Ethan in Germany with industrial AC and LiPO

Technical discussion on converting internal combustion to electric
User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Ethan in Germany with industrial AC and LiPO

Post by acmotor » Fri, 29 May 2009, 23:48

I don't know if anyone has been following this one....

http://rc-autopilot.de/wiki/index.php/O ... sa_Project

I have been emailing Ethan since last year with lots of discussion on DC bus connections to VFDs. He has just done some driving.
The battery pack is interesting. No BMS thought ??
iMiEV MY12     110,230km in pure Electric and loving it !

arnolde
Noobie
Posts: 10
Joined: Sat, 30 May 2009, 00:16
Location: Germany

Ethan in Germany with industrial AC and LiPO

Post by arnolde » Sat, 30 May 2009, 00:54

Thanks for the introduction, Tuarn :-)

Yes, no BMS yet. I know almost everybody will tell you thats suicide with Li-Ions. But even a simple BMS would cost >1000$ for all 150 cells, so I'm waiting to see if it's really necessary. There are projects without any BMS and they prove that it's not as important as most people think, if you're very careful about charging/discharging. (see video at http://web.me.com/mjrickard)

I'm currently looking very hard for a suitable charger. The basic idea is to take the 700VDC from a modified PC power supply (has all the PFC already built in, just need to change the way the rectifier is connected to double the DC from 370V to over 700V) and add a PWM circuit (and isolation) to make the output current/voltage adjustable. If that works (by manual knob turning at first), I'll add a microprocessor which can then be programmed with a charge curve.
Last edited by arnolde on Fri, 29 May 2009, 15:01, edited 1 time in total.

antiscab
Senior Member
Posts: 2490
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Ethan in Germany with industrial AC and LiPO

Post by antiscab » Sat, 30 May 2009, 23:49

The charger idea sounds interesting.
When you get it going, be sure to post an update :)

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

arnolde
Noobie
Posts: 10
Joined: Sat, 30 May 2009, 00:16
Location: Germany

Ethan in Germany with industrial AC and LiPO

Post by arnolde » Sun, 31 May 2009, 12:17

Its not an easy task. Wherever I ask I either get quoted prices >3000 EUR (from manufacturers) or get laughed at, or at least not much help, in electronics forums...

User avatar
Thalass
Senior Member
Posts: 741
Joined: Sun, 12 Aug 2007, 07:29
Real Name: Ben Rypstra
Location: Perth, WA, AU

Ethan in Germany with industrial AC and LiPO

Post by Thalass » Sun, 31 May 2009, 15:33

Nice conversion. How much did the battery pack cost, if you don't mind me asking?

And with a controller that price, I'd be tempted to buy two to run both my motors in parallel the same way the electric imp is set up!

So the controller has a serial input/output for realtime computer control? Or is that just a connection to tweak settings?


*edit* Looking at their website, these motors and controllers look pretty good! Shiny.
Last edited by Thalass on Sun, 31 May 2009, 05:49, edited 1 time in total.
I'll drive an electric vehicle one day.

arnolde
Noobie
Posts: 10
Joined: Sat, 30 May 2009, 00:16
Location: Germany

Ethan in Germany with industrial AC and LiPO

Post by arnolde » Mon, 01 Jun 2009, 00:34

The controller (MDV60A-15kW) I use normally costs about 2000 EUR new, you can get them around 1000 EUR used, I was very lucky to get it for 300.

The RS422 interface (long distance (>100m), potential free, can be easily adapted to RS232) is only for configuring the controller when not operating. But you can activate a data-capturing sequence (with adjustable resolutions as fine as 1ms) to capture parameters (torque, speed, battery voltage, motor current) in a loop and read that out afterwards, so you dont need to have a notebook recording data while driving.

I must admit this model (MDV-60A) is almost 10 years old, so newer models might be even better and more flexible. The downside is that it can only do voltage-to-frequency mode (=gas pedal determines motor speed), and not torque mode, unless you hook up an (expensive) encoder to the motor shaft. So thats what I'm looking for next.

arnolde
Noobie
Posts: 10
Joined: Sat, 30 May 2009, 00:16
Location: Germany

Ethan in Germany with industrial AC and LiPO

Post by arnolde » Thu, 04 Jun 2009, 23:36

Update on the charger:
On http://ecomodder.com/forum/showthread.p ... -6795.html we are developing and building a high voltage (>100V) charger. One guy is already charging his 144V battery with it at 20A, I'm working on my 600V version right now.
Last edited by arnolde on Thu, 04 Jun 2009, 13:37, edited 1 time in total.

User avatar
Johny
Senior Member
Posts: 3729
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Ethan in Germany with industrial AC and LiPO

Post by Johny » Thu, 04 Jun 2009, 23:54

The downside of a 600 VDC charger (IMO) is that the pack can't be broken up into say 48V or 60V sub-packs during charge. It's likely you may wish to work on the car during charge so charging with many, lower voltage chargers is (to me) preferable.

The general consensus here for 600 VDC systems has grown to be to have "break up contactors" as an adjunct to the main contactor. These break the battery system into non-lethal voltage sub-packs when the vehicle ignition is off (or when emergency etc. cause them to open).


arnolde
Noobie
Posts: 10
Joined: Sat, 30 May 2009, 00:16
Location: Germany

Ethan in Germany with industrial AC and LiPO

Post by arnolde » Fri, 05 Jun 2009, 00:05

Thats true, however I personally dont like the idea of having 10 contactors and 10 seperate chargers. It just isnt good engineering, and adds unnecessary weight and lots of wiring and components that can fail. I limit the HV wiring to a minimum and just never touch it again, period. I have no reason to open the battery casing once it's installed. The main contactor is inside, and it's all one box, so when the ignition is off, there's no HV. I do understand though that when using lead batteries, it's not that simple, because they're usually located in several different places.

As far as efficiency and engineering goes: The less you convert, the better. Losses are always introduced when converting. So when I can get my 600V practically right from the grid, I want to regulate that as little as possible. This way I can get the whole charger in the size & weight of a standard PC power supply and have well over 90% efficiency. Thats just not possible with 10 seperate chargers that step down 1:10 or so.
Last edited by arnolde on Thu, 04 Jun 2009, 14:08, edited 1 time in total.

User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Ethan in Germany with industrial AC and LiPO

Post by acmotor » Thu, 11 Jun 2009, 07:34

Arnolde, a sealed traction battery pack with internal fuses and contactors is a fair concept to work with.

My concern is that your LiPo pack does not have BMS yet.
What plan do you have to keep track of what is happening in the pack ?
Temperature monitoring ?

One of the reasons that people go for multiple low voltage chargers is that BMS is less critical in series battery strings with less cells and the risk of battery unbalance is reduced. It also means that any failure is at considerably lower voltages, bearing in mind that most EV 'incidents' seem to occur during recharge not discharge.

What intelligence do you plan for your charger ?
iMiEV MY12     110,230km in pure Electric and loving it !

arnolde
Noobie
Posts: 10
Joined: Sat, 30 May 2009, 00:16
Location: Germany

Ethan in Germany with industrial AC and LiPO

Post by arnolde » Thu, 11 Jun 2009, 10:58

Please don't call it LiPo, it is not Lithium Polymer but Lithium Iron Phosphate, a totally different thing, even though they both start with "lithium".

I plan lots of intelligence for the charger and pack, about anything you can think of. But first I need to prove that the power stage works in principle.

User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Ethan in Germany with industrial AC and LiPO

Post by acmotor » Thu, 11 Jun 2009, 16:49

Sorry, my mistake re the cells. They look like LiPo bags and you talk on your site about..........
"Using 3.7V lipo batteries, 631V means 170 cells in series (16Ah each for 10kWh)" My mistake for not paying enough attention.Image

Help me understand here. It is a very interesting battery pack.
You have 11 x 48V 'blocks' = 528V nominal and this is 150 cells ?
This is 3.5V per cell ? but 13.636 cells per block ?

The cells from HuanYu are HPPF70173248 ? Please clarify. Image
iMiEV MY12     110,230km in pure Electric and loving it !

antiscab
Senior Member
Posts: 2490
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Ethan in Germany with industrial AC and LiPO

Post by antiscab » Thu, 11 Jun 2009, 18:09

sounds like the battery description in the project web page hasnt been updated in quite some time.

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

arnolde
Noobie
Posts: 10
Joined: Sat, 30 May 2009, 00:16
Location: Germany

Ethan in Germany with industrial AC and LiPO

Post by arnolde » Thu, 11 Jun 2009, 21:53

Right, sorry about the confusion, I wrote the Wiki in different parts at different times. I do mention Lipo somewhere but that was before I got the battery, and was still in the discussion phase. Somewhere later on I describe exactly what I bought: 168 cells Huanyu HPPF70173248-20AH, of which 2 were damaged (deformed, they still work but I decided not to use them for now) when shipping and 1 which I tore a tab off when crimping together (before I found the right method). So I have 11 blocks of 15x3.2V (about 1kWh) each, totalling approximately 550V (each block had 50.2V when assembled, before any charge/discharge). Now I didnt know exactly how low a voltage the inverter would work with, the tech support couldnt tell me exactly, the specs suggested it would be 480V but I've successfully fired up the controller with less (450V), so I'm not exactly sure yet until I test it under high load. But anyway it works fine on 500V (sags to about 488V when under as much load as I can apply inside the warehouse when accelerating) so I decided to stick with 10 packs (500V) for now because they fit nicely into a 60x40x30cm basket. (The photo shows a 60x40x40cm open basket, meanwhile I've got a 60x40x30cm closed one).

The batteries can deliver 2C constant (20kW) and 3C (30kW) peak without reducing lifetime (at least, so Huanyu says). They can even deliver 5C (50kW) but they do heat up and it probably shortens the life, and max lifetime is top priority for me. By only using 70-80% DOD and never charging them up to the absolute maximum (3.6V is enough, no reason to go to 4V or even 4.2V like many people do, only adds a few % to the charge but really enhances the risk of overvoltage), I hope to get well over 1000 Cycles, which should be well over 5 years of use, I'm actually hoping for 10 years. Keeping the batteries under 30°C is also necessary for a long life. (Of course, not really possible when the car is standing outside in summer when its over 30°C, but that should only be a couple months per year in Germany)

Since I only have a 7.5kW motor (rating at 1500rpm), it will draw 15kW at 3000rpm, and maybe I can overdrive it a little, and so reach 20 or 25kW. The controller is rated at 15kW and can do 150% (22kW) as long as the temperature stays ok. So the battery pack should never get more than 25kW drawn (50A), and even that is a peak rating at max rpm under max load, which should not happen often. (Thats something I'll just have to find out when test driving, and make a driving profile).

For those who are doubtful about my no-BMS-approach, please watch the video here: http://web.me.com/mjrickard - Jack does just fine without any.

Ethan
Last edited by arnolde on Thu, 11 Jun 2009, 12:09, edited 1 time in total.

User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Ethan in Germany with industrial AC and LiPO

Post by acmotor » Thu, 11 Jun 2009, 23:12

Thanks for clearing that up. We are on the same page now (almost !)

Data sheet for HPPF70173248
........ still has me wondering as it does not mention Lithium Iron Phosphate although nominal voltage is 3.2V Image

I guess you have done more communicating with Huanyu though.

The internal impedence may limit your peak power expectations ?
15 x 10 x 3.2 = 480V - 12mohm x ~(22kW/480)A x 150 cells = ~400V at 46A
Yes, controller is likely to drop out at 470VDC. I'd put the extra cells in !

My experience with 3PIM is that you will not get more than 7.5kW at 3000RPM unless you can provide more voltage to maintain the v/f ratio however you will be easily able to get to 22kW on the 7.5kW motor at less than 1500RPM.   forum topic on AC

Re BMS.
I know you say 3.6V is charged enough (agreed), but how do you know you have reached that voltage on one cell and not another ? Out of sight and out of mind is not out of trouble. Image
I see early RC model aircraft used no BMS on multi cell Lithium packs, then they used BMS on charging then latest is at least individual cell voltage monitor on discharge.
I consider Jack and his Porche are only getting away without BMS while the cells are new. There are plenty of people who have rubbished Thundersky and other lithiumXX cells, but probably killed them without BMS.

I would be delighted to go without BMS, however the long term prospects are not good in my understanding. You may be about to prove us wrong !
The intellegence you build into the charger may make all the difference ! Image

Did Huanyu comment on the use of BMS ? Their website seems to be a bit incomplete at present. Image   
iMiEV MY12     110,230km in pure Electric and loving it !

Tritium_James
Senior Member
Posts: 683
Joined: Wed, 04 Mar 2009, 17:15
Real Name: James Kennedy
Contact:

Ethan in Germany with industrial AC and LiPO

Post by Tritium_James » Thu, 11 Jun 2009, 23:24

No BMS and a 10 year lifetime are not compatible things in my opinion. I might be biased, we make a BMS...

And that Porsche is NOT a good example of running with no BMS. I quote from his blog:
In any event, we did drive the car just under 100 miles (160km) with the two extra batteries. We DID DESTROY A BATTERY IN THE PROCESS proving rather emphatically the dangers of over discharge. One of the batteries added was just not quite balanced in with the rest of the pack, although we did add some energy to it when we put it in. After discharging, it ran out FIRST and went to zero, blocking any further current from the rear pack.

User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Ethan in Germany with industrial AC and LiPO

Post by acmotor » Fri, 12 Jun 2009, 00:01

At least someone read the blog. It was DC motor and no BMS on the baddery. I just looked at the pictures and vid ! Image

But, I'll start a thread on BMS free battery pack design.
Ethan may be showing us up ! Image

BMS free battery management
iMiEV MY12     110,230km in pure Electric and loving it !

djsharpe
Groupie
Posts: 90
Joined: Thu, 21 Feb 2008, 23:43
Real Name: David Sharpe
Location: Melb

Ethan in Germany with industrial AC and LiPO

Post by djsharpe » Sun, 28 Jun 2009, 16:18

I have only 46 LiPs in my humble Charade and it is series DC (AC folk heard to scoff in disgust). I dont use a BMS. I monitor the weakest cell & cut off charge when it hits 4.25V. The problem is that the weakest cell changes occasionally. Occasionally I go around with a 4V power supply & bring all cells up to 4.25V. Recently I bought an additional NG3 charger which has user changeable charging profiles. You can cut the charge at a safe voltage. I have found that there is little to be gained by going to 4.25V. 4.0V is probably enough.   

User avatar
coulomb
Site Admin
Posts: 3780
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Ethan in Germany with industrial AC and LiPO

Post by coulomb » Sun, 28 Jun 2009, 17:02

Tritium_James wrote: ... I might be biased, we make a BMS...

What sort of BMS do you make, TJ? I don't see it on the Tritium web page.

If the details are not ready for release yet, that's fine.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

Tritium_James
Senior Member
Posts: 683
Joined: Wed, 04 Mar 2009, 17:15
Real Name: James Kennedy
Contact:

Ethan in Germany with industrial AC and LiPO

Post by Tritium_James » Sun, 28 Jun 2009, 17:45

Hmm, I can't believe I forgot to show it to you the other weekend! We've got several different designs, they're all running in various cars or on the bench, but I want some more hours on them before doing the work to put them up for sale.

My latest design for the 90Ah TS cells is almost exactly the form factor you describe in the other thread. One board per four cells though, as I've got a micro and isolated CAN comms, so one board per cell adds up to too much $$. It monitors voltage on each end of the four cells (ie redundant measurements, and voltage drop across busbars gets removed) as well as the temperature of the cell it's physically connected to.

Crappy phone photo:
Image


antiscab
Senior Member
Posts: 2490
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Ethan in Germany with industrial AC and LiPO

Post by antiscab » Sun, 28 Jun 2009, 23:14

djsharpe wrote: I have found that there is little to be gained by going to 4.25V. 4.0V is probably enough.   


you are among the few who charge to (or above) 4v.
most of us charge to 3.6v (actually, anything above 3.4v is fully charged)

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

djsharpe
Groupie
Posts: 90
Joined: Thu, 21 Feb 2008, 23:43
Real Name: David Sharpe
Location: Melb

Ethan in Germany with industrial AC and LiPO

Post by djsharpe » Mon, 29 Jun 2009, 01:02

http://www.thunder-sky.com/pdf/2007030222.pdf

This indicates that the cell is only 50% charged if charging at 0.5C if you cut the charge at 3.7V. Need to see a family of curves for diff charging rates.

arnolde
Noobie
Posts: 10
Joined: Sat, 30 May 2009, 00:16
Location: Germany

Ethan in Germany with industrial AC and LiPO

Post by arnolde » Mon, 29 Jun 2009, 01:48

The thundersky batteries are not 100% identical to other manufacturer's batteries. The charging curve I have from Huanyu shows that at 3.7V the battery has received over 90% of its charge.

User avatar
coulomb
Site Admin
Posts: 3780
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Ethan in Germany with industrial AC and LiPO

Post by coulomb » Mon, 29 Jun 2009, 04:00

djsharpe wrote: http://www.thunder-sky.com/pdf/2007030222.pdf

This indicates that the cell is only 50% charged if charging at 0.5C if you cut the charge at 3.7V. Need to see a family of curves for diff charging rates.

I would stay it's a lot more than 50%:

Image

Note that the green, dark blue, and brown lines are official Thunder Sky curves from the above document. The orange line is one I drew in stopping at 3.7 VPC. The light blue and purple lines are my total guesswork, especially the light blue one!

Notice how even with the official charge to (ulp) 4.3 VPC, the cell seems to need 40 minutes of CV charging to get to full capacity. If you can take the curves literally, you can cut the charge off at about 0.1C. So you would not stop the charge the instant the voltage reaches 3.7 VPC; that would indeed be about 50% capacity.

In fact, you would keep the voltage at 3.7 VPC (the CV part of the CC-CV charge algorithm), and if my guess happened to be correct, you'd need to keep it there from about 75 minutes to about 200 minutes. (It could be much quicker than that.) It might take that long for the current to taper down to 0.1C, or as I happen to have drawn it, it might go to almost zero before maximum charge is attained. Whether the resultant SOC is about 95% as I have drawn it above or not is to be determined. (Or can some of you guys with long Lithium experience provide some better figures?)

But I think it's certain that stopping at 3.7 VPC, even for Thunder Sky cells, with a suitably long CV period, will lead to better than 50% SOC.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

antiscab
Senior Member
Posts: 2490
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Ethan in Germany with industrial AC and LiPO

Post by antiscab » Mon, 29 Jun 2009, 09:45

the thundersky graphs aren't all the accurate tbh.

If you do a single stage charge, and end when the voltage reaches 3.7v at constant current, of course it wont be fully charged.

on the scooter, i use 0.05C at 3.65vpc (average) as the end of charge cut-off.

when bench testing a 40AH cell, i found the difference between charging to 3.4v and 4.2v (and letting current fall to 0.05C before ending charge) was a less than 1Ah.

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

Post Reply