nicknorman

3.60v/cell I think. ie that is the BMS disconnect voltage. Which is perhaps a little cautious but probably a good thing.

Not totally sure, I think it is an epever PWM controller. Very cheap, but it has the interface for the MT50 display which can be used to adjust all the settings. I have been thinking of getting an MPPT controller but there is not a huge advantage when the Vmp of the panel is only around 16v

My point was a general one and not specifically all about me(!) - the point being that charging to a specified limit voltage (which is what solar controllers do) can result in either under or over charging. The latter being more of an issue for Li. Of course it is not an issue for LA. But to continue with my particular issue I have a 100Ah Li battery and a 100W solar panel. So even in summer on a sunny day (in Scotland) it only is likely to charge at about 5%C and in winter much slower, 1%C if I’m lucky! In winter I use it to run the telly and Sat box and the electric blanket for Friday evening and Saturday evening. So that probably takes out 40% of capacity. Meanwhile it may get between zero and 5% back in during the day in winter. So I leave on Sunday with it quite low and hope that when I return on Friday evening it will be replenished - otherwise I need to get the genny going. Then as we move towards spring, it gets more charge each day and at some point - varying each week according to the weather, it is going to get to 100% during the week and I am going to be using much less power of an evening. So there is a risk of it being overcharged. All of which is simply to say that solar controllers are not optimised to Li charging.

Having thought about this a bit more, I can see a problem. Li doesn’t like to be overcharged. As we know, the correct way to fully charge Li is to charge to 3.65v /cell and hold that voltage until the current falls to 5%. Then disconnect. If you hold the voltage 3.65 with the current below 5%, you are still adding charge to a battery that is already at 100% of its design capacity. Aka overcharging, which risks Li plating and accumulating permanent damage. With solar, the charge rates can be very low especially in winter or if you have a large bank and small panel. Certainly less than 5%. Let’s us hypothesise at 2% charge in winter - eg 4A for a 200Ah battery. If your solar controller is set to charge to 3.65v/cell (14.6v) then by the time you reach 14.6v, the battery has been overcharged just the same as if you charged at 50%C to 3.65v and then waited for the current to fall to 2%C. And of course it has been held “up the knee” for a long time. Which will shorten its life. It will probably still last many years, but not as many as if you hadn’t overcharged it. So even with solar, correctly charging Li is not easy. For my caravan battery (which is left unattended without significant power drain for days or weeks at a time), I have set the solar to 13.9v with very short absorption time, and float of 13.25v hoping that that is not going to result in the battery being held “up the knee” too far every day. But it is not ideal because when I am using the caravan in summer, it will take longer to recharge at that voltage.

Groundhog day.

True. However most boaters charge their batteries via an alternator at some point. And for many, it provides most or all of the charge.

On the other hand it also has to be said that IanD has no experience with fitting out a boat with lithiums. He paid someone else a lot of money to do it for him. Although I can see why you might think otherwise.

Well yes but your “pure resistor” is an imaginary concept that doesn’t exist in reality. So ohms law is very good at predicting the behaviour of an imaginary construct that doesn’t exist in the universe. Very useful (not)!

Yes. Apply some voltage to a resistor, what is the instantaneous current? You can’t work it out using ohms law because it varies. Shot noise. And thermal noise. So yes ohms law kind of works in an imprecise way at a cursory glance, but in a detailed way it doesn’t work. Then there is quantum tunnelling.

I blame DMR. I was going to let it drop because there is no point in arguing with a narcissist, but then Dave came along and drew some parallel with Gibbo, so I felt morally obliged to continue.

Have you ever thought of going into politics? You certainly have the prerequisite of a selective memory. As I explained before, if the thin wire is uniform they you are correct. But it isn't. As I explained, the "thin wire" circuit is from the shunt to the BMV, from the BMV to the BMS, and from the BMS to the cell 0v. All of which are reasonably similar and adequate cables that won't melt. However within the BMS there would be (if I hadn't designed it out) a PCB track connecting the BMV and cell 0v wires. A thin PCB track which at first glance would not be required to carry any more current than the sub-mA current arising from the TTL level 232 data lines. So likely very thin and thus concentrating nearly all the circuit resistance, and hence nearly all the heat, in a very small area with virtually zero thermal mass or ability to disspate heat. And even if the fat connection did get very hot and melt a bit, that is easily fixed. Whereas internal damage to the BMS pcb would require a new BMS. And then we have the problems that you are choosing to forget - a disconnection or near disconnection of the battery -ve which would keep services live but pass all the current through the alternative route and vapourise it. And the fact that the mV accuracy of the 0v BMS line would be compromised by even relatively small currents following the alternative path. All of which comes back to my point that it is bad practice to design systems that have alternative and "unintended" current paths. Also important to bear in mind that FMEA is only as good as the failure modes you think of. Building in weak points/bad design even though you can't immediately think of a scenario where it might cause a problem, is still bad design. If you are a good engineer, that is.

Shock horror you didn't invent it? I am amazed. Anyway as we both know Ohm's and Kirchoff's laws have long been discredited for being inaccurate.

There is no morse code for inverted commas. Or a question mark. Or a slightly smiley emoji. Nor is there any distinguishment between lower case and upper case. In fact do you know ANYTHING about morse code? Apart from the fact that you invented it, obviously.

Yes this is true, and not just modern diesels. But not the sort of diesels found on narrowboats. I just asked an AI chatbot why. This is the answer: Modern diesel engines may still feature throttle bodies, primarily for emissions control and Exhaust Gas Recirculation (EGR). The throttle body’s butterfly valve can be controlled by the ECU (Electronic Control Unit) or a solenoid-controlled vacuum line. Under low engine load conditions, the ECU restricts the amount of fresh air (from the turbocharger) entering the intake manifold. This creates a pressure differential, aiding the EGR system by allowing more exhaust gas to be recirculated. During engine shutdown, the throttle body acts as an Anti-Shudder Valve (ASV), ensuring a smoother shutdown. I also asked it why some BMW petrol engines don't have a throttle valve. This was the answer (quite useful these AI things!) Certainly! Some BMW petrol engines are designed without a traditional throttle valve, and here’s why: Valvetronic Technology: BMW employs an innovative system called Valvetronic in certain petrol engines. Unlike conventional throttle bodies, Valvetronic eliminates the need for a butterfly throttle valve. Instead, it electronically adjusts both the lift and timing of the intake valves. How Valvetronic Works: When you press the gas pedal in a Valvetronic-equipped BMW, you’re altering how much the intake valves open. The Valvetronic system varies valve lift instantly, allowing precise control over engine breathing. By eliminating the traditional throttle valve, BMW reduces pumping losses. Pumping losses occur when the throttle restricts airflow during partial load conditions, wasting energy. Benefits of Valvetronic: Improved fuel efficiency: Valvetronic optimizes air intake without relying on a throttle plate. Enhanced power delivery: The system adjusts valve lift dynamically based on driver demand. Drive-by-wire throttle action: Valvetronic operates electronically, providing a seamless response. Production and Impact: Valvetronic-equipped engines are manufactured at BMW’s Hams Hall plant in England. These engines enhance performance, reduce emissions, and contribute to BMW’s reputation for innovation. In summary, BMW’s Valvetronic technology replaces the traditional throttle valve, resulting in more efficient and responsive petrol engine

Yes. Certainly at idle I would have thought a diesel produced virtually no CO. Although 100% more is only twice as much, I would have thought it was much more.

No, massive difference between a petrol and diesel engine at low power/idle. The petrol engine has the throttle closed, this severely restricts both airflow and fuel - the throttle is typically a butterfly valve that is closed at idle. Very little airflow. This is because the mixture must remain within very close tolerances, so small fuel flow = small air flow. The diesel engine “throttle” only controls fuel flow. There is no restriction on the air inlet. A diesel doesn’t need or want a stochiometric mixture. So a 2 litre diesel shifts 2 litres of air every 2 revolutions at idle (roughly). A 2 litre petrol engine at idle shifts a tiny fraction of 2 litres of air every 2 rotations.

It doesn’t seem to get that hot in the under deck (actually very small under boards compartment of the “modern trad” stern. I think because there are large areas of uninsulated steel in contact with the cut, and when the engine sucks air into the inlet, new air comes in from outside. Being a diesel there is a lot of air going through the engine even at low power. So there is a kind of “forced ventilation”. Still, I suspect it gets to 40C in summer.

I think 70% of maximum is a reasonable target. My Iskra is dual fan (one at each end) and I can run it at around 125A/175A ie 71%, that with a limiting cooling air exit temperature of 90C which is fairly conservative bearing in mind the alternator spec says max AMBIENT temperature of 110C He has a "proper" engine room so alternator inlet temperature will be close to ambient with the doors open. Just a moment whilst I convert that to morse code...

This will depend entirely on how hot you want to run the alternator!

I will be really interesting to hear how you get on with the Zeus. It certainly looks a good product, although possibly a somewhat immature one. One of those products that has been taken to the market before all the intended features are actually functional. Seems to be the way of things these days! Definitely an issue with the weight lost, we had to get a couple of 28lb weights to replace the Trojans, even though we went up from 450Ah to 600Ah You could of course just get cells like I did, and add a cots BMS. At least that allows you to locate the cells in the existing space (and in our case, the 600 Ah Li had a smaller footprint than the 450 Ah Trojans, although they were taller). I used https://blslifepo4battery.com/ without any issues, pleased with the product and price although that was a couple of years ago. And has the advantage that you have direct access to the BMS and its configuration.

The "just" also applies to the expensive Wakespeed, Zeus, Alpha Pro etc in that you need to disconnect the existing regulator and access the brushes directly. Much the same amount of work, and although I would agree it is not an entirely trivial thing, someone with reasonable mechanical and electrical competence should be able to do it. I did! The advantage over the above alternator controllers is that it is a massively cheaper solution! You could of course also add a switch into the field circuit to "disable" the alternator gracefully once the Li was charged. I am always happy to share my alternator controller design but one problem is that it was designed to be integral to a system, not stand alone. It currenly looks at individual cell voltages and battery temperature (which it gets from the custom BMS), and battery current and SoC (from BMV712) to control charging, as well as being aware if the BMS is about to emergency disconnect the battery. The software could be re-written to be blind to all that but then there is the risk that a drop-in BMS would decide to disconnect the battery with the alternator controller being unaware until it detects a rapid rise in system voltage.

Yes just add a resistor into the field circuit. Around 4.25A maximum current at 14v seems fairly typical for a boat alternator field current, ie a field resistance of 3.3ohms. To limit the current to 2.5A (alternator limited to 60% of maximum output) you would need to increase the resistance to 5.6 ohm, ie add 2.2 ohm resistor. 2.2 ohms at 2.5A will dissipate ~14 watts. You could use a 25w resistor but I would choose a 50w wirewound chassis mount one, and it would need to be on a heatsink/metal plate because it would get extremely hot even at 14watts if not so mounted. I've been saying this for years, but I don't think anyone has ever tried it.

I suppose I could continue to repeat myself, but what would be the point? Your interpreter algorithm is malfunctioning. Maybe you should stick this into your OS, being binary your parser might understand it more easily: 01000001 01110110 01101111 01101001 01100100 01101001 01101110 01100111 00100000 01100001 01101100 01110100 01100101 01110010 01101110 01100001 01110100 01101001 01110110 01100101 00100000 01100111 01110010 01101111 01110101 01101110 01100100 00100000 01110000 01100001 01110100 01101000 01110011 00100000 01101001 01110011 00100000 01100001 00100000 01100111 01101111 01101111 01100100 00100000 01110100 01101000 01101001 01101110 01100111 00101110

Well at least you will be able to have a nice luncheon. Or several in fact.

Generally the advice from here for disconnecting a battery, is to disconnect the negative first. This is because if you touch the hull with the spanner, it doesn't matter. If you disconnect the positive first and touch the hull, you have a molten spanner. Once the negative is disconnected you can remove the postive connection and if the spanner touches the hull, nothing happens. You really should write a book. After all, you have designed almost everything from the entire Apollo moon landing software (which lets face it was pretty simple!) to the latest AI systems that can recognise a face from orbit. And everything in between. It would be fascinating, although rather long. Although I do curse you when I find that on my iPhone the ringer volume and the alarm volume are inseparable. What were you thinking of when you designed that bit? This is exactly what I said earlier. To paraphrase "just as well I decided not to have any ground connections through any data cables as there would be a risk of harm to them or the PCB tracks (or the BMS accuracy) if I had done." Having a ground connection would be the default, not having a ground connection was the result of considering the alternative current paths issue. This is the incredibly basic point of logic that you don't seem to get. Perhaps you are a badly trained AI after all? By your logic, I avoided a problem with alternative ground paths by not having an alternative ground connection, therefore the potential issue of alternative ground connections doesn't exist. Ever. At all. Does not compute. Back to the training data for you! Oh and forget the ad hominem attack thing, that only applies to humans.