[spoland]

Hi everybody,

I'm planning to do an EV conversion and right now I am considering converting an Audi A2. I like Audis and drive an A6 right now. The A2 is about enough size for my family (three persons and two dogs) to be able to commute 60km per day. I am hoping that its low weight will give good performance and range and that the aluminium body will make it last long.
I haven't been doing much auto mechanics for a while (I changed the transmission in a Saab about 20 years ago , but I know the basics about cars and how they work. I don't want to fabricate anything advanced (like adapter plates), but smaller basic stuff won't be a problem.
I'm hoping to get a range of 100-160 km (60-100 miles) and be able to cruise at highway speed 110 km/h (70 mph). My budget is about EUR 20,000 (USD 25,000) including donor car which I haven't got yet. I live in northern Sweden and in winter it gets quite cold here. -20C (-4F) is not uncommon and -30C (-22F) happens a couple of times each year.
The A2 has a Cd=0.28, A=2.2m2, m=1065kg (I'm guessing 1260kg after conversion=2780lbs)

The parts I have considered so far are:
Motor: HPEVS AC50 (I like the idea of not having to maintain the motor at all)
Controller: Curtis 1238-7601
Adapter: Canadian EV New Beetle adapter
Batteries: 36 Winston (formerly Thundersky) 200 Ah LiFeYPO4 cells giving a nominal voltage of 115V and energy content of 23 kWh. (Not sure if this will work since the nominal voltage of the Curtis is 96V, but some claim it will work just fine) The reason for using 36 instead of 30 cells is mainly to increase range. I don't really know the difference between LiFePO and LiFeYPO4 batteries. Some claim the LiFeYPO4s have better cold performance.
BMS: No BMS. I think I will follow Jack Rickard's advice and undercharge the cells to 3.5V instead of 3.6V to improve their life and not use a BMS.
Charger: A programmable charger that lets me create a constant current/constant voltage charge process with custom current and voltage. If it is possible I would like the charger to take as input 400V,16A which is standard in Sweden if you use the voltage between two phases in a three phase outlet. That would give me a charge power of a little more than 6kW and charge time of around three hours. I haven't found one yet. Any suggestions? Plan B is 230V,16A.
Vaccum pump for brakes: Yes
Power steering: Don't know if I need that. I will probably try without and get an electric 12V pump if needed.
AC: Will probably get a donor car without AC
Heater: Oh yes! I think I will need a 2-3 kW ceramic heater or a water heater

In Sweden we have some peculiar regulations regarding home built autos. To simplify it a little (a lot actually), if you convert a car older than 1993 you can do pretty much what you want, but if you convert a car newer that 1993 things get more compliated. If you only change the engine, the build is called an "altered car"' and then you need the car manufacturers certificate that the new engine will meet all emission an other demands on new cars. In practice this is impossible to get if you don't change to another ICE-engine that is part of the program for that car model. So this is not doable. If, however, you change the car more than this, for exampel change the engine, transmission, and brakes, it is called a "rebuilt car", and then you do not need any cerificates from the manufacturer and everyting gets much easier! That is apart from all the extra work that is needed to change the tranny and brakes that does not contribute neither functionally nor improves security of the final car. So unLEAN!
Since I want to use an Audi A2 from 2001-2003 I have to do this as well.
Transmision: VW Golf Mk IV, 5 speed (I hope the half shafts will fit)
Brakes: VW Golf Mk IV, discs front and rear plus main cylinder. The A2 has small discs in front and drums in rear so bigger discs will be OK.

Any thoughts on this conversion?
Thanks!
/ Peter

 

[EVEngineeer]

I found that motor on a website for 4.5k, but I also saw that motor and controller in a pack on http://evolveelectrics.com for 4.6k. Also, AC motors probably won't be a good way to go, because of your budget. You can achieve what you are looking for by going with a DC motor, like the WarP 9. That motor will only cost you 2k.

 

[cpct]

Can you give a reason why you don't want to use a BMS?

I mean, a large part of the budget is batteries, they will be used in harsh environment (very cold) and need to fullfill high range requirements. Leaving out a BMS will shorten the lifespan of your batteries significantly and it does not cost *that* much...

 

[spoland]

I found that motor on a website for 4.5k, but I also saw that motor and controller in a pack on http://evolveelectrics.com for 4.6k. Also, AC motors probably won't be a good way to go, because of your budget. You can achieve what you are looking for by going with a DC motor, like the WarP 9. That motor will only cost you 2k.

Thanks EVEngineer!
I think will also have a look at DC motors. I guess the Warp 9 would give me a little less power (32HP compared to 50HP for the AC50), no regen and also I would have to change the brushes from time to time. But, as you mention, with a Curtis 1231 controller for 1,700 I would save almost $1000.

 

[spoland]

Can you give a reason why you don't want to use a BMS?

I mean, a large part of the budget is batteries, they will be used in harsh environment (very cold) and need to fullfill high range requirements. Leaving out a BMS will shorten the lifespan of your batteries significantly and it does not cost *that* much...

I'm really not sure about this, thats why I'm asking everyone here. To me there seems to be a debate about BMSes where some argues that this is a must have and some, Jack Rickards of EVTV for example, claims BMS'es are nonsence.
I'm not sure of the precise arguments against BMS but i believe it has to do with the self balancing that occurs in packs of LiFEPOs anyway, the fact that a BMS turns off charging of a single cell based on the voltage over that cell whereas the charging state of the cell is much more complicated than that, and that it is better to undercharge your whole pack of batteries to prolong its life instead of trying to maximise energy content.
I sure am no battery specialist, but from what I can read in different foras there is no consensus regarding BMS in the EV community and they do cost a bit.

 

[drgrieve]

You don't need a BMS as such, but you do need to know how to charge the batteries and have a reliable charger (with a possible fail-safe - timer or secondary voltage or ah counter) and obviously a good ah counter for discharge.

In regards to parts since you are in Europe look at the kostov 9" 220v (or a kostov 10"). Match these up with a Soliton Jnr. Peak output should be around 130kW for the 220v in this combination. It's also a very light motor at 45 kgs.

BTW the warp 9 (with the new brushes) peak output is ~200kw when combined with a Soliton 1. Which is a little more than 25hp!

In regards to cells seriously look at the new calb greys. They are meant to have improved cold weather performance. Get the 100ah cells so you have at least 240v (~75 cells) to counteract voltage sag thus maintaining reasonable current output at a higher rpm.

In regards to brush maintenance. If you keep the motor accessible this is 20 minute job. Not something to worry about. Not sure on brush life but for normal street driving it would seem to be in the "last forever" category of things.

 

[steven4601]

For all the DC conversions in the E.U.... Warning: It is hard to get EMC (R10) & electrical safety (R100) compliance with (large) DC traction motors.
If you dont use C.E. certified components with according EMC tests for automotive usage you'll require to do an EMC vehicle test. ~ 2000 to 5000 euro for 1 test. which may fail and you have to do another one.
Testing can be more expensive than a battery pack.

Electrical compliance problems comes from the isolation requirement (described in the R100) between the traction pack & vehicle chassis. Carbon dust creates an electrical path between the contacts of the motor and the chassis causing it to fail the insulation test. Also this creates common mode currents causing higher chances it to fail EMC requirements.



Read first what is pubished regarding requirements, call officials regarding regulations towards vehicles. choose parts, budgetize, then buy parts and start converting.


//Steven

 

[Tesseract]

For all the DC conversions in the E.U.... Warning: It is hard to get EMC (R10) & electrical safety (R100) compliance with (large) DC traction motors. ...

You've been around here long enough to know better. Evnetics controllers from Rebbl have certificates of approval for on-road use, and so do Kostov motors.

 

[EVEngineeer]

I do not know how much if at all of the wiki you have looked at, but this one is important to you at this moment. http://www.diyelectriccar.com/forums/showthread.php?t=73869

 

[steven4601]

The rebbl emc filter package for the soliton motorcontroller are a solution for reducing the EMI emission, what about the isolation / chassis leak with a DC motor? The easiest way to tackle this would be to use a brand new motor for the test. (as the motor is clean, free from carbon brush dust )

 

[cpct]

I'm really not sure about this, thats why I'm asking everyone here. To me there seems to be a debate about BMSes where some argues that this is a must have and some, Jack Rickards of EVTV for example, claims BMS'es are nonsence.
I'm not sure of the precise arguments against BMS but i believe it has to do with the self balancing that occurs in packs of LiFEPOs anyway, the fact that a BMS turns off charging of a single cell based on the voltage over that cell whereas the charging state of the cell is much more complicated than that, and that it is better to undercharge your whole pack of batteries to prolong its life instead of trying to maximise energy content.
I sure am no battery specialist, but from what I can read in different foras there is no consensus regarding BMS in the EV community and they do cost a bit.

The problem is that without a BMS, cell differences (not one cell is exatcly equal tot the other) lead to one cell discharging or charging faster than the others. After some 10s of cycles, the differences get bigger and bigger until that one cell is consistently overcharged/overvolted or undervolted.

It will work for a while though.

 

[drgrieve]

The problem is that without a BMS, cell differences (not one cell is exatcly equal tot the other) lead to one cell discharging or charging faster than the others. After some 10s of cycles, the differences get bigger and bigger until that one cell is consistently overcharged/overvolted or undervolted.

It will work for a while though.

Just to be clear we are talking about LiFeP04 batteries.

<Begin Rant>

Huh? What planet are your from? Endless Sphere?

There is no such problem. FYI many members on this forum don't use a BMS. They have no issues. There is no cell drift.

Cell drift is a lie.

Perhaps something for you to read up on.

PS Individual cell monitoring BMS's can themselves cause of cell drift. Remove them and the issue is no more.

<End Rant>

 

[cpct]

Just to be clear we are talking about LiFeP04 batteries.

<Begin Rant>

Huh? What planet are your from? Endless Sphere?

There is no such problem. FYI many members on this forum don't use a BMS. They have no issues. There is no cell drift.

Cell drift is a lie.

Perhaps something for you to read up on.

PS Individual cell monitoring BMS's can themselves cause of cell drift. Remove them and the issue is no more.

<End Rant>

Touchy subject it seems.

I cant say anything about the extent, but drifting principle is not unique to the chemistry. The internal resistance of the cells will never be exactly equal, neither will the capacity. So voltage will drift inherently. Maybe if you equalise the cells initially when assembling the pack the problem can be reduced, but it will not be gone.

 

[spoland]

...
In regards to parts since you are in Europe look at the kostov 9" 220v (or a kostov 10"). Match these up with a Soliton Jnr. Peak output should be around 130kW for the 220v in this combination. It's also a very light motor at 45 kgs.

BTW the warp 9 (with the new brushes) peak output is ~200kw when combined with a Soliton 1. Which is a little more than 25hp!

In regards to cells seriously look at the new calb greys. They are meant to have improved cold weather performance. Get the 100ah cells so you have at least 240v (~75 cells) to counteract voltage sag thus maintaining reasonable current output at a higher rpm...

Thanks, I'll have a look at those.

 

[spoland]

For all the DC conversions in the E.U.... Warning: It is hard to get EMC (R10) & electrical safety (R100) compliance with (large) DC traction motors.
If you dont use C.E. certified components with according EMC tests for automotive usage you'll require to do an EMC vehicle test. ~ 2000 to 5000 euro for 1 test. which may fail and you have to do another one.
Testing can be more expensive than a battery pack.

Electrical compliance problems comes from the isolation requirement (described in the R100) between the traction pack & vehicle chassis. Carbon dust creates an electrical path between the contacts of the motor and the chassis causing it to fail the insulation test. Also this creates common mode currents causing higher chances it to fail EMC requirements.



Read first what is pubished regarding requirements, call officials regarding regulations towards vehicles. choose parts, budgetize, then buy parts and start converting.


//Steven

Thanks Steven!
I'll keep that in mind.
/ Peter

 

[spoland]

Touchy subject it seems.

I cant say anything about the extent, but drifting principle is not unique to the chemistry. The internal resistance of the cells will never be exactly equal, neither will the capacity. So voltage will drift inherently. Maybe if you equalise the cells initially when assembling the pack the problem can be reduced, but it will not be gone.

See what I meen about non consensus?

 

[Tesseract]

The rebbl emc filter package for the soliton motorcontroller are a solution for reducing the EMI emission, what about the isolation / chassis leak with a DC motor? The easiest way to tackle this would be to use a brand new motor for the test. (as the motor is clean, free from carbon brush dust )

The isolation requirement in ECR-R 100 is so easy to meet it hardly qualifies as an obstacle: 100 ohms per volt for DC circuits and 500 ohms per volt for AC circuits. Even the two WarP-9s in our dyno can still pass this test (and by at least an order of magnitude) despite 4 years of abuse.

 

[drgrieve]

Touchy subject it seems.

I cant say anything about the extent, but drifting principle is not unique to the chemistry. The internal resistance of the cells will never be exactly equal, neither will the capacity. So voltage will drift inherently. Maybe if you equalise the cells initially when assembling the pack the problem can be reduced, but it will not be gone.

Sorry for the rant just annoying when people post uninformed information as fact.

Please provide proof of cell drift where there is no BMS involved.

Will wait while the crickets chirp.

Some batteries types do drift. Hence the confusion. LiFeP04 don't drift as there is no internal mechanism for it. The ions can only move when the terminals are connected.

So no cell drift and no cell discharge.

 

[cpct]

Sorry for the rant just annoying when people post uninformed information as fact.

Please provide proof of cell drift where there is no BMS involved.

Will wait while the crickets chirp.

Some batteries types do drift. Hence the confusion. LiFeP04 don't drift as there is no internal mechanism for it. The ions can only move when the terminals are connected.

So no cell drift and no cell discharge.

So I have to prove what you don't have to, haha. But ok, here is what I know.

First, maybe you mean something else then I do when talking about cell drifting. Of course, if the cells are disconnected, there is only some limited self discharge (in the order of a few percent of capacity per month from what I've gathered).

However, when current is passed through them, the cells can be represented as an equivalent model of a resistor and voltage source. So for a complete pack we have chain of sources and resistors. If even one of the equivalent resistors is not the same as the others, the voltage drop over that cell will be different than for the other cells. This means that after a while, this cell will be critical (too low under load or too high voltage when charging).
As it happens, this equivalent resistance of a cell depends on a lot of possible factors, such as temperature, age, soc, manuf. process etc... Very hard to model. Anyway, no cell will thus behave exactly the same. Sooner or later, not properly equalising the cells in a Li pack will finish it.

 

[Ziggythewiz]

He's right. No two cells are identical, and over time they will drift...by a couple milliAH. They may drift closer, they may drift further depending on the volts, amps, gravitational distortions etc, so after years of use the imbalance may reach .1 V or so.

Who cares??? Don't try to max your voltage and no worries.