Search results

You've gotten excellent advice so far. A few additional comments/opinions: You can parallel multiple ATs to increase capacity. Generally two Victrons is the most someone needs in USA RV land. You can also buy or commission a custom AT for your desired power. Depending on the topology you...

Well, I certainly appreciate your rigor, in a very selfish way.

Yeah, [happiness] = [reality] - [expectations], doesn't it? Even the high end stuff (I'm looking at you, Victron) suffers from over-zealous marketing making its way into the datasheets... it's rampant.

Bingo. These cells are a bargain. Qualifying and then replacing a few at 96% of specification might feel good, but it is a Pyrrhic victory that places the value of our time at something like pennies per hour, for that little extra. Just assume they’re 90% of rated, or buy 10% more than you...

You can charge, but only at half of the capacity. Some people need that other leg to contribute as well. I imagine they are pretty deterministic, yeah. I know of one rig where it was a big issue, but they had a device introducing a delay on one leg.

Yeah, I think the main issue comes when using a 120V generator. You can't feed a single phase from one of those generators into both inverters if they are in split phase, so people in that scenario only end up able to use "half" of the generating capacity. I went up thread to look at this...

Yeah, definitely a small impact. But well within the realm of possibility if 3% is the difference people are observing in different places.

My (wild) guess is that maybe the electrolyte wicks down out of the jelly roll membrane a little bit, reducing the active surface area. I say that because new cells sometimes make a sloshing sound when you first handle them, but after a while they don't really seem to make as much noise.

There are definitely chemistries that require priming, or warmup cycles, before they deliver their full reversibility. There may be a small factor with LFP as well. It is the kind of detail that might not be interesting enough to make it into the public body of research knowledge. Only...

Well, I think there's a pretty clear indication. Look at the measured capacities at 0, 25, and 40C. Fit a curve to those values (pick your poison). You'll at least have an indication of what the temperature-related effect might be. I really think everyone is splitting hairs here (3% on a...

Temperature definitely affects capacity. Here's one study of small form-factor LFP:

Sounds good -- I'll be intrigued to see if anything changes over time.

How many cycles have you applied?

It's entirely possible that some of these large format prismatics actually undergo a capacity change during their initial few uses (and during their long passive period sitting on the pallets and boat). It stands to reason, if there are mechanical changes (electrolyte expansion, case-widening...

Those sentences are not literally true, but they are substantially accurate. However, they don’t have much to do with the OP’s question. (Why did you write them?)

I would marry them at your balancing "side" of the voltage curve. For example, if you top-balanced both strings, join them up at a high voltage close to the balancing value you started with. You won't get much power inrush if the voltage is pretty close, because discharging LFPs drop in...

Oh, hmm, that's interesting. So the CAN protocol doesn't include a way for the BMS to tell Victron to switch to Float? You just get charge current as the parameter, and so the Quattro holds the higher CV voltage? If so, that doesn't sound ideal. Coupled with the possibility that you might...

It's pretty hard to do dual-channel BMS with an inverter-charger combination unit, since there is only one pair of wires coming out of it. You really have to step up to digital control (CAN or whatever) where the BMS instructs the unit(s) to stop or start charging and/or discharging. In...

I can think of no reason I’d want a BMS to automatically reapply a battery after a disconnect in a single-channel circuit. Even on the bench, I would want a configuration that never does that. If you’re saying you need a safety net until you get your BMS programmed properly... okay, sure. But...

If you just want to buy something, have a look at Crouzet controls, or any other industrial programmable timer. Something like MUR3 will do what you want, for $$, at any VDC you care to provide it. You know, though, that you don’t need automated precharge. You’re over-engineering. :)

Oh wow that is such a good price. I was just going off digikey or something, I think — my notes from years ago say “$300 per pole; physical fit issue”. (I think I had a secondary problem with the size and how to fit big cables onto it or something.) If you don’t get the automatic precharge...

Those are Carlings, something like $300/p IIRC. And you really want to be breaking both wires in a floated scenario if this is a safety OCPD. It’s hard to justify $600 for OCP when there is a $30 solution.

~C/8 here for tests. I definitely had at least a couple out of 66 cells test at 270Ah or possibly even a little lower (didn't push the voltage to the range you have). So I don't know that your result is too big of a surprise.

So is the plan 16s then 2p? Two separate batteries and BMSes?

That sounds like a solid approach. Yeah, I mean, it's the only other thing left to test given the video (I know, you're only showing us a subset of what you've done). If one or more is running out low, and you re-charge and they balance, then for sure you have a capacity issue. Doing it that...

What does "paralleled them at 3.65" mean exactly? You put a dedicated charger set to CV = 3.65V on them and held it for 20 minutes beyond the point when it dropped to 0.5A of charging current? If you can disable balance bleed on the BMS, I'd be inclined to charge the whole string back up...

Hmm. 2.5V is pretty low. I measured >270Ah to 2.8Vpc at 0.05C for 65 of these cells. I didn't actually go any further to see how much they could each really hold. It's very hard to get that much capacity out of them in real life if you're staying conservative for lifecycles, so the fact that...

What makes you suspect them? Are you validating the readings with your meter every time? Are all the connections staying cool during the full-load capacity test?

I'm using MX14 (a little better mounting arrangement for my setup, and a little more current headroom; my system is a bit larger than yours) with great success. Note that the holding power is 2.0W with the 48V coil; also a rounding error. Speaking of watts, what's your plan for dealing with...

We've heard other reports of Chargery being very sensitive to ripple. I guess something like that is possible. Indeed, just a few controlled loads, all colocated... float it is fine. Well, 250 A DC, yeah, probably big and expensive. I'd just use two fuses, but they're also pretty big and...

Those specs are coil inrush for 75ms. Surely the BMS drive line rating is for continuous? I would guess you'd be okay, but if you really want to use Chargery you could ask them to clarify the capability for momentary current.

It sure does. You could have some kind of wiring issue, I guess. I'm not sure why bringing shore power online would cause much current at all. Even if charging is going to occur, it ramps current pretty slowly. I like floating a dedicated 48V house battery, but there are tradeoffs with...

Based on the information you've shared so far, it does sound like one cell is in one of the two states you suggested. I'm not sure how much it really matters. Why not just balance them and then do a capacity test? Put a tag on the sketchy one and see how it behaves over time. If they test...

Dang, that sucks. We had a Dutch Star for a few years, and it was great. I guess there's a lot of randomness in the builds over time. We would consider something the size of a New Aire when it's just the two of us one day, but I'll record your vote in the "no" column now. :D I don't think...

I'm amazed you're going from New Aire to anything (except maybe a SpaceCraft). 2x5kVA Quattro is pretty heavy. What kind of CCC do you have in your new rig? I'm not at all amazed that you're not over the moon with Chargery. It's not the kind of product that exactly exudes quality or inspires...