ad: elecraft

DIY 5Ah 64wh LiFePO4 battery tutorial

Discussion in 'Amateur Radio News' started by OH8STN, Nov 20, 2017.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: abrind-2
ad: Radclub22-2
ad: Left-3
ad: Left-2
  1. WU8Y

    WU8Y Ham Member QRZ Page

    But a 25 A supply doesn't actually provide 25 A unless the load demands it. In the case of a battery, the supply will provide whatever current that's required by the battery to reach the 14.4 V output of the supply, following the formula V = V0 + IR, where V0 is the open-circuit voltage of the battery and R the resistance of the system, including the internal resistance of the battery at its present State of Charge (SOC) and temperature. As the battery SOC approaches the charger voltage, current will drop. If that's more than 25 A, it'll provide 25 A and the battery voltage will be less than 14.4 V.

    For a 5 Ah battery, 25 A is 5C-rates, which is a fast charge but not especially challenging. I routinely tested (in my previous job) charges at up to 20C-rates. Those really cut back on cycle life...
     
    KC8VWM likes this.
  2. WU8Y

    WU8Y Ham Member QRZ Page

    A 5s pack, using LiFePO4 cells, would have a 100% SOC voltage at room temperature of 18 V, and a fully discharged voltage (also at room T) of 14 V. So, with a DC/DC buck converter to 13.6 V output, that'd be very good for ham gear.
     
  3. WU8Y

    WU8Y Ham Member QRZ Page

    In really fast charges, the lithium ions don't have time to intercalate, they just kind of glom onto (technical term) the surface of the electrode. Too much of that, and they'll actually plate out, removing that area from the active surface and in the worst case, begin to form dendrites. That'll destroy your capacity.

    100% DOD cycling - full to empty to full again - at low C-rates, below about 3C, isn't really too bad. Above that, though, you get the aforementioned plating and self-heating effects causing electrolyte degradation. Both of those shorten the cycle life of the cell. It's non-linear; as I said, there's not a whole lot of cycle life lost comparing a 3C cycle to a C/3 cycle, but, above that the cycle life vs. C-rate curve rolls off quickly.

    (For those who aren't familiar, C-rate refers to the rated capacity of the cell. So, for example, if you have a 5 Ah cell, then 3C-rate would be 15 A, while C/3-rate would be 1.67 A.)

    Source: I am a battery manufacturer. :)
     
    KC8VWM likes this.
  4. WU8Y

    WU8Y Ham Member QRZ Page

    OK, I shouldn't have written the worst case above. The true worst case is when the dendrites grow through the separator and short-circuit the cell. That is Not Good.

    The main factors affecting cycle life are temperature, rate, and maximum/minimum State Of Charge (SOC). In general, for room-temperature cycling and the same charge/discharge currents, the greater the SOC swing (Depth of Discharge) the shorter the cycle life. For the same DOD, the higher the charge/discharge rates the shorter the cycle life, although there is a noticeable difference between the cycle lives of 100% to 50% SOC swing, 75% to 25% SOC swing, and 50% to 0% SOC swings.

    Don't even get into asymmetric charge/discharge currents (e.g. charge at 3C and discharge at C/3)...that testing and modeling made me lose my hair. The general conclusion is that it gives longer cycle lives to charge slowly and discharge fast, than the reverse but any more specific statement is too situation-dependent to say.
     
  5. AD0GW

    AD0GW Ham Member QRZ Page

    I have been using the MFJ-4416C to overcome this issue. T'aint cheap, though neither are Lithium batteries.

    Jim - AD0GW


     
  6. N6YFM

    N6YFM XML Subscriber QRZ Page

    Hi Julian:

    In the video, I may have missed a few items that seem important to an amateur building their first battery;

    A. Where is a reliable source for buying the 26650 LFP cells with tabs? And how much does each cell typically cost?

    B. Where is a good source for buying a quality Battery Mgmt System (BMS) controller board?

    C. What was the total cost for building the entire battery pack?

    I watched your excellent video, and also read the book "DIY Lithium Batteries" by Micah Toll, but am left with these questions above.
    The book even warned that there is huge industry in counterfit/fake cells on ebay and Alibaba/AliExpress, but then he mentions that
    he buys all his cells from Asia. Also, the book author mentions that many of the names on ebay and amazon (he does not list them)
    buy cells from the major manufacturers that failed testing, and then relabel them and sell them as no-name cells on ebay and Amazon.
    But again, he does not list a few "safe" suppliers for either cells or the BMS. He simply says he buys good ones, which does not help us.
    The "cheap" BMS boards can be built with such cheap parts that they fail and cause battery failure or fire.
    The cheap ones may even exclude cell balance, and only have low voltage cutoff. Or not.

    So it would be a HUGE help to those of us trying to learn to build our own pack, if someone would share where to find
    quality, non-fake, non-reject cells and good quality BMS boards that have balancing circuitry, master LVC (low Voltage Cutoff),
    as well as individual cell LVC protection.

    If anyone can point us in a good direction, please share.

    Cheers,

    Neal
    N6YFM
     
    AK5B likes this.
  7. K6BRN

    K6BRN Premium Subscriber QRZ Page

    I guess what I'd like to know is why MOST applications (car/motorcycle batteries, Tesla, Bioenno, etc.) are using small cylindrical batteries rather than a prismatic approach (built like a lead-acid car battery - much bigger cells packed back to back with much less packaging material used). Current LiFePo cells are mostly packaging by weight and volume, which is a huge inefficiency. Which means they may look pretty good, but are still performing way below their potential.

    A possible driver to this approach (making big batteries out of many discrete, small cells) is some sort of defect density issue where battery cell yield drops exponentially with cell size, active surface area, etc., making prismatic LiFePo cells very expensive.

    Brian - K6BRN
     
  8. VK4HAT

    VK4HAT Ham Member QRZ Page

    Maximum charge current depends on the cells being used. But a typical rule of thumb is 1C which is 1 x the pack capacity in amps is the MAXIMUM charge current you should use with LiFePO4, so 4000mha LiFePO4 pack, 4amps max charge current. Lipo are a different story, 1C is often the minimum charge current for pack, again it depends on the cells.
     
    KC8VWM likes this.
  9. N6YFM

    N6YFM XML Subscriber QRZ Page

    Hi Julian (or others who know):

    I am ready to try building my first pack, but your video seems to have left out one critical detail; Before physically connecting two individual cells in parallel, how close do you charge the voltage to? I am doing a 4S-3P pack with A123 18650 cells.
    These have internal resistance approx 3 to 5 milliOhms. When I do an initial full test charge on 3 new cells, and let them sit
    for an hour, I can meter wide variation for each cell such as 3.3, 3.5, 3.47 volts. Let's do the math; If I happen to pick up a cell
    that is at 3.3 volts and try to connect a parallel cell that is at 3.55 volts, the initial current flow will approximate .25 volts divided by .003 milliOhms, or, 83 Amperes. I tried charging approx 5 of my new cells and metering them, each shows a little different voltage.

    QUESTION: What process do you use to ensure close enough matched voltage, and small enough inrush current, when physically connecting two cells in parallel?

    Thanks,

    Neal N6YFM
     
  10. N0DNE

    N0DNE Ham Member QRZ Page

     
  11. N0DNE

    N0DNE Ham Member QRZ Page

    The debate of how best to power our equipment will continue for many years to come. Each power source has its pros and cons as does output power of the rig, frequency used and antenna vehicle/horizontal propagation. I guess it’s whatever is required to successfully complete the mission. As a ham for thirty years and public safety service operator for more than that, I have spoke on every band from 1.6 to 800 MHz on every type of radio from tube to solid state.

    Reading some of these blogs is like going through a short arm inspection. In all my years of experience in two-way communications I have learned to gain the knowledge and experience that I have yet to receive. My quest is to learn from others so that I may successfully complete. Think about ot
     
  12. AK5B

    AK5B XML Subscriber QRZ Page

    Hoping that Julian will answer these great questions sooner than later; you're not the only one wondering all of the above:)

    73,

    Jeff
     

Share This Page

ad: M2Ant-1