DC-DC charge regulation in a mixed system

[SteveK]

@DDW I have not been to the boat for 14 days. After reading post 30 I went and checked the start bats for engine 12.8 & GEN 13.
This is without any charging for 15 days, the way it is usually found between visits.
What are you trying to solve with Start if yours are left charged to 100%?

 

[Jeff F]

The problem with charging the AGM after the house is done, is that could take say 4 hours, then the AGMs need another 2 - 4 hour soak. That's a long time charging.

If we accept that this is the biggest problem, what is the negative impact?

I wouldn't lose any sleep over having my AGM batteries sit for a few hours - or days - at 80 - 100% SoC, assuming they get fully charged at the end. That's not a harsh environment for batteries IMO. Even if it is, what's the implication? Somewhat shorter battery life?

I dunno. You're using a battery well within its operating range. This seems like a contrived problem to me.

 

[DDW]

@DDW I have not been to the boat for 14 days. After reading post 30 I went and checked the start bats for engine 12.8 & GEN 13.
This is without any charging for 15 days, the way it is usually found between visits.
What are you trying to solve with Start if yours are left charged to 100%?

Jeff's proposed solution was to keep the start battery at 60 or 80%, to use it as a buffer to keep the DC-DC charging the house from cycling. An AGM battery lasts a long time if it is kept fully charged, or charged frequently to full. Often under appreciated is that because of the long charge tail, this takes 4 - 6 hours charging if you are ~80% or below. There is nothing you can do to speed that time, it's inherent in the chemistry. Treated well, I routinely have them last 10 years. The set in the sailboat was replaced at 16 years and still tested at 80% of original capacity. They self discharge at a few percent per month, so left alone, they need to be fully charged (many hours...) every few months in storage. Technically speaking, keeping an AGM somewhere between 12 and 13V is "within the operating envelop". But life can suffer if not frequently charged, and severely too. This is why people are replacing AGM batteries at 4, 3, or 2 years - they have murdered them.

The proposed solution is to solve problems with the proposed topology: many DC-DC chargers charging an LFP bank from an alternator charged start bank. The idea is you would set the start voltage of the DC-DC so that it sucks the start battery down a ways, as a method to keep the alternator putting out maximum power, and the DC-DC from cycling on and off. Since the LFP has an almost flat acceptance rate, this will continue until the house is full. for example, down 600AH with 100A available, 6 hours of run time. When the house reaches full, the DC-DC will shut off, and the alternator will recharge the AGM - which requires another 3 - 4 hours to reach 100% SOC. Our daily runs are typically 4 - 6 hours. So over a month or four cruise, the start battery might never get returned to 100%. No acute damage done, but much shortened life, and starting each day with a half flat start battery. Not a contrived problem at all. When we toured the South Sound, daily runs were 2 hours. The AGM house bank never got to 100% for a month even though charged directly.

There are many alternative topologies for this mixed system (I've got schematics drawn for at least 10 variations), each has pros and cons. This particular one seems popular, but for me in my use, well down the list. May work well for others in a different use profile. I asked the question to see if there was some magic setup to overcome the issues.

The one I'm likely to implement is to charge the house, start, and thruster through a three output isolator. There are two cons of this system, one is that a single charge profile has to be used for all. But the charge profile for AGM and LFP are very similar, and for a start and thruster battery kept at near 100% SOC (as they typically are) the charge profile is pretty much identical. The other downside is that unlike schemes using solenoids or ACRs, there can be no current sharing for heavy loads. Against that it is much easier and cheaper to install, more reliable, and less opaque in operation. Every decision you make on a boat is a tradeoff, different boats, different uses, different tradeoffs.

 

[Jeff F]

Jeff's proposed solution was to keep the start battery at 60 or 80%, to use it as a buffer to keep the DC-DC charging the house from cycling.

Well, not exactly. My proposed solution is to allow the batteries to help power the chargers until SoC drops to 60-80% before interrupting charging.

This would never happen unless demand exceeds supply for some period of time, and once the house bank is charged the start battery gets taken to 100%.

I'm not advocating keeping them there, only to use available capacity as needed on a transient basis.

 

[Jeff F]

When the house reaches full, the DC-DC will shut off, and the alternator will recharge the AGM - which requires another 3 - 4 hours to reach 100% SOC. Our daily runs are typically 4 - 6 hours. So over a month or four cruise, the start battery might never get returned to 100%. No acute damage done, but much shortened life, and starting each day with a half flat start battery. Not a contrived problem at all. When we toured the South Sound, daily runs were 2 hours. The AGM house bank never got to 100% for a month even though charged directly.

I think that's the life of AGM batteries. Before solar AGM house banks for unplugged cruisers were routinely cycled between 50-90% SoC. That was considered perfectly acceptable IIRC.

If you care that much about the start battery throw a small solar panel up to keep it at 100%, or run the charger off the inverter. Or start a battery replacement fund.

And 80% SoC isn't half flat. It's 1/5 flat.

 

[SteveK]

Well I am interested in reading what the two of you are working on.
I use my start batteries as a buffer between the LFP which gets charged thru DC@DC from ALT's via the start batteries. I mentioned my engine start was 12.8 and GEN start 13 after last being ALT charged two weeks ago. I should mention the ALT's are putting out 14.4v up until shutdown as the house is not 100%.
I consider my start batteries at full or near charge, unless you tell me otherwise. The 3 second they crank the engine does not discharge dramatically and they get an immediate recharge the rest goes to the house bank.
The right way is how you want to do it after learning the options.

ETA: My ALT's are putting out 14.4v for 4-6 hours. So I would say that the start batteries are well fully charged and that voltage is sent to the house but amps are limited by DC2DC to protect the ALT's.

 

[Diep]

I've been listening in on the conversation.

I don't have a good sense of the overall DC system we are talking about here and maybe I missed it. In post #16 DDW mentions a hypothetical 25A of maintenance loads on the start bank system. Are you powering systems other than engine starting and engine electronics circuit on this bank? If you are powering a lot of potentially large and variable DC loads I can see an issue. But with a big LFP house bank, why would you not switch most/all such DC circuits to the house bank? This shifts all the "buffering" to the LFP bank and lets the effective charge current vary with the supported loads. The alts and start batteries don't feel any of it.

In this case you are left with starting bank charging and thruster bank charging in addition to the DC-DC. I would suggest that starting bank charging is complete very quickly in an ordinary situation. If you are concerned or see alt overload or cycling you can set one or both Orions with a longer delay to allow the alt to recharge the start bank first.

There is still the thruster bank charging to consider. I don't know your use case particularly well but it sounds like long term cruising trips without access to shore power. Obviously all of this is moot if you had daily shore-side AC.

How do you usually use your thrusters? My guess is: anchoring - never; departing the dock - little to none; landing at a dock - some to a lot.

So the only time the thruster bank is going to be hungry for a lot of charge is when you have just landed at a dock. I would guess that shut down is imminent and you are about to plug in. If so, all batteries are full when you depart and there is no problem.

If everything I postulate is correct, and you are able to move significant loads off of the start bank, I don't see many realistic scenarios where the alternator is doing much other than charging the start bank for 10 minutes and then supporting the Orions. If you have a rare situation like using thrusters pulling into a fuel dock and again when departing you can always temporarily reduce the orion current and/or turn them off until the house and thruster bank are back. But that seems really rare.

Just encouraging a zoom out on the situation. Are we talking about edge cases here? Making perfect the enemy of good?

 

[DDW]

In this sort of system, the demand (from the DC-DC) has to exceed the supply, or at least match it, or you are wasting charge capability. And the nature of the DC-DC is you set it for 100A and that is what it does - it cannot back off if the supply is reduced, except cut off entirely. It is a course tool. If you are willing to have DC-DC capacity well below the alternator capacity, then yes, the start bank gets recharged alongside the house charge. My assumption is that the start and thruster batteries are never less than about 98% SOC, because you never use much energy from them. But if you are going to suck them down to make the DC-DC behave, then that assumption is off.

AGM house banks were never cycled between 50 and 90%, unless you were really intent on murder. That was done with flooded. Do that with AGMs and you better have a large replacement fund. Flooded cells have the ability to accept an equalize charge to remove the sulfate created. AGMs do not, cycling them at Partial State of Charge will ruin them, over even a few cycles. Plenty of literature and testing on this is available. Read Rod Collins web articles, or Nigel Calder. Both have tested this and published results. That is perhaps the #1 advantage of LFP for the house, they actually prefer being operated at a PSC.

The maintenance load on the start bank are engine and alternator, and a few other minor things. This is an electronically controlled, common rail engine, it needs about 10 - 15A just to keep it running. The alternator field is another 10 let's say. The measured amount is in excess of 20. Moving these to the LFP house risks them in the event of a BMS disconnect so, not graceful failure. It does nothing to solve the problem of how to keep the DC-DC chargers using the maximum current available.

As far as perfection and overthinking go, yes, I am known for it - you should see my custom built sailboat. Many or most of these schemes work, the question is how well, how long, and what are the failure modes? I am starting more or less with a clean sheet here, new batteries, new alternator, new chargers. I try to do everything for a reason, and the best way possible. It's a curse really, sometimes I envy others who willingly settle for 'good enough'. But I've also seen lots of 'good enough' boats stranded in anchorages or under tow, and the envy goes away.

 

[DDW]

There is one more aspect to this. I believe in big data collection, it is the only way you really know anything. There are many charging systems out there that are sort of working, but far from optimized. But the owner doesn't know it, because they look at the voltage or amperage every half hour and the values seem reasonable. On the last boat I finally set up data logging to see what was actually going on. I had a Balmar regulator that seemed to do some odd things, but it was hard to tell. It eventually charged the battery, and nothing smoked. It wasn't until I logged the data for several full charge cycles that I saw how badly it behaved.

Does it matter? Maybe to most people with less demanding usage, or lower expectations, no. After all the goal is to enjoy the cruise, not obsess on the technology. But once I saw what was going on, I was able to improve it a great deal, and observe that it was improved. I have a hard time seeing something working badly, then leaving it alone to continue.

 

[Jeff F]

There is one more aspect to this. I believe in big data collection, it is the only way you really know anything.

I had a career in big data collection. 20 years ago I might have agreed with you.

 

[SteveK]

I updated post 36
ETA: My ALT's are putting out 14.4v for 4-6 hours. So I would say that the start batteries are well fully charged and that voltage is sent to the house but amps are limited by DC2DC to protect the ALT's.

 

[SteveK]

There is one more aspect to this. I believe in big data collection, it is the only way you really know anything. There are many charging systems out there that are sort of working, but far from optimized. But the owner doesn't know it, because they look at the voltage or amperage every half hour and the values seem reasonable. On the last boat I finally set up data logging to see what was actually going on. I had a Balmar regulator that seemed to do some odd things, but it was hard to tell. It eventually charged the battery, and nothing smoked. It wasn't until I logged the data for several full charge cycles that I saw how badly it behaved.

Does it matter? Maybe to most people with less demanding usage, or lower expectations, no. After all the goal is to enjoy the cruise, not obsess on the technology. But once I saw what was going on, I was able to improve it a great deal, and observe that it was improved. I have a hard time seeing something working badly, then leaving it alone to continue.

What data and how are you collecting, just curious because;
I currently have two sources for the house bank monitoring. The BMS monitor and a shunt monitor. While they never agree 100%, they are close. When I am at 50% SOC and they show about 400Ah left, I would say they agree enough. when they show x amps and an hour later the total amps resemble the hourly increase expected, I think it is working.
If I can manage 36 hours from the bank (so as not to completely deplete), I am satisfied. But, I still read in case I can improve.

 

[Diep]

I hear where you are coming from. However it seems to me that the other loads on the start bank are relatively consistent. It seems like the DC-DC could be tuned pretty well such that total load would not need to be too far below the steady-state alternator output.

I concede this does not make MAXIMUM use of the available alternator output. The DC-DC also has efficiency losses.

If you're really set on alternator and house bank charge maximization/optimization it seems like the starting point needs to be direct charge of the house and then DC-DC or an ACR or something to support the start/thruster banks. This way the main part of the current flows to the battery without loss and the smaller demands are skimmed off and suffer the efficiency drag.

Not trying to disagree or convince you of anything in particular, just trying to dig a bit and understand/test the priorities and assumptions.

 

[DDW]

If your AGM batteries are gettign 14.4V for 4 hours you are probably good. But the definition of full charge is 14.4V, with charge current below 0.02C - 0.005C, depending on which manufacturer. Practically, anywhere in that range will be good for any manufacturer.

I started data collection looking at amps and volts using a recording oscilloscope/meter I had. When I installed the Wakespeed to replace the Balmar, it puts out a stream of data on a serial port which can be logged with a terminal program. That takes some effort to digest, so I wrote an app that logs and graphs it in real time. I've shown this before but here it is again, below. I hardly ever look at SOC, it is a derived calculation depending on a few assumptions for values which change with time, temperature, etc. Not very accurate, and tells you nothing about charging profile. Amps and volts give you all the information available, accurately.

The engine maintenance loads might be fairly constant, but the alternator output is not. So there are several moving parts. I have no investment in DC-DC manufacturers, and no need to fit them where they don't fit well. I am struggling to see any advantage to doing it that way, vs. directly through an isolator. The fet isolators have less loss than a DC-DC, require no setup, are more or less stone ax reliable, and cost far less than the DC-DC. So let me ask, what is the argument for using a DC-DC instead? I'm not seeing it.

 

[Diep]

I think the main use case for DC-DC is with smaller stock, internally regulated alternators. The DC-DC output control serves as the regulator that keeps the alternator from burning up. In that way it is a crude tool, but effective.

If you have an externally regulated alternator the case for DC-DC is not nearly as good.

 

[DDW]

That is an argument for it, but for longer distance cruising, surely you would have an external regulator with alternator temp sensor and a large alternator? and if you have that, then the smaller DC-DC seems to be just in the way. Different use cases may have different conclusions. We use the alternator (and what small solar that can be fit) pretty much exclusively, for live aboard months at a time. I don't plug the boat in even in the occasional stay in a marina, don't need to, the DC system can easily handle a day or two at anchor or on the dock. I've turn the genset to charge only very rarely, as a result of being weather bound for several days at anchor. I'm almost never at anchor two days in a row unless there is weather, so moving every day.

So yes, the decisions I've made assume a large, externally regulated alternator.

 

[Barking Sands]

With an internally regulated alternator and isolator...what voltage will the lithium bank see? What voltage will they finish? WIll the lithium bank get the proper absorption once full? Will the lithium bank hit the necessary marks for SOC reset as determined by the BMS settings?? When full will the voltage be reduced to a desirable level? Certainly you can put back capacity with that method. But control of voltage to the lithium bank is the name of the game IMO.

With an Orion DC2DC it will provide the lithium bank with the exact profile you set. Lets just take a single Epoch Essential. In the Orion you would set 14.3 absorption, an absorption time of your choosing, say 15 minutes, and then finally the float voltage of 13.5. In the case of the Epoch 300ah Essential the BMS must see 14.3 to hit the SOC reset to 100%. So on an 8 hour cruise, you'd hit 14.3, absorb for a short period, and the rest of the cruise you would be sitting at a nice float voltage or 13.5, essentially at 100% with the voltage pressure off ready to hit anchor.

You can use your method certainly, to get some capacity returned to the lithium. But I prefer the tight voltage control the Orion provides. Thats my main preference.

ETA*** Yes externally regulated assumes you can do all of that too.

 

[DDW]

I'd agree that with an internally regulated alternator, a DC-DC to control the charging of an LFP bank is a must, not an option. But that is really true of AGM as well. The internal regulators are intended to replace a few amp hours used in starting, then maintain the the car at somewhere above 13V the rest of the time. They are not intended for, nor are they very good at, charging a house battery. If you spend much time away from the dock and don't want to run the genset all the time, external regulation is the first mod you should make.

 

[tpbrady]

I might be missing something here, but don’t DC-DC chargers appear as a load to the alternator? Saying that there must be a battery in between DC-DC charger and the alternator, seems like a misstatement. The battery is not in series with the DC-DC charger, the charger and battery are in a parallel relationship with the alternator output. The only reason for the battery is to excite the alternator and to protect the alternator from disconnect of the DC-DC charger. If this is the case how does the charging of the battery affect the alternator’s output to the DC-DC charger if demand is less than the total output of the alternator. The only way to keep the alternator from seeing an excessive demand is to put both the LFP bank and start AGM bank on DC-DC chargers who’s combined demand doesn’t exceed the alternators operating envelop or externally regulate the alternator.

I am an amateur at this so if I have made an egregious error let me know.

Tom

 

[Barking Sands]

Yeah...its just priorities I guess. I never cared too much about extracting the last amp out of the alternator for house bank charging. I have the gen and dont mind running it. Down here I have to run it for AC anyways...lol.

 

[DDW]

The DC-DC is a load, but not a simple one by any means. It has fairly complex servo feedback systems both on the input and output side. The alternator also has a servo feedback system. That makes a lot of quickly moving and variable parts trying to deal with each other. It isn't inconceivable that it would work, but a much harder problem to make it work, and I can see why Victron wants a battery in there - it slows down all the changes by a lot, a low pass filter with a lot of dampening.

Sterling makes a device that is intended to do just this, so it is possible. It loads the alternator sufficient to get a stable output, then does a boost-buck type convert to push current into the battery. The advantage they tout is you don't need an external regulator, they control the alternator indirectly by how much they load it.

I don't see it as a great solution to any problem, but it is another way of many to do the job.

 

[Frosty]

I might be missing something here, but don’t DC-DC chargers appear as a load to the alternator? Saying that there must be a battery in between DC-DC charger and the alternator, seems like a misstatement.

I can only speak to the Victron Orion XS (actually I think this applied to the older TR-Smart as well, but not positive now). They require a battery be in between the charge source and the Orion (in other words, you hook the Orion to a battery which is being charged, NOT straight to the charge source).

I can't remember why exactly anymore, but when I was helping a buddy re-do his electrical system, one of our potential setups would have had the Orion attached directly to the engine charging without the Start battery in between. Until I re-read the manual (and then asked also) and found that it was a total no-no. At least the Orion XS DC-DC charger has to be connected to a battery on the input side. Basically as a voltage buffer, IIRC.

 

[Jeff F]

AGM house banks were never cycled between 50 and 90%, unless you were really intent on murder. That was done with flooded. Do that with AGMs and you better have a large replacement fund. Flooded cells have the ability to accept an equalize charge to remove the sulfate created. AGMs do not, cycling them at Partial State of Charge will ruin them, over even a few cycles. Plenty of literature and testing on this is available. Read Rod Collins web articles, or Nigel Calder. Both have tested this and published results.

OK, assuming this is true, why not use FLA batteries for the start bank? This would seem to address your primary concern.

 

[DDW]

Flooded cells would cause many other concerns. They outgas, they have to be equalized periodically, they have to be watered. In the current installation of the thruster battery, that requires disassembling cabinetry.

Turning this around, why not just dispense with the DC-DC and use an isolator? DC-DC converters seem to be the Pixie Dust Du Jour. People want to sprinkle them in everywhere, even when seemingly simpler and better solutions exist. Given that I have an externally regulated alternator, what do I gain using a DC-DC to charge the house, vs an isolator? I've already explained at some length what I've lost.

 

[SteveK]

The DC2DC is the poor man external regulator.

 

[DDW]

Well yes, but if you have a large alternator and large house bank, you will either charge them very slowly or buy a lot of DC-DC, which starts to get expensive. In the Victron world, an Orion is about $330, with a 160A alternator you will need three, so $1000. External regulator is around $350, or $750 for the fanciest ones on the planet. Of course you can buy cheaper than Victron from Amazon or Alibaba, but I don't carry enough fire extinguishers on the boat for that.

 

[Diep]

Starting with a premise of a high amp externally regulated alternator already on board you are right, DC-DC is probably not the best answer.

DC-DC is a nice solution if you have existing low-amp internally regulated alternators.

For example, hardware for a new house bank, comprehensive modernization of high amp DC distribution, and new inverter/charger and associated AC distribution modernization for our boat is coming in around $4,500 all-in. This includes wire and consumables but not batteries. This includes 1 Orion for each of the stock alternators (2 total).

To add an externally regulated, high-amp alternator would run at least $2,000-$2,500 additional and that is after eliminating the Orions. In addition to the regulator ($750), alternator (1,000) I would either need to modify the current engine mount (Custom bracket, new idler, additional pulley to re-route belt, modify belt cover, new belt and spare) OR mount the new alternator externally (New pulley assembly for flywheel, custom brackets to mount to stringers and tension belt, new belts, modify existing belt cover)

We can run a lot of gen hours for $2,500 and even so I concede that depending on our actual power usage and cruising experience, we might still want to make this upgrade. But for $660 we will try the DC-DC route first.

The other nice thing about the DC-DC is it will allow us to harvest an additional 100 amps of house charging off the start bank charger when we are running the gen. Without that we would probably be inclined to add another AC charger to keep the gen loading at or above 50%.

 

[DDW]

DC-DC is a nice solution if you have existing low-amp internally regulated alternators.

I'd say only solution, not just nice.

But high amp externally regulated alternators aren't nearly as expensive as all that. I just bought a new Delco 28Si, 160A. It took me about 2 hours to noodle through converting it to external regulation (the next one would take maybe 45 minutes 'cause I wouldn't have to think). This alternator cost me $330. It is a much better alternator than a $1200 Balmar. It will mount in place of the existing alternator, no changes in belts, pulleys, or brackets. A Balmar regulator for $350 will work ok, but a Wakespeed is better at $750. I'll have $1080 into a much higher output alternator, with much better regulation, temperature protected, delivering (as much as) 160A. The alternative would be to keep the old alternator and put in 2 or 3 Orions at around the same cost. But then we are back to the original question of how do you set them up to work properly? That seems not possible without serious compromises.

 

[tiltrider1]

The DC2DC is the poor man external regulator.

I kind of agree with you. In my case, there isn’t room for large frame alternators. Even length is an issue as I only have 1/2” of clearance between the back of the alternator and the block. So while it was the cheap solution it was also the most logical.

 

[Frosty]

@SteveK
"The DC2DC is the poor man external regulator."

Sometimes. Some of us "poor men" have outboard-powered boats (since non-trawlers are allowed here). One doesn't really have a choice with those.