Update our our update to our lithium installation

[booster]

Yep, it has been a while since I put anything up about what is going on. That is mainly because I really didn't know where it was all going .

Bottom line is that the initial results looked like the system addition of the Wakespeed alternator regulator was going to work well, but when it got to the details of accuracy, repeatability, and stability, the Wakespeed fell quite a bit short for our application.

The system was a two alternators in parallel setup with both alternators running off the Wakespeed regulator. A shunt is used to control the amps out of the Wakespeed and end charging appropriately. Sounds simple and our old AGM setup worked fine that way, but the Charles/Amplepower regulator was not adjustable enough to handle lithium batteries.

Without going in to too much detail, here are the major issues encountered in trying to get it to run properly.

The Wakespeed uses a "sense" wire to activate it like many alternators do, but the way they have it set is that it requires a 12v signal to start. The sense wire on modern vehicles is not a 12v signal like the old points or early electronic ignition modules were, it is a 5v data level signal from the PCM. Our van is a 2007 Chevy so surely not a recent change from the 12v signal days that commonly came from the oil pressure sending unit to prove actual running. Wakespeed says to go to the key "run" position to get the 12v but that doesn't confirm engine running. Big hazard in that in that if the key is on but the engine not running, the regulator will go to full field current and burn up the regulator very quickly. Many diagnostics are run in "key on/engine off" mode so asking for problems there. The only 12v switched, and confirmed running, source I could find was the fuel pump power so I tapped into that and it worked fine. Newer vehicles with variable speed pumps might not be able to do that and neither would those that switch ground instead of power to the pump. I lucked out there I think. The real bummer is that the Wakespeed has an input connector to a stator tap which the could use to confirm actual running, but they don't use it for that which makes little sense to me.

The next major thing was that the Wakespeed just didn't like the system without the coach batteries connected. It would still see the starting battery, but it had problems controlling the voltage even though it was told to not try to cause a reading in the shunt by programming in zero amps when in float. I finally just decided to take the second alternator to being a standalone to simplify it all. The system was designed and built to be able to do that easily so it took about two hours, only.

It ran much better as a standalone, and appeared that it would all work out until I did the actual accuracy testing on the charging. The goal was to charge to a voltage and tail amp setting to terminate charging at a point under true 14.6v charged full lithium batteries to help preserve life and capacity. Since the tail amp control is a "must be under" test it was pretty much doomed to fail in a system that runs at various speeds, like a vehicle does. It would be charging away, but when I slowed down the amps would drop and it would think it was done and go to float. This is pretty much going to happen to all systems like this because they even have settings to intentionally turn down the alternator charging when at idle or low rpm to prevent alternator overheating from it trying to maintain the high charge rate. On a setting of 60 amps or lower it would often trip out at over 100 amps and leave the charging short of where we wanted to be.

I really, really didn't want to use a DC to DC charger because there really aren't any that I could find that were settable enough to be able control charging to stay in the mid SOC range accurately, plus they are very large and relatively inefficient in the higher amp version.

Finally, I remembered that the Victron battery monitors had a switched output built into them that could be used for various things so I decided to take a look at them to see if they could be made to do what I wanted to do.

As it turns out, the Victron monitors will do what I want very easily and the testing indicates they also do it very well. Using one actually even makes the system totally automatic if I want and will only come on with the engine running and SOC having dropped below 40% of our arbitrary full point which is at about 90% of full capacity. It will charge every time the van is running until it hits 90% SOC of our full setting and then shut off the regulator until the SOC is again under 40%.

The relatively obvious issue that could come up is defining the arbitrary full point and consistently being able to get to it accurately and synchronize the Victron to that point as it will drift a bit due to charge/discharge efficiency differences. We can charge to that reference point accurately with the shore power charger, which is a Magnum MS2000 running on an ARC-50 remote. We charge to 13.6v and 53 amps tail current and we are at reference point. I just do a many synchronize on the Victron and good to go. Based on what we saw on our last fall test trip with the old regulator system, the monitors are not going to drift much at all because of the high lithium charge efficiency and relatively low C rate charging. Maybe needed every 30 charges which would still be months of use continuously. We only need to charge every 5 days or so to stay within the 40-90% range. My guess is that I will only being doing the referencing when the batteries have been disconnected for a while during storage which is usually once or twice a year.

The proof or not of it all will come when we get out in the real world of use as you can't test for stuff that you don't know is going to happen

We have a thee week spring trip coming up that we probably never bother to plug in for at all, like we did our fall test trip. Exception would be if it gets horribly hot, which can happen where we are going that time of year. in the last 5 years we had 3 that were very hot and two that were nicely cool. 15 years ago it never got all that hot on the same trip.

[booster]

The description of the system can get long, but the control wiring to the Wakespeed regulator and Victron battery monitor is very simple and basic but still gives all control needed. All the rest of the wiring is the normal way that remote regulators and monitors are wired, so nothing odd there.





Basic on/off the regulator control, a leg to turn on the regulator from the Victron, a second leg for a manual override to charge even in the Victron is not in charge mode, a time delay to keep the van starting motor kickback voltage from faulting the Wakespeed.

[booster]

The operation of it all is just based on 4 reference points/settings in the various bits and pieces of equipment.


1. 100% full lithium batteries per SOK spec of 14.6v with 15min hold time.


2. 90% of full lithium batteries determined by discharging 62ah form the full 618ah pack. Empirically test to be able to charge back to this point with the Magnum shore charger.
This point is also used as the synchronization point for the Victron monitor to read 100% (90% of total full batteries).



3. 90% reading of SOC on the Victron (really 90% of the 90% full the Victron calls 100% and 81% of totally full batteries). This is the point that the Victron will shut off the regulator when charging.


4. 40% reading on the the Victorn (really 40% or the 90% full the Victron calls 100% and 36% of totally full batteries). This is the point that Victron will turn on charging by the regulator.


That is all that the system does, so pretty simple of a way to stay in the midrange of SOC all the time with alternator charging, even with no operator intervention if desired.

[avanti]

which Victron monitor are you using?

[booster]

Originally Posted by avanti

which Victron monitor are you using?

I put in a BMV 700 first as I didn't really need the Bluetooth and it worked just fine for all the controlling. What it didn't have was a feature that I thought was in all of the monitors, but isn't. Only the BMV712 can be set to scroll whatever readouts you have set for display on the single line screen. Since this is going to be used as an evaluation tool for at least a while, having it scroll is way safer than constantly pushing buttons to get the next reading while driving.


For that reason, I picked up a BMV 712 and put it in, which was very easy as all the other parts stay the same. The Bluetooth does make the programming easier by a whole bunch also. The bonus I got is that I can easily turn on the display full time from the phone instead of going through the many layers of settings via the buttons. I need it on when driving with sunglasses on in the bright daylight as the lcd screen is nearly impossible to see in those conditions without backlight.

[@Michael]

Are you using the programable relay functionality as the switch that controls the Wakespeed?

[booster]

Originally Posted by @Michael

Are you using the programable relay functionality as the switch that controls the Wakespeed?

I am using the programmable relay connected to the Wakespeed, but it doesn't really control anything in the Wakespeed except turning it on and off at the right times to control charging on and off points. The Wakespeed independently controls the charging voltage and amperages and is set to parameters that make it not go to float in the time it takes to do our charging. Basically, the Wakespeed is working as a voltage controlling, current setting device. It runs at 13.7v and 120 amps to the batteries and that is shunt controlled so no load affects at all on it as it will just increase output to hold 120 to the batteries. At idle it dials back to about 60 amps to reduce load on the engine and keep the engine and alternator cooler.



Separating the controls to the units that do them best is what made the system work well, as the Wakespeed couldn't do it by itself with any reasonable accuracy.

[@Michael]

Thanks.

It looks like I can do something like that - use relay outputs on my Cerbo GX to turn on and off my Redarc B2B. Right now I'm doing it manually using a Bluetooth controlled SSR.

FWIW - I ran into an interesting issue when running internal battery heaters this winter. Victron shore power was fixed at a constant 13.2v. Current for the heaters came from shore power through the Victron shunt. The Victron BMS was fooled into thinking the battery was 100%, while the internal BMS was fooled into thinking the battery was drained. the actual state of the battery was unknown - but at 13.2v, likely pretty well charged.

[booster]

Originally Posted by @Michael

Thanks.

It looks like I can do something like that - use relay outputs on my Cerbo GX to turn on and off my Redarc B2B. Right now I'm doing it manually using a Bluetooth controlled SSR.

FWIW - I ran into an interesting issue when running internal battery heaters this winter. Victron shore power was fixed at a constant 13.2v. Current for the heaters came from shore power through the Victron shunt. The Victron BMS was fooled into thinking the battery was 100%, while the internal BMS was fooled into thinking the battery was drained. the actual state of the battery was unknown - but at 13.2v, likely pretty well charged.

You likely can do that with the Cerbo, as using a Cerbo with the Wakespeed is Wakespeed's preferred method unless the battery BMS will communicate directly with the Wakespeed.


Things like you mention with the fooled BMS vs BMS thing is, I think, an example of the how easy it is for all this stuff to interfere with each other when not a complete system like a total Victron mega cost one.


I don't know if you have a Victron monitor on the system also, or if the Cerbo can do the same things, but you may want to try to find a reliable, and repeatable way to find a reference point for the SOC. Totally full charge to 14.6v is probably the easiest, but lower SOC points can also be done like we did. Once you have an easy to get to and use reference point, correcting any weird glitches that leave you with unknown SOC gets way, way easier to do.

[booster]

Michael, I just reread what you said about the unknown SOC and some of it is a bit confusing to me.


First is that you said the heaters were from shore power and going through the Victron shunt, but in most systems that is impossible by the definitions of where the connections on the shunt go to AFAIK.


The shunt should have all the 12v loads negatives tied to the load side of the smart shunt and the batteries, and only the batteries go to the battery side of the shunt.


The shore power feeding 12v heaters would go through the charger to the heaters and come back on the common ground to the load side of the shunt, which also includes the negative from the charger 12v output. None of the current should go to the battery at all under those circumstances if it is really balanced as you think it was.


There is also the question of two BMS units in the system. What are they controlling? Normally the BMS would be for the lithium pack control and protection and can also in more and more cases have real time communication through a Cerbo or other device to turn on and off various other components like the b to b, solar, or shore charger. I don't know how two BMS units would work together, especially if they are communicating with a master control. Perhaps the second BMS was actually a Victron BMV monitor coming of the smart shunt? In that case with the proper programming of the system both would likely work together fine as they would be used just for data input and system would decide what to do and then execute it.

[booster]

As I mentioned earlier, there is nothing special in the installation of the upgrade. Not a lot changed visibly although I did do a whole lot of wiring cleanup and replacing as the system has been changed a lot over the years and needed to get cleaned up.


The AGM 440ah bank was hung underbody just behind the rear axle with the cables coming up through the floor on the passenger side just forward of the electronics/charging area. The lithium pack is inside the van and on the driver side directly behind the left side rear bolster which ends at the front of the driver side wheelwell. The batteries are right next to the wheel well and are 3 SOK 206ah 12v batteries. The cabling to the electronics and charging now runs across the van at the main bedframe crossmember and sits in an aluminum channel bolted to the frame and open upwards where it will be right under the bed plywood deck so not in danger of damage.


Here are the batteries, associated cables, Blue Sea remote disconnect, inductive pickup for the old amp/volt meter we use to see real time amps and volts at the batteries from the cab while driving and charging.





Here is the channel across the van to the electronics area





Here is the mostly unchanged electronics and charging area that has been in use for nearly 10 years now. The only change is in the electronics area under the charger where I had to replace our normal shunt with the Victron smart shunt. You can tag other devices to the shunt also and we have 3 of them stacked on top of the Smart shunt circuit board. The shore charger, solar controller, and the Wakespeed regulator all read amps from the shunt and control from those readings.





The dash cleaned up a bit as I could remove the larger panel I used to control the old Charles/Ample Power regulator and just put 3 switches directly into the dash cubby along with the inductive meter, the remote disconnect switch, and some 12v outlets for the GPS and rear camera.





The inductive meter is on the left.


The first, red, switch is the main switch for the regulator initiate wiring. It is fed 12+ from the fuel pump circuit only when the engine is actually running so regulator can't start up without it running. The switch lights up red when this switch is on which is needed because this switch can be on without the regulator being activated if the Victron is not calling for charging.


The second, blue, switch is used as an override switch so I can turn on the regulator manually if I want to when the Victron has not yet gotten to the recharge SOC point. Handy to do a quick charge addition on a short trip the store or such if desired. The blue switch lights up whenever the regulator is initiated and running whether it is because of the Victron calling for the charging or the override switch calling for the charging. You can tell which it is by looking if the blue switch is on or off.


The third, green, switch is our old and still active electric cooling fan control switch. It is hardly ever used since I reprogrammed the torque converter lockup years ago. It lights up green when the fans are on.


The larger red switch that is next is the remote battery disconnect switch which is now only used as an indicator that the batteries are connected or not. I disabled the remote switching by removing the connections for the switching at the disconnect. The dash switch shows a red light when the batteries are connected and no light when they are disconnected. We now use it as a manual only switch which we can reach through the rear doors.


The Victron monitor is velcroed to the doghouse top cup holder tray and faces the driver. It is that way so it can be easily pulled off the doghouse cover tray so the tray can be removed and then the doghouse taken off if needed for engine service of some sort.

[Luv2Go]

Originally Posted by @Michael

Thanks.

....

FWIW - I ran into an interesting issue when running internal battery heaters this winter. Victron shore power was fixed at a constant 13.2v. Current for the heaters came from shore power through the Victron shunt. The Victron BMS was fooled into thinking the battery was 100%, while the internal BMS was fooled into thinking the battery was drained. the actual state of the battery was unknown - but at 13.2v, likely pretty well charged.

Hi Michael:

I glanced at your blog posts and schematic and now understand how this is happening. I was considering batteries with built-in heaters for our new build and rejected them due to other criteria and this is another reason to use external heaters. With these batteries the heater negative is internally connected to battery negative so it will always be on the wrong side of the Victron shunt as booster pointed out.

There is a possible solution, the SFK batteries now offer a RS485 option which will hook up to the Cerbo GX through USB with a data cable SFK makes (Link to a video and the Cerbo GX driver on their website here). The Cerbo GX would then be set to use the state of charge from the SFK battery instead of the Smart Shunt. Not sure what the RS485 retrofit to your battery involves.

[Luv2Go]

Hi booster:

Great series of posts! I'm in the middle of specifying components for our new Sprinter build and the info regarding the Wakespeed regulator could not have come at a better time. There is a relatively new alternator regulator on the market, the Arco Zeus. They say they have solved some of the issues that their competition has, I'll have to look into it more to see if they were able to resolve this one.

[booster]

Originally Posted by Luv2Go

Hi booster:

Great series of posts! I'm in the middle of specifying components for our new Sprinter build and the info regarding the Wakespeed regulator could not have come at a better time. There is a relatively new alternator regulator on the market, the Arco Zeus. They say they have solved some of the issues that their competition has, I'll have to look into it more to see if they were able to resolve this one.

I think I may have even mentioned the Zeus earlier here or in a different post. I have contacted them to get some detailed information but once I got very detailed of function they quit responding, which is too bad because the installation manual doesn't cover how it actually works. I do know it is a lot like the Wakespeed in its functions and even claims to be a Wakespeed type regulator without the Wakespeed problems, which I thought was a novel way to put it. They have bluetooth in it, which is a big deal compared to Wakespeed which you have to hardwire, but always remember to disconnect or it keeps the regulator active even without the engine on if the phone is attached, which is really dumb. I couldn't get an answer from Arco on what voltage is needed on the sense (activation) wire. The WS needs at least 9volts which meant I couldn't use the existing PCM sense wire. They say to use a key on sense but then if the key is left on the regulator is still working even with engine not running and can take the non rotating alternator to full field and fry it. I moved my sense wire to the fuel pump feed 12 after the fuel pump relay but before the pump so it now only comes on with the engine running.


How well a Wakespeed or a Zeus will work for you is somewhat dependent on what state of charge targets you will be shooting for. It gets easier if you are going to the full 14.4-14-6 volts all the time because you can use voltage and hold (usually for balancing) time and the Wakespeed or Zeus should do that well. Where they don't work well is if you are trying to stay mid SOC range all the time and only going full if you need balancing (which we still don't need after a year of use). The varying rpm of driving tends to make the WS, and probably the Zeus end charging at the wrong SOC because the amps change with engine speed, especially is you load shed intentionally at idle to keep the alternator and engine happier.


Combining a WS or a Zeus with a fairly inexpensive Victron battery monitor can cure most SOC accuracy issues in most any system with the WS or Zeus, I think. As mentioned above it put us in a very usable setup that lets us set any SOC range to stay within.


If the Zeus had been out when I collected all the parts for ours, and I knew all the issues that the WS has, I would have gotten the Zeus to try.


I still believe that a remote regulator on a standalone high amp alternator is a great setup for charging batteries when compared to B to B chargers. The regulators can set any voltages you want and give you exact charging to the batteries because they measure amps at the batteries. They also allow you to load shape the output of the alternator by rpm so you can cut amps back at idle and low speed, which I think is good feature. We charge at 120 amps and idle at 545rpm so have low alternator speed. If I don't cut back the output at idle it will take the regulator to full field current at idle which is a recipe for an overheat, or at least a turn down from the temp sensor for the regulator. It is also a drag on starting acceleration and smoothness.


It will be interesting to see how you finally decide to set it all up and how it works once you are done. I know ours has been a very good education as I have been able to see first hand how the batteries, regulator, alternators and the rest of the systems work, or don't, together well.


I think you pinpointed Michael's power and SOC issue.



I have never understood the whole deal with internal heaters that work of the batteries themselves as it makes no sense to me. Wire them to the outside of the battery so you can warm the batteries from shore power or engine running easily. 12v or 110v probably doesn't matter much if the connection is outside as with a charger and an inverter you can supply both for outside sources. I made sure I ordered our batteries without heaters so no chance of issues. We probably will never be in a situation when they are needed anyway, based on our 15 years past history of travel, even though we live in Minnesota.

[@Michael]

Originally Posted by Luv2Go

With these batteries the heater negative is internally connected to battery negative so it will always be on the wrong side of the Victron shunt as booster pointed out.

There is a possible solution, the SFK batteries now offer a RS485 option which will hook up to the Cerbo GX through USB with a data cable SFK makes

SFK will send me a BMS with the RS485 enabled for a small fee. Mine has the port, but I've not tested it to see if it's lit up.

I don't think this will help though. The situation was that data from the internal BMS (via the UART & Bluetooth serial adapter) showed a low (20%) state of charge when the battery had been held at 13.2V for a couple of months. I'm pretty sure that the RS485 data would show the same.

As to why, I'm not sure. Apparently, a small fraction of the current drawn by the pads is showing up in the BMS's internal shunt, causing a slow downward drift in the SoC.

A solution is to have internal, 12-volt heating pads powered externally from the far side of the shunt. Then that current would be accounted for in the external shunt. The pads would function when on battery or on shore power.

I can modify mine by putting a connector in the lid and rewiring the heat pads. I'm not too concerned about it though. It's a minor issue, more curiosity than anything.

[booster]

Originally Posted by @Michael

SFK will send me a BMS with the RS485 enabled for a small fee. Mine has the port, but I've not tested it to see if it's lit up.

I don't think this will help though. The situation was that data from the internal BMS (via the UART & Bluetooth serial adapter) showed a low (20%) state of charge when the battery had been held at 13.2V for a couple of months. I'm pretty sure that the RS485 data would show the same.

As to why, I'm not sure. Apparently, a small fraction of the current drawn by the pads is showing up in the BMS's internal shunt, causing a slow downward drift in the SoC.

A solution is to have internal, 12-volt heating pads powered externally from the far side of the shunt. Then that current would be accounted for in the external shunt. The pads would function when on battery or on shore power.

I can modify mine by putting a connector in the lid and rewiring the heat pads. I'm not too concerned about it though. It's a minor issue, more curiosity than anything.

Any current that is used inside the battery won't show up on the battery monitor, but will be missing from the battery itself, so there is going to be a mismatch on any system that has internal BMS when compared to an external shunt based battery monitor.



The most common fix for this, if one is worried about it, is to get rid of the internal BMS and install and external one so all the current used by the BMS is captured if the BMS can be wired to catch it. Since that current might be on BMS circuit board, it may not be available there but it should show all current in the power and ground to the BMS wherever it went. Of course, heaters also have to be moved to outside the box control as you proposed.


The limited amount of time sitting testing I have been able to do on our batteries, which are SOK internal BMS with no heaters, has shown the sitting over time loss, based on voltage only measured at the batteries with a 5 good digit multimeter, is still less than an AGM loses sitting. If that is correct I am not going to worry about it, but will need to do a resynchronization of the external monitors when it comes out of it's off duty time, and I have that cycle in our shore charger. I have had that built into my procedures for a long time because the AGMs also needed it. The SOK BMS goes to sleep after a settable amount of time so uses very little power at that point. I don't know if other brands do the the same.


I can't speak for other brands, but the internal SOC measurement in the SOK batteries is pretty useless as it only calibrates correctly if you go from high cutout voltage and then all the way down to low voltage cutout and then back up. Useless for my use, so I don't worry about it, and almost all other SOK users say the same thing.


You probably also my find that the BMS reported voltages, both for whole battery and cells will be quite accurate and repeatable, but the amps are not very accurate at all and not all that well suited for determining amp hour use. It makes sense that on a small internal BMS you don't have the room for a decent shunt to measure accurately, I think.


If I were building from scratch, based on what I have seen now, I would use an external BMS that had fully adjustable active balancing so it could be set at the resynchronation point. That point would only be used occasionally as we cycle at a different high point. Controlling the daily charging off the external monitor is good way to have SOC controlled charging that is probably more accurate than the BMS units can give unless they are very good ones.