even a GX if you have two MPPTs. You can make DIY ve.direct cables but these lack the galvanic isolation of the Victron ones.

 

With VenusOS Large on a Cerbo GX, you also have access to NodeRed which makes the programming possibilities almost limitless. You can add 3rd party Ethernet relay boards for unlimited outputs and so on.

 

I have resisted bus systems and other technology on the boat, its too close to been back at work, and as off grid liveaboards I don't want to be too dependant on high tech electronics though its hard to avoid. Some low tech redundancy is also good.

I would be tempted by a RPi system but finding the time is difficult. Cost saving is not everything and DIY is good, its nice to get to really inderstand things and making them is a good way to do this.

Anyway the Zeus is the first step in the bus direction.. I have some Victron stuff but the solar controller predates some of the more advanced features, and I will be getting the new DC-DC charger when it becomes available.

[cheesegas]

1 minute ago, dmr said:

and as off grid liveaboards I don't want to be too dependant on high tech electronics though its hard to avoid. Some low tech redundancy is also good.

I'm also an off grid liveaboard and had the same thoughts - however, my setup hasn't changed for just over a year now in its current incarnation and it's required absolute zero maintenance or manual intervention other than replacing/tensioning the alternator belt. 60-70a is asking a lot of a single V belt!

 

My advice is to avoid the Pi setup as it's quicker, easier and only a tiny bit cheaper than a Cerbo. 

 

What solar controller do you have? Only the seriously old/PWM ones don't have a ve.direct/ve.can port.

[dmr]

27 minutes ago, nicknorman said:

I'm sure I saw you on a bus the other day?

Rather than buying some "closed source" "drop in" Li batteries, I suggest it might be better to get some bare cells and add a commercial BMS such as JBD or Overkill. Then you have full control. Depending on your usage it might be better not to use FET based switching in the "emergency disconnect" switch, you can use eg a bistable relay like the Tyco BDS-A which obviously has a better ability to cope with large current surges than a FET based system - if you can find a BMS that drives it correctly (pulse on, pulse off).

 

Haven't been on a bus for years....except for a quick trip from Stourport to Bewdley in the summer, the bus driver was having a bad day and scared the life out of me, its safer to travel by boat.

 

Its very tempting just to buy a 500+ Ah lithium with CANbus and drop it into the batterey box but it does not use the space well....though the wasted space could take some 56lb weights to restore the trim. However if I made my own system from EVE cells or similar (plus BMS) I could just about squash in 1200Ah 😀and configure it as two batteries to get some redundancy.  I am thinking about this, its not urgent, but if I do go the DIY route will be seeking advice from you.

[IanD]

42 minutes ago, dmr said:

 

Haven't been on a bus for years....except for a quick trip from Stourport to Bewdley in the summer, the bus driver was having a bad day and scared the life out of me, its safer to travel by boat.

 

Its very tempting just to buy a 500+ Ah lithium with CANbus and drop it into the batterey box but it does not use the space well....though the wasted space could take some 56lb weights to restore the trim. However if I made my own system from EVE cells or similar (plus BMS) I could just about squash in 1200Ah 😀and configure it as two batteries to get some redundancy.  I am thinking about this, its not urgent, but if I do go the DIY route will be seeking advice from you.

 

Bear in mind that DIY may not save you much -- if any! -- money, if that's why you're doing it... 😉

 

If you compare the Fogstar prices (one of the cheapest for both cells and batteries) the 304Ah cells are £130, 8 of these to get 12V 608Ah would cost £1040, plus the BMS (£105 with CANbus) and other bits. A complete 12V 608Ah battery (Bluetooth only, no CANbus) complete with all the bells and whistles is £1250...

 

But is really meant to be standalone, not control a system or other charging sources -- the PRO one with CANbus is £1900 for 560Ah so a *lot* more expensive.

Edited February 27 by IanD

[nicknorman]

46 minutes ago, dmr said:

 

Haven't been on a bus for years....except for a quick trip from Stourport to Bewdley in the summer, the bus driver was having a bad day and scared the life out of me, its safer to travel by boat.

 

Its very tempting just to buy a 500+ Ah lithium with CANbus and drop it into the batterey box but it does not use the space well....though the wasted space could take some 56lb weights to restore the trim. However if I made my own system from EVE cells or similar (plus BMS) I could just about squash in 1200Ah 😀and configure it as two batteries to get some redundancy.  I am thinking about this, its not urgent, but if I do go the DIY route will be seeking advice from you.

It’s a sign of the times that you can get 1200ah of Li cells from the same place I got mine, for the same price as I paid for 600Ah 2 or 3 years ago - around £1600.  But 1200 Ah - that will take a lot of charging! Why not though!

[IanD]

23 minutes ago, nicknorman said:

It’s a sign of the times that you can get 1200ah of Li cells from the same place I got mine, for the same price as I paid for 600Ah 2 or 3 years ago - around £1600.  But 1200 Ah - that will take a lot of charging! Why not though!

 

Once you get up to "big enough" -- long enough to keep going between acceptable charging intervals in winter -- then more capacity doesn't help much, if (for example) it means running the engine/generator every other day for 2 hours instead of every four days for 4 hours. I'd say a battery bank that can support your boat power use for at least a couple of days is plenty, so it all depends on how much power you use. But if money doesn't matter, bigger is always better... 😉

 

(says the man who has the equivalent of a 2800Ah 12V battery -- but that's on a boat with hybrid propulsion and electric cooking, so not remotely the same case as the others here...)

 

Also depends where you're going to charge from, if an engine/generator you certainly want the batteries to be able to accept all the current these can provide, and that might dictate the size rather than Ah capacity.

 

Then there's the "should I go 24V or not?" debate... 😉

[dmr]

4 hours ago, nicknorman said:

It’s a sign of the times that you can get 1200ah of Li cells from the same place I got mine, for the same price as I paid for 600Ah 2 or 3 years ago - around £1600.  But 1200 Ah - that will take a lot of charging! Why not though!

1200Ah is over the top and a luxury but it might be nice to fill the battery box up 😀. There are two cases (that I can think off) when a big bank would be good. One is an engine failure, it would be good to have a few days in reserve to get time to go and buy a little generator, or get one delivered etc.

We did have an injection pump failure a few years ago and were very lucky that mtb was nearby and lent us a genny pretty quickly, but can't bank on that.

The second case is the real heatwave spells that are going to get more common. A drawback of a proper engine room is that it makes the boat much too hot in summer. We don't have enough solar to live on so the option of going several days without an engine run would be good. Charging should not be an issue as usually the bank would be moderately full so just need topping up.

 

Doing a propoer lithium install is a big project so the cost of an extra few hundred amp-hours is probably not a big deal.

 

 

 

Had a bit of a diversion today. While doing our first run with the Zeus yesterday the TravelPower died (burning smell from the black box but turned it off before the smoke came out) so this morning was spent getting it out and having a look inside. I can maybe fix it but think its better to drive the whole lot down to that Mr Cox tomorrow, the generator is just about due a new set of bearings anyway.

It's tempting to think that this somehow relates to the Zeus but Im convinced it's not, unless one believes that machines have feelings and the arrival of newer technology upset it.

Cheeesegas was not too impressed by the electrolytics in Zeus so here is what the TravelPower does, a set of joined together plastic plugs between the capacitors and a cable tie round the whole lot. (i've pulled the plugs out a liitle to get the photo).

[dmr]

I've just hit a very interesting problem with the Zeus. I have not installed the alternator thermistor yet so have set the maximum field duty cycle to 55% to limit the current and temperature. Performance is random with often no charge at all from the alternator. I thought I had a bad connection but its more fundamental....

I have stuck with an N type alternator install. This has the advantage that the ignition light still works  (I thought) but this is negated by using a switch-on delay and ramp up. (Note the Zeus has a "require tacho" option so the delay does not start till the engine starts which is rather nice.)

What I believe is happening is that with initial excitation coming via the ignition light and with a limited duty cycle there is just not enough current to start up the alternator 😀.

Tomorows first job is to get back into the alternator and convert to a P type system,

[cheesegas]

4 minutes ago, dmr said:

I've just hit a very interesting problem with the Zeus. I have not installed the alternator thermistor yet so have set the maximum field duty cycle to 55% to limit the current and temperature. Performance is random with often no charge at all from the alternator. I thought I had a bad connection but its more fundamental....

I have stuck with an N type alternator install. This has the advantage that the ignition light still works  (I thought) but this is negated by using a switch-on delay and ramp up. (Note the Zeus has a "require tacho" option so the delay does not start till the engine starts which is rather nice.)

What I believe is happening is that with initial excitation coming via the ignition light and with a limited duty cycle there is just not enough current to start up the alternator 😀.

Tomorows first job is to get back into the alternator and convert to a P type system,

Hmm, something's not quite with what you've said. There should be nothing connected to the ignition light terminal on the alternator at all - the Zeus controls the light's functions directly, and it's not used as a current limiter like with an internally regulated alternator.

 

There's no need for initial excitation either as I expect (like the Wakespeed) the Zeus applies a small current to the field to get an output on the tacho wire, then when it knows it's spinning, it begins to ramp up to full field current.

[nicknorman]

Yes surely the device becomes a “6 diode machine” with the field current supplied from the B+ terminal, ie direct from the battery until the alternator is spinning and making its own juice.

 

I would definitely have the w connected, otherwise the controller doesn’t know the speed. You want it to know the speed so you can have reduced max field current at low rpm, but allow max field current at high rpm (reduced if it gets too hot, of course).

 

Are you sure you have wired it correctly?

Edited February 29 by nicknorman

[dmr]

32 minutes ago, cheesegas said:

Hmm, something's not quite with what you've said. There should be nothing connected to the ignition light terminal on the alternator at all - the Zeus controls the light's functions directly, and it's not used as a current limiter like with an internally regulated alternator.

 

There's no need for initial excitation either as I expect (like the Wakespeed) the Zeus applies a small current to the field to get an output on the tacho wire, then when it knows it's spinning, it begins to ramp up to full field current.

I'm pretty sure that I'm right, and setting the duty cycle limit back to 100% sorts it out.

Its all down to doing a N install while the Zeus (and Wakespeed), are really aimed at P installs.

The Zeus has a lamp output (which is not working yet but thats another issue) but I have connected that to a separate indicator lamp which was previously used by the Adverc. I left the old ignition lamp connected to the D terminal hoping that it would still work, and it does, but the start up delay and ramp makes it a bit useless.  With an N install the Zeus has no access to the positive brush, only the negative, so it can't supply any current itself, it can only control the current flow between negative brush and ground. With a P install the Zeus does supply the current (via a positive field wire) and it now looks like this is another reason to favour the P system. I want to be 100% certain of this before pointing it out to Arco, so any thoughts????

 

 

17 minutes ago, nicknorman said:

Yes surely the device becomes a “6 diode machine” with the field current supplied from the B+ terminal, ie direct from the battery until the alternator is spinning and making its own juice.

 

I would definitely have the w connected, otherwise the controller doesn’t know the speed. You want it to know the speed so you can have reduced max field current at low rpm, but allow max field current at high rpm (reduced if it gets too hot, of course).

 

Are you sure you have wired it correctly?

 

 

 

The Zeus works nicely (almost) with a W input and there is an option that the delay and ramp-up only start when speed is detected rather than detecting ignition. This will be good for those old vintage engines that take a lot of cranking.

The linitation is that the speed resolution is only 50rpm which is not so good for slower revving engines. Don't get a Zeus if you have a Kelvin 😀.

 

The alternator remains a 9 diode machine with N wiring, but becomes a 6 diode machine in P mode???????

[IanD]

4 minutes ago, dmr said:

I'm pretty sure that I'm right, and setting the duty cycle limit back to 100% sorts it out.

Its all down to doing a N install while the Zeus (and Wakespeed), are really aimed at P installs.

The Zeus has a lamp output (which is not working yet but thats another issue) but I have connected that to a separate indicator lamp which was previously used by the Adverc. I left the old ignition lamp connected to the D terminal hoping that it would still work, and it does, but the start up delay and ramp makes it a bit useless.  With an N install the Zeus has no access to the positive brush, only the negative, so it can't supply any current itself, it can only control the current flow between negative brush and ground. With a P install the Zeus does supply the current (via a positive field wire) and it now looks like this is another reason to favour the P system. I want to be 100% certain of this before pointing it out to Arco, so any thoughts????

 

 

There shouldn't be any operational difference between N and P installs, the Zeus field current controller and the alternator field coils are in series so carry the same current, the only difference is that one goes VBAT-ALT-ZEUS-GND and the other goes VBAT-ZEUS-ALT-GND.

[dmr]

1 minute ago, IanD said:

 

There shouldn't be any operational difference between N and P installs, the Zeus field current controller and the alternator field coils are in series so carry the same current, the only difference is that one goes VBAT-ALT-ZEUS-GND and the other goes VBAT-ZEUS-ALT-GND.

 

In N mode the Zeus takes the field wire from the bottom brush and "connects" it to ground via its internal (lower) transistor, it has no means to supply current to the top brush so this must come from the diode trio or ignition light? 

In P mode the Zeus uses its (top) transistor to supply the top brush, with the bottom brush grounded and so the Zeus has full control over the current?????

Ahh I see what you are saying....in an N install, and unlike the Adverc, I should connect the top brush directly to the main supply, in effect joining B+ to D. 

I don't like that idea but not sure why not, I think going for a P scheme is better.

[IanD]

9 minutes ago, dmr said:

 

In N mode the Zeus takes the field wire from the bottom brush and "connects" it to ground via its internal (lower) transistor, it has no means to supply current to the top brush so this must come from the diode trio or ignition light? 

In P mode the Zeus uses its (top) transistor to supply the top brush, with the bottom brush grounded and so the Zeus has full control over the current?????

Ahh I see what you are saying....in an N install, and unlike the Adverc, I should connect the top brush directly to the main supply, in effect joining B+ to D. 

I don't like that idea but not sure why not, I think going for a P scheme is better.

So long as the only thing supplying the current to the field coil is the Zeus, it makes no difference which scheme you use -- this assumes that either both field coil terminals are accessible, or that one terminal is connected to GND (P) or VBAT (N). That's what the Zeus manual seems to show anyway... 😉

[cheesegas]

39 minutes ago, dmr said:

I'm pretty sure that I'm right, and setting the duty cycle limit back to 100% sorts it out.

Its all down to doing a N install while the Zeus (and Wakespeed), are really aimed at P installs.

The Zeus has a lamp output (which is not working yet but thats another issue) but I have connected that to a separate indicator lamp which was previously used by the Adverc. I left the old ignition lamp connected to the D terminal hoping that it would still work, and it does, but the start up delay and ramp makes it a bit useless.  With an N install the Zeus has no access to the positive brush, only the negative, so it can't supply any current itself, it can only control the current flow between negative brush and ground. With a P install the Zeus does supply the current (via a positive field wire) and it now looks like this is another reason to favour the P system. I want to be 100% certain of this before pointing it out to Arco, so any thoughts????

It functions exactly the same whether in P or N mode, as Ian says - the same current passes through the regulator in either, and the same current flows in the same direction in the field winding. The alternator doesn't know any difference what mode the regulator is in; the field winding is totally independent of the rest of the alternator when externally excited. I'd suggest that you remove the lamp connected to the D terminal, but I do think the internal regulator may not be completely disconnected by the behaviour of the lamp connected to the D terminal you describe.

[dmr]

13 minutes ago, IanD said:

So long as the only thing supplying the current to the field coil is the Zeus, it makes no difference which scheme you use -- this assumes that either both field coil terminals are accessible, or that one terminal is connected to GND (P) or VBAT (N). That's what the Zeus manual seems to show anyway... 😉

 

I think you are correct that Arco intend the N system to have a live feed added to the top brush, this will explain why they suggest putting a fuse in the field wire, I have been puzzling over that. However unless some physical layout of the alternator internals makes a P install difficult I can see no advantage at all in an N install with a live feed to the rotor, and the P arrangement becomes the obvious way to go.

[nicknorman]

29 minutes ago, dmr said:

 

In N mode the Zeus takes the field wire from the bottom brush and "connects" it to ground via its internal (lower) transistor, it has no means to supply current to the top brush so this must come from the diode trio or ignition light? 

In P mode the Zeus uses its (top) transistor to supply the top brush, with the bottom brush grounded and so the Zeus has full control over the current?????

Ahh I see what you are saying....in an N install, and unlike the Adverc, I should connect the top brush directly to the main supply, in effect joining B+ to D. 

I don't like that idea but not sure why not, I think going for a P scheme is better.

I’m going to go upstairs and check on the big screen computer, but this sounds wrong to me. I would expect the field circuit to be: from B+ direct to the brush. From the other brush via the regulator to ground. The D+ and field diodes are not used. It becomes a 6 diode machine. The alternator warning light would be operated by the regulator.

Just now, dmr said:

 

I think you are correct that Arco intend the N system to have a live feed added to the top brush, this will explain why they suggest putting a fuse in the field wire, I have been puzzling over that. However unless some physical layout of the alternator internals makes a P install difficult I can see no advantage at all in an N install with a live feed to the rotor, and the P arrangement becomes the obvious way to go.

I certainly have a P type circuit on my setup, the other brush connected to ground. But I would still fuse the wire from the B+ to the regulator.

 

I don’t think it makes much odds whether you use P or N. In my case, the reg chip was P only and one brush was already connected to case so it was a no-brainer.