[aquabiologist]

Hi

Looking for my 144 volt 28.8 kw battery pack I stumbled upon this 4.4 ah 36 volt lg mf1 pack
http://www.ebay.com/itm/201916138726


Here's a spec sheet on the cells:

http://www.houseofbatteries.com/documents/5216-BatteryLines/LGLithiumIon18650MF1Rechargeable.pdf

This guy tested the pack and the bms that comes with it
https://youtu.be/Q2w5Ka2oNv8


So i figure a 4s45p would give a nice 200 A 144 volt pack @ approx 180 kg.
Much lighter way cheaper than lifepo4 especially as a bms is included.

Also it seems the pack can be hooked to a charger, while discharging
https://youtu.be/Ua6PpLWqEgc

Series hybrid anyone?

I wonder, since lithium batteries do not balance in series but will in parallel (or am i wrong), this should perform better than a lifepo4 pack, which could be 45s2p with 100 ah battery. With the lifepo4 pack one dead cell drops out 50 % of the pack. Where in the other case just a bit over 2% would drop out. So still alot of range left in that case.

But at the price of less cycles (500) and a less safe chemistry?

What do you think. Anyone using these?

Gesendet von meinem E5823 mit Tapatalk

 

[aquabiologist]

Found the answer myself

After some more digging found a list of lithium chemistry and battery codes here:
https://batterybro.com/blogs/18650-.../18880255-battery-chemistry-finally-explained

The lg 4.4 packs sold on ebay are lico chemistry. So no. Not safe for ev purpose.

Gesendet von meinem E5823 mit Tapatalk

 

[Solarsail]

Series hybrid is the way to go, unless you have 200 kWh of EV storage. You have a very nice architecture: 4s45p10s2p

And the 2.2 Ah cells are cheap, about $1.30/cell as opposed to $3.25 for a Panasonic 3.4 Ah. At $164 /kWh they are the cheapest on the market. I have decided to go with the Panasonic at $266 /kWh for my 12 kWh pack, as it will be 35% smaller and lighter.

May I ask what motor and controller are you using?

 

[Solarsail]

The problem with parallel cells I understand is that if one cell in a group or bank that is in parallel is underpeforming, it will be hard to find out which cell that is. And one bad cell can make the whole bank and pack useless by triggering underdischarge voltage protection.

I am surprised that you say the LCO chemistry is not safe for EVs. Isn't that exactly what Tesla has been using? Tesla is not using LMN. So why do you need to be safer than a Tesla?

 

[Duncan]

$266 dollars a kWh??

I paid $1800 for a 16 kWh Volt pack - thats only $112 a kWh

And I got all of the fixing hardware, five contactors, the BMS modules, the main fuse and a current sensor all part of the package

The Volt modules are water cooled/heated and are capable of delivering oodles of power - I'm drawing 400Kw!! - for a couple of seconds

 

[Solarsail]

That is a great deal Duncan. The price of the Panasonic drops with volume. The price I quoted was for a 1.3 kWh pack. For a 12 kWh pack, the consumer price will drop to about $230 /kWh. These are new cells under warranty. Also, it gives you a lot of flexibility on how to configure the pack, as I am building from scratch. And if I were to expand the pack, I can do that in increments, and need not find used auto packs. Availability is not an issue.

In my application, I don't need to draw a lot of power. 25 kW would be plenty.

Five contactors? Where these all in the power pack on the power main? Why would there be five contactors in a Volt power pack?

 

[Duncan]

five contactors
two big ones were for the main HV lines

Three smaller ones
One was for the pre-charge
One was for the heater for the water circuit (did I mention the heater - it was in the pack as well)
I can't remember what the third one did

If you are buying to make something and sell it - then you really have to buy new

If it's for your own project then buying something like a Volt pack makes a lot of sense - everything is made to automotive quality levels - a LOT higher than normal consumer levels

 

[Solarsail]

I am just building for my own use.

Two heavy contactors for the HV line. Is that because the negative also has a contactor or is it that there are two modules and each has its own contactor? Why would the HV need two instead of one?

So the battery heater is HV. Yes, used EV batteries are a very good deal. I would love to pick up a used T3 pack with the 2170s.

 

[Duncan]

Hi Solarsail

I think all electric cars break both the positive and the negative HV lines - for safety
The Volt also has a fuse/service plug that breaks the pack into two

I know that is a legal requirement for Home builts in NZ to have a contactor in both lines

I was able to use one of the large contactors for my pre-charge but I draw too much current to use them as my main contactors and I had to pay money to get a pair of larger ones for my main contactors

 

[john61ct]

How do the 3.2V prismatic cells compare pricewise? Just a bit fuzzy on the units conversion. . .

Say I can get them at $130 per 100AH (C/20 rating) delivered, how many of those would it take to get to 12 kWh?

 

[Solarsail]

Hi Duncan - thanks for the info, it is very helpful.

May I ask what is your 400 kW application? Racing?

Can I ask why do you have to pre-charge the inverter controller? Does it have a huge capacitor? Are you using a microcontroller to manage the battery?

As you are in Kiwi, can I ask you if you have experience with e-drive for cats? That is the purpose for my 100 kWh project - to put it on a cat with solar panels (100m2). Do you know of people doing this? If I do this, it will most likely be based in Brisbane (only 13 hours from Vancouver).

 

[Solarsail]

How do the 3.2V prismatic cells compare pricewise? Just a bit fuzzy on the units conversion. . .

Say I can get them at $130 per 100AH (C/20 rating) delivered, how many of those would it take to get to 12 kWh?

I believe you mean LiFePO4 prismatics. My understanding is that they are more expensive, because their unit capacity is less. Assuming the prismatic is 1S, then energy is 320 Wh/cell. You will probably need at least C/4, so the capacity probably drops to 280 Wh/cell. That is 3.5 cells/kWh or 455 $/kWh. This is almost twice the price of 18650 LCO.

You would need 12K/280 = 42 cells or $5,500 per 12K pack. Problem is that these cells are such huge capacities that it will limit you to the voltage that you would get. 42S1P will give you 134V. 21S2P gives you 67V. So you can only get 33V, 45V, 67V, and 134V. If you want to get 96V or 192V, you have to greatly increase the capacity to 17 kWh, or drop it to 8.5 kWh for 96V.

My argument is out of billions of 18650s in laptops, how many have recently caught fire? And note that laptops don't have microfuses. Do you hear of laptops catching fire anymore? And when they did, it were just a few, and probably related to the Sony problem. All these millions of smartphone portable chargers, 1 cell and 4 cells, etc. are they catching fire? I have done nail tests on loaded 18650s, and I don't get a fire. When my batch of NCR18650Bs arrive I will do some more tests.

Can you please give me the size and weight for this prismatic?

I think the 18650s mentioned by Aquabiologist in this tread is a much better deal than the prismatic. Or as Duncan has said, a used EV module.

 

[Solarsail]

I am not sure if the test attached (Panasonic 3.4Ah 18650) indicates there is a puncture of the shell or if the pressure release valve has been activated. The 18650 survived an impact. Temperature rose to 170C.

 

[Duncan]

Hi SolarSail

My car has a Paul & Sabrina prototype 400 v - 1400 Amp controller with great big disc shaped capacitor
I'm using 340 v and 1200 amps in - Duncan's Dubious Device
http://www.diyelectriccar.com/forum...dubious-device-44370p15.html?highlight=duncan

You will find that all controllers have and chargers have decent sized capacitors so there is always some form of pre-charge system to avoid a huge inrush current when you switch them on

Mine is purely manual - I switch the 12 v power on which switches one of the main contactors on - then switch the pre-charge (I'm using a kettle element as a pre-charge resister) then when I reach about the right voltage I switch the second contactor on

You get the same with the charger - you need to couple it through a resister until the voltages get close or you get huge surge currents and your contactors welding together

 

[Solarsail]

Thanks for the link. That is one cute hot rod you got there. At 400 kW, that would be equivalent to (400/.74)*1.5 = 800 HPe? I am told ICE HP is 50% exaggerated when compared to electric, due to torque curves and very tight optimal operational window for an ICE.

Have you thought of putting in a 2 speed gearbox, to get better starting acceleration, and lower end run RPM? How about a planetary gear set, a lock, and electronic synchronization? Let's say the planetary gearset is 2:1 ratio. The universal gear is locked to ground (stationary) by a stationary solenoid (#1). This will be low gear. To change gears, the solenoid #1 retracts and releases the universal. Motor RPM goes to half which will synchronize the universal with the planets. Another solenoid #2 that is bolted to the universal locks the two, so now you have high gear. To go back to low, lock 2 opens, motor doubles its RPM (in relation to the planets) so the universal becomes stationary and lock 1 closes. No clutches needed, no synchromesh needed, and you get the benefits of a planetary gearset. Is this easy to build?

 

[brian_]

How about a planetary gear set, a lock, and electronic synchronization? Let's say the planetary gearset is 2:1 ratio. The universal gear is locked to ground (stationary) by a stationary solenoid (#1). This will be low gear. To change gears, the solenoid #1 retracts and releases the universal. Motor RPM goes to half which will synchronize the universal with the planets. Another solenoid #2 that is bolted to the universal locks the two, so now you have high gear. To go back to low, lock 2 opens, motor doubles its RPM (in relation to the planets) so the universal becomes stationary and lock 1 closes. No clutches needed, no synchromesh needed, and you get the benefits of a planetary gearset. Is this easy to build?

No, I don't think it is easy to build, to work reliably and handle significant power. As far as I can find so far, there are no solenoid-shifted two-speed transmissions (planetary or otherwise) currently available... although there are many transmission companies which are quite capable of designing the product and manufacturing the components.

 

[Solarsail]

But you see, there is no load on the solenoid. It is not moving gears or shafts or rings or clutches. All it does is to lock the universe gear to either ground or to the planets. Even a 12V 1A solenoid could do that. And multiple solenoids can be used for each lock. With a strong traveler, the lock can handle lots of torque, I think, as long as it can overcome the centrifugal force in high gear. A brush would be needed to deliver power to lock #2. The #1 lock is stationary so no issues there.

It would be best to put the lock#2 on the planetary set. Less centrifugal force and also can use the same ribs on the universal gear as lock#1 uses for latching.

I think the Volt already uses electrical locks (and a clutch) in its very complex transmission, which I understand is based on one or two planetary gearsets.

 

[Duncan]

Hi Solarsail

I can spin my tires from stationary - and I just don't have the space for any gearbox at all

As far as top speed is concerned - for the 1/4 mile it would be handy but for the road the overall speed limit is 100 kph - and I can do over 140 kph

For the motorsport they are quite sensible here - the faster the sport the more safety is needed
For the basic stuff I do they make sure we stay below about 100 kph

To go up to the next level I need a license and a lot of other stuff

 

[brian_]

But you see, there is no load on the solenoid. It is not moving gears or shafts or rings or clutches. All it does is to lock the universe gear to either ground or to the planets. Even a 12V 1A solenoid could do that. And multiple solenoids can be used for each lock. With a strong traveler, the lock can handle lots of torque, I think, as long as it can overcome the centrifugal force in high gear. A brush would be needed to deliver power to lock #2. The #1 lock is stationary so no issues there.

It would be best to put the lock#2 on the planetary set. Less centrifugal force and also can use the same ribs on the universal gear as lock#1 uses for latching.

I'm not sure quite what mechanism are meant by "locks", but in practice they would be dog clutches if they are going to handle any significant torque. There are bicycle transmissions which use pawls (such as the Pinion series, and similar to what is normally used as a park mechanism in an automatic or an EV gearbox), but in the bikes that's a one-way drive (no regenerative braking) and these transmissions only handle a few hundred watts of input power (although at significant torque).

There are transmissions intended for electric vehicles which shift gears, such as the GKN two-speed eAxle, used for the front axle of the BMW i8. GKN runs synchronizers with a motor which slides a conventional shift fork (like a traditional manual transmission) to slide the moving part along the input shaft... and that's in an integrated drive system, developed in close cooperation with major auto manufacturer for their most technically advanced model, boasting "a complex GKN shift control software which combines eMotor control with a precise synchroshift control using advanced sensor and actuator technology."

I think the Volt already uses electrical locks (and a clutch) in its very complex transmission, which I understand is based on one or two planetary gearsets.

The Volt transaxle is mechanically pretty simple (compared to a modern automatic), like other similar hybrid transaxles (such as those in Toyotas). It does use a planetary gear set, although as a power splitter rather than as a shifted reducer. It does have shifted elements to change modes; they are shifted by engaging and disengaging multi-plate hydraulically actuated clutches, just like conventional automatic transmissions.

Any conventional automatic transmission controls planetary gear sets as you are describing, but they use multi-plate hydraulic clutches (although long ago they used drums and brand brakes). There are a very few dog clutches used, although ZF has resorted to them in the HP9 transverse transaxle to save space... and they actuated hydraulically, just like the multi-plate clutches.

 

[Solarsail]

Thanks a lot. I'll study this and get back ...