Advice wanted on lithium battery choices

[Peter Thompson]

Hi Folks,

I've figured out several different configurations of lithium batteries
for the EA 914AC kit racks, and have come up with some interesting
possibilities.

Individual Battery Battery Pack
Battery V Ah Weight Cost # of Batt Pack V Avail Kw
Pack Lb Cost
SE120 min 3.2 120 9.9 $132.00 51 163 13,709
505 $6,732.00
SE100 min 3.2 100 7.04 $110.00 52 166 11,648
366 $5,720.00
TS90 min 3.2 90 7.04 $99.00 60 192
12,096 422 $5,940.00
SE60 min 3.2 60 5.28 $66.00 72 230
9,677 380 $4,752.00

SE120 max 3.2 120 9.9 $132.00 76 243 20,429
752 $10,032.00
SE100 max 3.2 100 7.04 $110.00 63 202 14,112
444 $6,930.00
TS90 max 3.2 90 7.04 $99.00 89 285 17,942
627 $8,811.00
SE60 max 3.2 60 5.28 $66.00 109 349 14,650
576 $7,194.00

All of the costs are based on the posted values on the
evcomponents.com site.

Assumptions: cost is not a huge factor in this assessment. Space has
been predetermined by the battery racks from the kit. This is powering
the AC24 / DMOC445 motor and controller.

Questions:
1) should I go for max voltage?
2) or should I go for max
3) or should I try for a compromise.

If there is a desire, I can post the spreadsheet used to calculate
these values on the group list.

Thanks in advance for any advice/suggestions.
Peter

[David Hale]

Peter,

I know you have been considering this and my recent woes posted both here and EVDL have swayed you into lithium. Li is going to give you less mass and more voltage in the same space, so it's a great thing. And numbers I see about FLA & LFP lifetimes suggest that LFPs may be cheaper to own over the long run; of course, you have to pay substantially more for this up front. If that is not a concern then that's great. But in light of the recent responses I've gotten from EVDL and especially with Andy's car coming on line and getting 0-60 in 20s, it is clear that my car has a unique problem. So you shouldn't let my case be the sole decider for you.

Now, if you are sure about LFPs then you should talk to Electro Automotive about it. Bill mentioned once that you should not use less than 160Ah cells, although JungleMotors is using 90Ah units. I e-mailed Jungle about this and they didn't respond. Dave @ evcomponents said he'd make some calculations. 

Power is linear with voltage so in theory you want as high a voltage pack as you can fit, but I've heard (or read) somewhere something to the effect of the Azure motors having a certain range where voltage increases don't help until you cross certain thresholds. You might want to try contacting them about that.

Also, my figures are slightly different from yours. I determined the following for SE cells in the ElectroAuto battery boxes:

180Ah: 59 batteries for 188V pack and 728 lb

120Ah: 76 batteries for 243V pack and 754 lb

100Ah: 86 batteries for 275V pack and 588 lb.

Anyway, now that it seems clear that my car has a unique problem I am no longer pursuing the idea of LFPs -- at least not until the FLAs need to be replaced (and assuming we've fixed the current problem).

--Dave

[Evelyn]

All,
My 0-60 is more like 45 secs with a power setting of 30000, not
in the 20's as I stated before.,
(0-40MPH around 16 secs with a power setting of 30000).
I guess you shouldn't believe the readings you get from an iphone with
a dyno app.
What I notice is that the ISR2RealTorque drops off after ca 5 secs,
(the ISR2TorqueDesired) is still high.

Andy

On Jun 13, 4:00 pm, David Hale <astronomerd...@gmail.com> wrote:
> Peter,

[Peter Thompson]

I've pretty much decided to forget about FLAs - the conversion was
intended to be a learning experience, and totally has been one. :)

I'm definitely leaning towards the SE60AHA set (349 volts, 14.65kw
available, 576 pounds) for many reasons - really high voltage, less
weight and cost.
I do suspect that some of the cost savings will be eaten up with BMS,
due to the large number of batteries.

A close second is the TS90 set (285 volts, 17.94kw availalbe, 627) for
the extra power in the pack.

I've sent email to Azure Dynamics asking about the upper limit on the
voltage - I suspect that will really settle the issue. :)

I'll definitely keep the group posted.

Cheers,
Peter

[joe.obrien]

You might also want to contact Doug over at Blade Electric Vehicles.
"doug.f...@bev.com.au"

They use Azure controllers with L-ion packs. So they have a lot of
experience that directly relates to what you're thinking about.
Joe

[John RA Benson]

Doesn't range matter? the downside of the higher voltage and smaller
pack is of course less range. with my 100ah lead pack at 192v, I've
got great performance (even with that 1300 lb pack!), but going more
than 40 miles is something I don't want to do anymore. half that
weight and 1.8x capacity would be great - IMHO anyway.

a reason you might want a larger ah rating could do with the battery's
C rating - I don't know lithium, but if a 90 ah battery can't deliver
at least 300 amps for more than a minute, you'll probably be on your
way to premature failure. higher ah batteries like the 180 are more
likely to be capable of delivering at least 2C continuously. My 100ah
nicads could only do 2C and 500 peak for 10 seconds - it wasn't enough.

just food for thought ;)
Good luck!

jrab






On Jun 14, 2009, at 18:13, Peter Thompson <PThomp...@gmail.com>
wrote:

[Peter Thompson]

Range is indeed somewhat of a concern, with a range of 68 to 48
miles. Naturally, I would like to get the 150 miles advertised, but
that's not gonna happen without a Mr. Fusion.

As for battery ratings, I've checked them all out. The worst case I
saw was with the SE120, where it was demanding 2A per battery. Best
case was with the SE60, with 1.7A per battery. Of course, this was
using the 150A max for the Wye configuration. The 240A for the Delta
configuration is another matter, but not too far off - the SE120 uses
3.15A per battery, and the SE60 uses 2.2A per battery.

It is odd, as I've heard the term "soft battery" when talking about
the SE60. Yes, it has less capacity, but ganging up more of them
boosts the pack voltage, even though you get less range. *shrug*

Investigation continues. :)

Cheers,
Peter

On Jun 15, 3:10 am, John RA Benson <j...@e-m-w.com> wrote:
> Doesn't range matter? the downside of the higher voltage and smaller  
> pack is of course less range. with my 100ah lead pack at 192v, I've  
> got great performance (even with that 1300 lb pack!), but going more  
> than 40 miles is something I don't want to do anymore. half that  
> weight and 1.8x capacity would be great - IMHO anyway.
>
> a reason you might want a larger ah rating could do with the battery's  
> C rating - I don't know lithium, but if a 90 ah battery can't deliver  
> at least 300 amps for more than a minute, you'll probably be on your  
> way to premature failure. higher ah batteries like the 180 are more  
> likely to be capable of delivering at least 2C continuously. My 100ah  
> nicads could only do 2C and 500 peak for 10 seconds  - it wasn't enough.
>
> just food for thought ;)
> Good luck!
>
> jrab
>

> On Jun 14, 2009, at 18:13, Peter Thompson <PThompson...@gmail.com>  

[DanP914]

I'm in the process of emailing Doug. In the meantime, I'm seriously
considering evcomponent's sale mentioned on the evdl list for
ThunderSky LiFePO4 at $1.05 / Ah. So many questions, so little $$$$.
1. Will 90 Ah do? Or do we need the 160 or 200?
2. What V will provide the performance we were expecting? 192 @ 160
Ah? 224 @ 90Ah? 192 @ 90 Ah?
3. What else needs to be upgraded?
a. Charger? Zivan NG3 can be programmed, I'm told, only to 170 or
190 V
b. DC-DC converter
4. BMS - evcomponents also sells.

Thanks for any ideas.
Dan

On Jun 14, 1:05 pm, "joe.obrien" <joe.obr...@cox.net> wrote:
> You might also want to contact Doug over at Blade Electric Vehicles.

> "doug.falco...@bev.com.au"

[David Hale]

On Wed, Jun 24, 2009 at 10:10 PM, DanP914 <dab...@ieee.org> wrote:

I'm in the process of emailing Doug. In the meantime, I'm seriously
considering evcomponent's sale mentioned on the evdl list for
ThunderSky LiFePO4  at $1.05 / Ah.  So many questions, so little $$$$.
1. Will 90 Ah do?  Or do we need the 160 or 200?

I would be very nervous about 90Ah until getting some more feedback from the people that are currently running with them.

 

2. What V will provide the performance we were expecting?  192 @ 160
Ah?  224 @ 90Ah?  192 @ 90 Ah?

This is a delicate balance between at least two, possibly three things. The obvious two are power (linear with voltage) and mass. With 160Ah batteries you might be looking at only marginal returns: 160s are still pretty heavy, maybe only a few hundred pounds less than lead, and the voltage may not be much more than our present voltage. I would expect an improvement (depending on the next question), but just how much, I don't know. 

The third possible variable is whether or not there are certain voltage levels that do not provide an improvement with the Azure system. The generic question is, do power & torque increase linearly with voltage? For example, Azure lists two voltages on their data sheet, 156V and 312V, and torque, power, etc. figures for each. If one moves from 156V to 170V is that better? to 190V? to 200V?, etc.? Or does one reach a plateau at some voltage, after which one doesn't see an improvement until one crosses a new voltage threshold?

3. What else needs to be upgraded?
  a. Charger?  Zivan NG3 can be programmed, I'm told, only to 170 or
190 V 

  b. DC-DC converter

(a) not sure, check with Electro Automotive

(b) yes, but where? check with E.A.

4. BMS - evcomponents also sells.

There's another ~ $2500.

Keep us posted; a lot of people I know are eager for more details on this subject.

--Dave

[Peter Thompson]

Hi Folks,

I talked to Doug, and he said that the 60Ah are too weak, but that the
90Ah would work. He also said that the 100Ah would be even better.
I've also gone and done a LOT of space calculations, and have decided
on the following configuration:

Thundersky 100Ah batteries. 34 in the front, 11 in the middle and 41
in the rear.
This creates a pack voltage of 275, with 19.2Kwh available.

I'm sending email to get the price of the components, and will be
ordering soon. I'll also let everyone know the cost of this setup.

Cheers,
Peter

On Jun 24, 11:17 pm, David Hale <astronomerd...@gmail.com> wrote:

[Mike S]

Hi Peter,

Not sure 275V will be good a good choice for the AC24 - DMOC445. I
believe it is too high for the AC 24 windings to be connected in delta
configuration. The other choice -- wye connection -- increases the
maximum nominal pack voltage to 336V. With a 275V pack, available
torque and power will be noticeably lower than if you had the full
336V. Beth Silverman at AZD will be helpful on this question.

Mike

[David Hale]

This is exactly what I was saying a few days ago that we needed to know the answer to. It may be that certain voltages provide no improvement.  You need to get the details on this before making a battery pack selection.

--Dave

[Peter Thompson]

Hmm, well, I'll try again. :) There is a configuration that will get
us to that voltage and will still fit into the battery racks, but that
uses the SE60, with 349v (can keep a few batteries out to get it down
to ~336), available power of 14.65kw. The good thing about going with
the Wye configuration is the higher torque and it uses less amperage.
Yes, the SE60 is a "weaker" battery, but the amount of amperage each
battery has to supply is less - 1.37A at max current. The SE60 can
handle up to 2A without trouble according to the various reports I've
read.

I'll be sure to pester, er, ask Beth about the preferred voltage for
the AC24.

Cheers,
Peter

On Jun 27, 2:21 pm, David Hale <astronomerd...@gmail.com> wrote:
> This is exactly what I was saying a few days ago that we needed to know the
> answer to. It may be that certain voltages provide no improvement.
>  You need to get the details on this before making a battery pack selection.
> --Dave
>

[David Hale]

hi Peter,

Something just occurred to me, pertinent to your battery decision. Even if the power provided by the DMOC does not increase linearly with voltage beyond 156V there may still be a benefit to a higher voltage. For example, maybe you get the same power by running at 190V as you do at 156V. But with the FLA I get 152V only when I'm fully charged and sitting still. As soon as I press the accelerator the voltage sags. Now at 138V I am immediately underpowered and I'm only 8 seconds into a fresh charge. I believe that LFPs don't sag as much like FLAs, but their voltage is going to drop with usage. So if you start out with more to begin with then even if the DMOC doesn't provide more power you are likely to remain above the 156V level for a longer period of time, and hence have the power you do have for longer.

What made me think of this is last night I decided to finally measure my real range. So I set myself up for a long drive. I got 60 "useful miles", which included just a few freeway miles and mostly 35-45 mph surface streets. After 60 mi I was limited to about 15 mph, less on steep hills (so those last 6 miles home took awhile). When I would stop the car, my pack voltage was ~ 128V, which is only a 15% DOD, but pressing the accelerator caused it to sag (below the limit of my meter, so I don't know how low) to such a level that I was limited to about 15 mph.

So, while it will be interesting still to hear what you find out about how the power output increases with voltage, I think any voltage you pick over about ~200V will be of some use.

best,

--Dave

[Peter Thompson]

HI Dave,

I just got through with a very nice email conversation with Beth at
Azure Dynamics. I told her the possible combination of batteries
(using 120Ah, 100Ah, 90Ah and 60Ah). After she consulted with the
various engineers and the capability of the AC24, she suggested that I
go with the SE60Ah pack (109 batteries, 349v nominal). The other
packs are good, but they are in the "no-man's land" between the Delta
and Wye configurations for the motor.

Advantages: Wye configuration on the motor. 150A max current. 82 Nm
torque. 17kw continuous, 47kw peak. Lighter pack.
Disadvantages: more connections between batteries. BMS required for
the batteries (true for all lithium packs tho). New charger. New DC/
DC converter.

I'll definitely see more pep out of the car using this pack due to the
higher output with the Wye configuration and the lower weight of the
pack. I'll have to be careful to not exceed the 400v max voltage of
the controller.

Now that I have the cash in the bank account, it is time to go
shopping. :)

Cheers,
Peter

[David Hale]

Peter, We will all be waiting with bated breath for you to go shopping. :) I'm glad to hear Azure was responsive to you. Did you find out what happens when the voltage sags? This is the plot that is missing from their data sheet, torque/power output vs. voltage. We have a spec for 312V. You're starting well above that, at 349V. What happens when it gets to, say, 305V; does it drop linearly, or does it drop like a rock?

It would be very valuable (for all of us, regardless of pack voltage or chemistry) to see torque/power output vs. voltage, for both configurations.

Thanks for working on this.

--Dave

[joe.obrien]

An additional quote you should at least follow up on; since you're
looking at quite a few batteries; is Brad at Jungle Motors. They have
palates of batteries, since he goes for bulk purchases. He might be
interested in how this works out for you.
Joe

[David Hale]

Peter,

I came across this Hyundai this morning. It's not completely relevant but somewhat, and interesting nonetheless. He has been running with LFPs for 6 months, and he's getting astonishing range. Unfortunately it's not with the Azure drive. I still think it will be absolutely necessary to know the power curves with voltage before making a battery purchase.

http://www.evalbum.com/2314

also an S2000 that has a fair number of miles on LFPs:
http://www.evalbum.com/2394