Homemade EV HVDC Battery Charger

BMW i3 Forum

Help Support BMW i3 Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

iainventor

Member
Joined
Jun 12, 2018
Messages
14
hi there

I am experimenting with building my own EV charger and range extender. I am hoping someone on here can help me. I am thinking of using a few 5kw hr refurbished Tesla battery packs = 90 x 4.1 volts charged lithium ion in series something like 369 volts full pack.

And what is the nominal voltage of the i3 60 AH battery. what constant charging voltage does it take to recharge the battery ?

And then creating my own battery charger for the 5kw hour brick , maybe to match the BMW i3 pack will need to take a few 4.1v lithium cells out and then though what voltage is the nominal and charge voltage that I would need for my homemade charger to match the onboard EV pack?

thanks Im probably leaving out something want to build my homemade range extender to tie into BMW i3 HV DC power cables and flip a switch while driving to charge and drive with my added lithium ion packs.

thanks in advance
 
i just found tech manual said nominal voltage 360 VDC can be as low as 290 VDC and as high as 396 VDC. Said 98 batteries X 3.75 VDC nominal in series = 360 VDC

So then the 369 VDC of the refurbished Tesla Pack should at least charge to the output voltage of 360 VDC when put in parallel then going to splice the V+ and V- at power line from battery to EME and put blocking diode and NO contractors as well between my extra pack and EV battery to turn on only after EV startup then turn on the contractors to my pack and feed high voltage dc into the path of battery and EME.

And I think just like on a Leaf I seen the guy had to turn off the power of his pack before the EV was shut down.

might work will know in about a week or so.

has anyone on this site ever done this before? If so, any luck or what issues came up?
 
THere is a communications protocol that must be negotiated for a CCS charge to properly occur. The two line voltage pins, when trying to do a CCS charge get serial data superimposed over it that communicate between it and the car. This tells the CCS unit what voltage the car wants at the moment so that the charging rate is not going to damage things.

I'm not sure the car will actually charge itself without that communication and interlock functionality. It could be a very expensive experiment if you get it wrong. The actual specifications document, I'm sure, can be purchased, but you might have trouble finding it on-line for free.
 
thanks, for your input. the 2014 model I have doesn't have a ccs just level 2. I was just going to splice the hv dc battery cable between battery and inverter power.
 
iainventor said:
thanks, for your input. the 2014 model I have doesn't have a ccs just level 2. I was just going to splice the hv dc battery cable between battery and inverter power.
IMHO, seems like a good way to fry your i3...dealing with high voltage, high current requires some very good understanding of the system. While the standards for hooking up a CCS certified device to a vehicle with the proper hardware is available, figuring out what the actual car's hardware is is proprietary verses being an open standard for the CCS unit itself and it's interaction with the vehicle attached. There are both safety interlocks and serial datalinks between a CCS unit and the vehicle. Then, you'd also have to activate the cooling circuits and charging logic that monitors charging to prevent overcharging or overheating, or risk damaging the cells themselves, and any other devices that may be connected to the HV.
 
It sounds like you are looking at:
  • parallel 5kWh tesla and 18.4 kWh BMW battery - if they are at the same voltage, connect once inside the power relays and pray the different battery chemistries don't let the smoke out. The probability of identical electro-chemical, battery specs match is null and void. Don't do it.
  • DC-to-DC buck-boost converter ~500W - the car uses ~250W/mile. Using a DC-to-DC converter you can feed the BMW battery without risking mismatch of the Tesla pack and BMW. You can also control the current feed and ramping in at a rate that doesn't "let the smoke out."
The earliest, homemade, plug-in hybrid experiments with the Gen-2 Prius failed because they tried to 'relay' connect the external pack to the traction battery. The successful versions used a DC-to-DC converter to managed the current and voltage differences.

Did I guess right about your goals and objectives?

There are more technical issues but I need to understand your approach.

Bob Wilson
 
Bob

Hi sorry been a few weeks since I had a chance to be on here and look around. What I am doing is similar to your approach but made dc-dc boost converter making a homemade battery charger that uses say 8-10 kw hr old school FLA group 27 12v and exports 375-380 VDC to i3 battery pack by splicing into main HV cable between inverter and pack.


Will be testing next week , but using the wisdom as REX wont turn on until 25 % discharge to prevent over-charging , And like in Nissan leaf homemade REX just need to turn on BMW first put in drive mode and then use my exported power when down to say 45 miles range left on my Guess-ometer. and then turn on my added contactors and put a blocking diode in with HV V+ on my Aux Rex, so if it goes into regen mode it wont back feed into my aux battery REX if regens while I'm Aux REX charging. Then when get to my destination , simply de-contact my aux battery REX before the i3 is turned off and main battery in EV de-contacts

So I am going to charge and drive hopefully like it should of always been.

I made a park and charge aux battery kit that fits in my 2013 Ford Focus Electric. Here is a local news story I live in rural Iowa and its 40 miles to any Level 2 Public Charging and only about 8 in a city of 200,000. http://www.cbs2iowa.com/news/local/recharging-your-ride Hope this is ok, about the homemade invention at that time not an offer for sale.


I am sure its going to throw a million codes and then I have a freind that can do a work around. I am sure though there are built in BMS safe guards that will shut down EV to prevent overcharging.


James
 
Bob

Also forgot to mention in my prior post , I was thinking of using refurbished Tesla pack, since I thought 1 module was 90 x 4.2 v cells in series giving me 369 VDC but that's not the case

That's why I thought it was easier to try Lithium to Lithium when I first posted. So when I found out the modules were 5kw hours but put in series think its like 45-48 volt per module , that I could not put in parallel.

But actually was as easy to get the right voltages and created 2-3 current limiting levels at a steady CV charging.

My motto is, "Defund Oil -- Defund Terrorism 1 EV @ A Time" If things like this that can limit range anxiety and get further down the road could be implemented in mass it could help be tipping point to mass EV adaption.

Thanks

James
 
Sounds like you're effectively increasing the BEV capacity by an alternate approach. The only caution is a ground-fault technique the Prius used but I've not seen documented in the BMW i3.

The Prius uses an AC signal, injected on the battery DC cables and the body. If the AC signal has an unexpected current to ground, the car declares a fault and the traction battery is isolated. I don't know if the BMW has a similar ground fault detection.

Sounds like you're already well on the way. If you assemble a set of photos before doing the smoke test, please share.

GOOD LUCK!
Bob Wilson
 
Bob

Yeah we have encountered the ground fault issues. We are stepping up the voltage up through dc/ac modified sine inverter then using another step up AC voltage transformer to around 440 VAC and then rectify that AC and after smoothing cap and such so we are isolated the lv 12vdc in my mobile battery charger just like the lv BMW 12 volt battery share common ground.

But my HV DC when contractors are closed shares BMW HV System already isolated ground.

The modified sine power is junk for precision things and some motors because of heat loss of the almost square wave with a little ripple. higher efficiency by about 17-18% over pure sine wave.

We are working hard over the holiday all day today Sunday through Thursday . Will hopefully have someone film me doing all the things so we can get better on next one.

Who knows if really get it squared away by a team of experts, who once they see it what my crude thing is could take it light years forward, "compared to me"

Maybe the LAPD could be our first real customer see this link: https://electrek.co/2018/01/18/lapd-bmw-i3-fleet-barely-used

TTYL

James
 
Back
Top