Saturday, May 19, 2012

More Charging Experiments

Not much to report for a few weeks... Thankfully.

The Winston battery pack continues to do nothing other than sit there and do it's job without fuss, chemical smells, or sudden death.

April's final generation figures were interesting in that they were the same as March's figures.  That's interesting because March was one of the sunniest on record and April was one of the rainiest on record.

May is turning out to be a mixed bag and a bit disappointing.  We had a long run of gloomy days and the main lithium bank hit the bottom a few times.  On one run of bad weather I also depleted both of the backup Ritar lead acid batteries as well, using an old 1kW inverter to run a pair of the lab power supplies to charge the lithium bank during the night.

The lithium battery protection works well, sounding the police car siren at low battery and then shutting down the inverter when the weakest cell gets to 2.999V.

Then during the day after using the Ritar battery, I turn the lab power supplies the other way round and use the solar to recharge the fragile lead acid batteries first.  It doesn't matter if the lithium bank sits at the bottom for a few days without being charged much but the lead acid ones need charging as soon as possible.

I might set up a more permanent hybrid battery system to make use of the lead acid batteries as a "battery of last resort".  Used that way they might last for many years, if only discharged once in a few weeks, rather than every day.

I've also been playing with the lithium bank charge settings.  Cell no. 8 has taken to reaching full charge a bit before the others.  This is a feature of the bank being bottom balanced and the longer absorption charge times I've been playing with (now up to 50 minutes).  At 28.0V absorption level, the cell was getting up to 3.58V before charge end, while the others were getting up to 3.51V.  But cell no.8 rockets up to this high in a few minutes at the end of the 50 minutes, causing the other cells to actually fall a bit.  So I've been winding the absorption Voltage down, a bit at a time, to see at what point the cells will remain close together.

At 27.7V, or 3.462V per cell, this seems to be the case.  Today the pack sat at absorption for the full 50 minutes and all the cells remained fairly close together in Voltage.

6 comments:

  1. Hello,
    nice to see, your system works well.

    I think that your cells are drifting without balancer at some point. At least that's the experience in our German PV forum.

    A good possibility is the new balancer "LiPro1-1" by www.ecs-online.org. They are cheap - use little energy into heat - can be directly controlled solid-state relays. I will use it.

    Yes, I´ve now completed my planning and I want 4 each of LyFeYPOs use in our motor home. Unfortunately, nothing on stock - I will see.

    Best Regards Gerd

    ReplyDelete
    Replies
    1. Hi Gerd,

      The pack isn't drifting as such. Just pushing too close to the top. Being bottom balanced, it will go a bit wonky if you try to charge it to a high level. 3.50V is right on the knee in the charge curve where it all starts to get non-linear.

      When sitting at normal charge levels (at or below 3.375V per cell) the whole bank sits within 7mV of balanced when delivering the house base load of about 250-300W.

      Even when scraping the barrel in bad weather we had in May, the bank stays within about 20mV total delta between the best and worst cell right before the protection shuts down the inverter. Using a top balancer would only make the pack perform worse at the bottom (when you need it to be balanced so that all the cells get to be empty at roughly the same time).

      I'd rather lose some top end capacity from "castle wall" balancing (uneven at the top), to have more usable capacity at the bottom before the protection turns the lights out in the house.

      Delete
    2. Agreed. I think Jack in his multitude of experiments and cell losses can show the folly in bothering with them. His show in which he uses the different length sticks for illustration purposes drove it home for me.

      Delete
    3. The 3.5V Knee.... I agree. I've noticed it when using 56V (16 cells) as my charging parameters in the Midnite Charge Controller. It causes two cells to shoot up further than the others. Instead, I have backed it off to 55V (3.44) and it works happily there. 4 months later. 117 Days /235Kwh Later.

      Delete
    4. Hi would it be a feasible to run two winstons - the 12v 90ah solely using a good solar charge controller (Midnite Solar Kid) - No BMS?

      Delete
    5. If you mean those 12V monobloc packs then, yes. I don't think you have access to the individual cells anyway so you've no choice but to charge them with no BMS.
      If you're using two packs, I'd use them in series, not parallel. You'd be bound to get into balancing trouble if you tried to use them in parallel.

      Delete