[dimitri]

Originally Posted by tomofreno

I'll ask the obvious just to be sure, you plan some way to warn a fuse has been blown correct?

My idea of a single wire Normally Closed signaling loop is such that ANY failure ( blown fuse, loose wire, dead cell, LVC or HVC on any cell , blown component ) would result in open circuit and produce the alarm event. Head end control board will decide what to do with this alarm event, sound a siren, drop a charger, drop or reduce throttle, eject the driver from the seat, up to EV designer's imagination

[tomofreno]

My idea of a single wire Normally Closed signaling loop is such that ANY failure ( blown fuse, loose wire, dead cell, LVC or HVC on any cell , blown component ) would result in open circuit and produce the alarm event.

Would be great!

[tomofreno]

One idea could be just to add 1 individual cell SoC meter on the weakest battery.

I think I already have this capability. The charger supplies current through the shunt so I can read DC current into the cells during charging on my TBS gauge. Then I can record time and know how much Ah charge I put into the cells. After I have gone through a couple charge/discharge cycles monitoring cell voltages, I should be able to identify the lowest capacity cell as the one with the largest voltage change during charging, and know it's capacity from the Ah data. Then I just enter this Ah capacity on the TBS gauge for the pack capacity, so it reads soc based on the capacity of this cell (the TBS monitors discharge current through the shunt and integrates over time to track Ah of charge discharged). I would then not be in any danger of over-discharging this cell, nor any others if the cells remain fairly well balanced, if I stop driving by 20% soc. This should be quite safe if I monitor cell voltages at regular periods to see if capacities/and or balance have changed. If another cell should become the lowest capacity, I just enter its capacity in the TBS. If they become too imbalanced, diminishing pack capacity, I re-balance them by charging lower voltage cells with a ps. I install Dimitri's HLVC system as a safety backup to ensure I don't over charge/discharge if balancing changes more than I expected between my monitoring periods.

Tom

[etischer]

Originally Posted by dimitri

Tom,

the problem with lower shunting level is that chargers can't lower their CV current enough to be less than shunting capacity, so you end up overrunning shunts and create more problems that you solve.

High Voltage Cutoff is what protects the cell.

Whether you shunt balance or not, the HVC protection is there to shut off the charger.

Whether your shunts are over run by charge current or not, the HVC is still protecting each cell.

[dimitri]

Originally Posted by etischer

High Voltage Cutoff is what protects the cell.

Whether you shunt balance or not, the HVC protection is there to shut off the charger.

Whether your shunts are over run by charge current or not, the HVC is still protecting each cell.

HVC is just a signal, it doesn't help if your charger is dumb enough not to be able to react on it.

HVC from first cell can be interpreted in 2 ways, either throttle the current to less than shunting and let the rest of the pack safely catch up, or shut down the charger and forget the balancing. I don't see how you can have both with a dumb charger.

How many commercial chargers do you know which are capable of throttle down on HVC signal? How much do those cost?

[dimitri]

Someone mentioned BMS from EV Power in Australia. After reading their Web site I have to admit that their product is almost exactly what I had in mind, although its more expensive and some data posted on their pages seems to be contradicting, but perhaps its just due to various versions and ongoing development. It would be nice if someone with real experience with these modules came in here and gave us an independent review. Anyone?

They state support for SE cells, but operating range of 2.5V - 4.1V does not match SE range, which would cause problems. This means separate product selection for SE should be offered, but its nowhere to be found.

I like physical design of the unit, but it appears to be a single PCB and I am not sure that pressure of fully torqued terminal bolt will not crack and destroy fiberglass layers of PCB. Its possible they use special PCB designed to withstand pressure, don't know.

They claim simple NC signaling loop, and in another page they refer to some proprietary heartbeat protocol, so which is is?

They have some shunting ability to slow down first cells to reach 3.65V and let others catch up, although with 0.6 Amp shunting it will only take a minute for first cell to signal HVC and shut off the charger, so I am not sure how useful their shunting is, but it seems to be designed with safety in mind even if its not so useful. This could work well if your charger only pushes 1 Amp in CV mode, but I am not aware of such chargers for large cells.

Again, on paper and pics looks like real nice product although not so cheap. They are looking for US and European distributors, so I wonder why none of many US EV shops jumped on this yet? Maybe someone did? Anyone?

Even though I have plenty of design ideas, I doubt I can compete with this nice product without cutting corners, so if someone is waiting for me to start selling BMS, I have not decided if I want to step into this market.

BTW, here are useful links, enjoy and please come back with comments.

http://www.ev-power.com.au/-Thunders...g-System-.html

http://www.evworks.com.au/index.php?...t=BMS-CM160-V6

http://www.evworks.com.au/store/data..._datasheet.pdf

[dimitri]

Originally Posted by etischer

High Voltage Cutoff is what protects the cell.

Whether you shunt balance or not, the HVC protection is there to shut off the charger.

Whether your shunts are over run by charge current or not, the HVC is still protecting each cell.

Eric, I think I misunderstood your statement at first, sorry about that. You consider shunting point and HVC as 2 different voltages, just like abovementioned EV Power system, right? So, shunting starts first, then after some time ( depending on how much current is pushed by the charger and how much this current is over shunting ability of the modules ) HVC hits and charger is forced to turn off. In my mind I always had shunting and HVC to be the same voltage point, which relies on charger to shut off on its own when max pack voltage is reached while modules must shunt ALL current to prevent overcharge, this is how VB was designed. Just 2 different ways to do it. One is designed to make sure most cells have a chance to reach HVC, but requires proper match with the charger ( this is what my EV does, BTW , and it seems to work well for me ), the other is focused on safety shutoff, even if that means some cells never get to catch up all the way. After all recent discussions I definitely see merits to this approach, although it would require additional voltage sensing components to distinguish shunting voltage from HVC voltage, which adds cost.

[Dolphyn]

The NC signalling loop idea has got me thinking, so I'll toss out one more idea for battery management and monitoring.

It seems vitally important to monitor the "low-cell voltage". That is, I always want to know the condition of the lowest cell. Instead of (or in addition to) a "pack state-of-charge", I want an indication of the weakest cell's state of charge. During actual driving, that could be more important than knowing which cell is weak!

One approach would be to have multiple NC signalling loops, say, one each for 2.5 volts, 2.7 volts, 2.9 volts, 3.1 volts, and 3.3 volts. Each could be wired to an LED on the dashboard of the vehicle, so that in this example "three LED's on" means that every cell is at least 2.9 volts. This would provide instant visual feedback to the driver if any cell (or the whole pack) is getting weak.

Another approach would be to have a BMS output that simply matches the lowest-cell voltage, and wire that output to a voltmeter on the dashboard.

Alternatively, one could do something like the "low cell voltage output" described here:

Originally Posted by liionbms.com

Low Cell Voltage Output: Analog voltage that remains at 5 V as long as the battery is able to provide current, and decreases down to 0 V as the lowest charged cell's voltage is reduced from 2.9 V down to 2.1 V.

I think their idea is that it would be wired to the throttle somehow, to reduce the power available to the driver, but it would be equally useful to provide a visual queue. (Note, I've asked evcomponents.com and eeveemotors.com about that BMS, and it doesn't actually seem to be available.)

Anyway, regarding EV Power ...

Originally Posted by dimitri

They claim simple NC signaling loop, and in another page they refer to some proprietary heartbeat protocol, so which is is?

I think both are correct! If you buy the Master Unit, the heartbeat protocol is used, but otherwise the loop is used as a simple NC signalling loop.

[dimitri]

Originally Posted by Dolphyn

One approach would be to have multiple NC signalling loops, say, one each for 2.5 volts, 2.7 volts, 2.9 volts, 3.1 volts, and 3.3 volts. Each could be wired to an LED on the dashboard of the vehicle, so that in this example "three LED's on" means that every cell is at least 2.9 volts. This would provide instant visual feedback to the driver if any cell (or the whole pack) is getting weak.

This will require additional wire for each signal, which creates spaghetty wiring mess, conflicts with requirement of easy install. It also adds voltage sensors for each level, which adds cost.

Anyway, regarding EV Power ...

I think both are correct! If you buy the Master Unit, the heartbeat protocol is used, but otherwise the loop is used as a simple NC signalling loop.

You can't use one wire for both without major complexity of design and I believe their modules are just as simple as what I listed in my original list. One could say that I came up with my list after reading EV Power specs , but its not true, I came up with the list by listening to feedback from many forum members here and my own LFP experiences.

[Dolphyn]

I'll admit after reading the site again, I'm not entirely sure what purpose the heartbeat serves. Maybe I'm missing something, but it doesn't seem like the control unit does anything that would require any sophisticated feedback from the modules.