Charging lithium from the Alternator

[Robmac]

Shhh, it's all bollox, I made it all up! It's amazing what you can get away with if you just act confident and look like you know what you're doing.

Yikes!

 

[Fisherman]

Yikes!

Never trusted him Rob, better get David to check your fire hazard of a van out

 

[merl]

Never trusted him Rob, better get David to check your fire hazard of a van out

Crickey Rob, Wallet alert!! Most stuff'll be the wrong colour for starters.
I was actually quite careful about the fire issues Bill and made sure Rob had a fire extinguisher in the van and to cover myself against the possibility of anyone taking it out of the van at some point in the future I welded it firmly in place to the rear bulkhead so it can't be removed. See... No.mug me mate

 

[Robmac]

Crickey Rob, Wallet alert!! Most stuff'll be the wrong colour for starters.

No, remember the rules Merl - 'if it's Blue I'll have two'

And cables you need a pipe bender to work with them - as you know!

Sizewell 'D' lives on!

 

[Geek]

I guess you've got the two battery types quite close with relatively short interconnects compared a much longer run of common cable supplying them from the B2B?

The b2b is within half a metre of each of the three batteries, and each of them is connected by 25mm cable (it's 50mm cable from the starter battery to the b2b). The AGM batteries are pretty high-spec ones. The impedance of the cables is not specially high. Probably about 0.35mΩ for each, and the cable runs are about the same length for each of them.

Today, on a short run, the Li-ion was showing 13.8v when charging at 33A while the AGMs were at 14.1V when charging at 7A or so. There must be some resistance between them to allow a 300mV difference, but it is difficult to work out where.

I suppose I could investigate with a voltmeter and find out exactly what is happening, but I tend not to worry to much about what they are doing.

They've been connected this way for several years (since before lockdown, I think) and have never given any problems, except on a couple of occasions where battery terminal connectors have corroded a bit (buying cheap quick-release ones was a mistake).

 

[Fisherman]

No, remember the rules Merl - 'if it's Blue I'll have two'

And cables you need a pipe bender to work with them - as you know!

Sizewell 'D' lives on!

Merils favourite quote,

Red to Green, black to Red, orange to yellow,


Blew tae f—k

 

[merl]

Today, on a short run, the Li-ion was showing 13.8v when charging at 33A while the AGMs were at 14.1V when charging at 7A or so. There must be some resistance between them to allow a 300mV difference, but it is difficult to work out where.

Yeh, it's not easy to get your head around exactly what's happening because the resistances governing the charging rates are so small. I've often wondered if the 4 milli ohms figure for lithium is theoretical from.measuring one cell and multiplying by four because if it is then it's inaccurate because it doesn't take into account the 2 short lengths of cable inside the battery or the resistance of the BMS's FETs and circuit board tracks. Not much but significant when you get into single digit milliohms figures.
I built speed controllers for model racing cars back in the 90's/naughties, every month a new FET came out with a lower rds (drain to source resistance) figure as they gradually improved, I've not checked the rds of a modern FET used in BMSs but I bet they are amazingly low by comparison but ultimately they still get warm so there's significant resistance still.

 

[Fisherman]

Yeh, it's not easy to get your head around exactly what's happening because the resistances governing the charging rates are so small. I've often wondered if the 4 milli ohms figure for lithium is theoretical from.measuring one cell and multiplying by four because if it is then it's inaccurate because it doesn't take into account the 2 short lengths of cable inside the battery or the resistance of the BMS's FETs and circuit board tracks. Not much but significant when you get into single digit milliohms figures.
I built speed controllers for model racing cars back in the 90's/naughties, every month a new FET came out with a lower rds (drain to source resistance) figure as they gradually improved, I've not checked the rds of a modern FET used in BMSs but I bet they are amazingly low by comparison but ultimately they still get warm so there's significant resistance still.

Ah now I get it Merl

Ja, es ist nicht einfach, genau zu wissen, was passiert, weil die Widerstände, die die Laderaten bestimmen, so gering sind. Ich habe mich oft gefragt, ob die 4-Milli-Ohm-Zahl für Lithium theoretisch aus dem Messen einer Zelle und der Multiplikation mit vier stammt, denn wenn ja, dann ist sie ungenau, weil sie die 2 kurzen Kabellängen innerhalb der Batterie oder den Widerstand der FETs und Leiterplattenspuren des BMS nicht berücksichtigt. Nicht viel, aber signifikant, wenn man in einstellige Milliohm-Zahlen geht.

Ich habe Geschwindigkeitsregler für Modellrennwagen in den 90er Jahren/Naughties gebaut, jeden Monat kam ein neuer FET mit einer niedrigeren rds-Zähigkeit (Drain-zu-Quelle-Widerstand) heraus, da sie sich allmählich verbesserten, ich habe die rds eines modernen FET, der in BMSs verwendet wird, nicht überprüft, aber ich wette, sie sind im Vergleich dazu erstaunlich niedrig, aber letztendlich werden sie immer noch warm, so dass es immer noch eine erhebliche Widerstandsfähigkeit gibt.

 

[Geek]

I built speed controllers for model racing cars back in the 90's/naughties, every month a new FET came out with a lower rds (drain to source resistance) figure as they gradually improved,

It's also possible that the ratings have changed.
Like appliance energy efficiency, where an A used to be good, but that same appliance is now a F.
Or like microwave ovens, where a 500 watt oven in The Old Days would now be rated as an 850 watt one.

 

[merl]

Ah now I get it Merl

Ja, es ist nicht einfach, genau zu wissen, was passiert, weil die Widerstände, die die Laderaten bestimmen, so gering sind. Ich habe mich oft gefragt, ob die 4-Milli-Ohm-Zahl für Lithium theoretisch aus dem Messen einer Zelle und der Multiplikation mit vier stammt, denn wenn ja, dann ist sie ungenau, weil sie die 2 kurzen Kabellängen innerhalb der Batterie oder den Widerstand der FETs und Leiterplattenspuren des BMS nicht berücksichtigt. Nicht viel, aber signifikant, wenn man in einstellige Milliohm-Zahlen geht.

Ich habe Geschwindigkeitsregler für Modellrennwagen in den 90er Jahren/Naughties gebaut, jeden Monat kam ein neuer FET mit einer niedrigeren rds-Zähigkeit (Drain-zu-Quelle-Widerstand) heraus, da sie sich allmählich verbesserten, ich habe die rds eines modernen FET, der in BMSs verwendet wird, nicht überprüft, aber ich wette, sie sind im Vergleich dazu erstaunlich niedrig, aber letztendlich werden sie immer noch warm, so dass es immer noch eine erhebliche Widerstandsfähigkeit gibt.

I just spent time translating that

 

[marchie]

I just spent time translating that

I must have become bi-lingual; I don't understand the message in both English and German ...

Steve

 

[Pudsey Bear]

Ah now I get it Merl

Ja, es ist nicht einfach, genau zu wissen, was passiert, weil die Widerstände, die die Laderaten bestimmen, so gering sind. Ich habe mich oft gefragt, ob die 4-Milli-Ohm-Zahl für Lithium theoretisch aus dem Messen einer Zelle und der Multiplikation mit vier stammt, denn wenn ja, dann ist sie ungenau, weil sie die 2 kurzen Kabellängen innerhalb der Batterie oder den Widerstand der FETs und Leiterplattenspuren des BMS nicht berücksichtigt. Nicht viel, aber signifikant, wenn man in einstellige Milliohm-Zahlen geht.

Ich habe Geschwindigkeitsregler für Modellrennwagen in den 90er Jahren/Naughties gebaut, jeden Monat kam ein neuer FET mit einer niedrigeren rds-Zähigkeit (Drain-zu-Quelle-Widerstand) heraus, da sie sich allmählich verbesserten, ich habe die rds eines modernen FET, der in BMSs verwendet wird, nicht überprüft, aber ich wette, sie sind im Vergleich dazu erstaunlich niedrig, aber letztendlich werden sie immer noch warm, so dass es immer noch eine erhebliche Widerstandsfähigkeit gibt.

Bovvered

 

[merl]

It's also possible that the ratings have changed.
Like appliance energy efficiency, where an A used to be good, but that same appliance is now a F.
Or like microwave ovens, where a 500 watt oven in The Old Days would now be rated as an 850 watt one.

I asked chatgpt and it appears the Rds (on) resistance of MOSFETs has improved around 40 fold since the mid 90s.

Observed Trend​

• 1990s → 2000s: From ~25 mΩ down to ~7 mΩ as trench cell density improved and die resistance dropped.
• 2000s → 2010s: Further reduction to ~1–2 mΩ as advanced trench silicon met with better package (QFN, copper clip).
• 2020s: Cutting-edge SMD packages in ultra-compact footprints now hit sub-mΩ territory—< 0.65 mΩ—making them far more efficient than older TO-220 planar parts.

Overall, that's about a 40× improvement in absolute per-package RDS(on) from the mid-’90s (~25 mΩ) to today’s best (~0.6 mΩ).

Amazing little devices.

 

[Fisherman]

I just spent time translating that

And I thought you were a techie kind of a guy Merl

 

[Geek]

I asked chatgpt and it appears the Rds (on) resistance of MOSFETs has improved around 40 fold since the mid 90s.

Observed Trend​

• 1990s → 2000s: From ~25 mΩ down to ~7 mΩ as trench cell density improved and die resistance dropped.
• 2000s → 2010s: Further reduction to ~1–2 mΩ as advanced trench silicon met with better package (QFN, copper clip).
• 2020s: Cutting-edge SMD packages in ultra-compact footprints now hit sub-mΩ territory—< 0.65 mΩ—making them far more efficient than older TO-220 planar parts.

Overall, that's about a 40× improvement in absolute per-package RDS(on) from the mid-’90s (~25 mΩ) to today’s best (~0.6 mΩ).

Amazing little devices.

Sorry, I consider chatgpt to be even less dependable a source than the Daily Hate Mail. AI merges opinions without evaluating them.

 

[Pudsey Bear]

Sorry, I consider chatgpt to be even less dependable a source than the Daily Hate Mail. AI merges opinions without evaluating them.

If you ask its opinion, I sort of agree; however, the results I've been getting, I check, and it's been infallible so far.

I have been tasking it with some difficult (for me) jobs, and although it has taken a while, with much toing and froing, to fine-tune things, we've got the results I was looking for.

Agree about the Mail, though, and the Express.

 

[Technispark]

That’s why as highlighted in the video you need a B2B which controls the current draw whilst maximising charging from a smart alternator. This video from five years ago shows the old 30a victron b2b which was far less efficient, larger and required a heat sinc. The new 50a B2B is a game changer not only producing more power but doing so so efficiently that no heat sinc is required due to 98% efficiency, far more efficient than the old 30a B2B which I think ran at around 70% efficiency. (The less efficient electrical equipment is, the more heat it creates necessitating a fan or a heat sinc)
As for drop in lithium that depends on the age of the vehicle and the charging equipment supplied with it. Any van produced in the past 4-5 years will most likely come with all of the ancillary equipment required, and drop in lithium is a realistic expectation. My van came with lithium compatible charging equipment, inc a Nordelectronica 40A B2B which as I said I recently replaced with a 50a victron B2B. But older vans will require lithium charging equipment. But the benefits of lithium far outweigh any disadvantages for those who require more power, or who wish to go off grid in winter.

Lead batteries are relics from the past, heavy, low usable amperage, short life spans, varying voltage, greater internal resistance making them more difficult to charge, and in the medium term far more expensive. My old Varta lead batteries died after two years, lithium batteries can be charged thousands of times before reaching 80% capacity. Even then they are still better than lead batteries still in use today.

If you have the right lithium battery it is drop on compatible. I have an Ecobat 100ah and it dropped in to my 18 year old Bilbo with Sergeant charger and works fine . I ran this through with Ecobat technical before purchasing and was told it would be fine and it is . Always a good idea to talk to the battery manufacturers before buying and avoid unnecessary additional electronics

 

[Technispark]

Cost is a big issue Barry, and I don't use much when not driving, so 100ah is more than enough I think.

Go for an Ecobat 100 from go outdoors for £300 great value and a direct drop in replacement

 

[Pudsey Bear]

Go for an Ecobat 100 from go outdoors for £300 great value and a direct drop in replacement

Sorry, but why would I pay 4 times the price of the 110AH LA I bought, when I clearly didn't and don't need it? I did very nearly get sucked into Lithium, and at much less than the £300 you quote.

We are100% wild campers (as far as that goes in van). This not only covers where we stay, but also how we stay; we're very minimalist within the constraints of a motorhome.

 

[Geek]

Today, on a short run, the Li-ion was showing 13.8v when charging at 33A while the AGMs were at 14.1V when charging at 7A or so. There must be some resistance between them to allow a 300mV difference, but it is difficult to work out where.

I'm guessing there must be a 9mR resistance somewhere in that circuit. Won't be easy to find!