DIY Electric Car Forums banner
41 - 60 of 126 Posts

·
Registered
Joined
·
141 Posts
Discussion Starter · #41 ·
I'm looking for a programmable can bridge, is yours Arduino based/programmable/available for sale?
This device is very specifically to allow communication between the LG batteries I'm using, and the CAN bus (for master devices like SimpBMS). It may be available for people who have the same batteries.
I'm looking for a programmable can bridge, is yours Arduino based/programmable/available for sale?

Also, not sure if you are aware but the openinverter board is quite limited in its protections, it is therefore very easy to destroy the power stage with incorrect settings
With that said, what kind of mistakes are likely to cause damage? I'm familiar with inverter programming, but since I didn't write openinverter I need to be mindful of what might cause damage. I'd hope iacmax would be able to prevent any damage by limiting the max current through the inverter and motor.
 

·
Registered
Joined
·
221 Posts
what kind of mistakes are likely to cause damage? I'm familiar with inverter programming, but since I didn't write openinverter I need to be mindful of what might cause damage. I'd hope iacmax would be able to prevent any damage by limiting the max current through the inverter and motor.
There is no hardware interlock, i.e. it is entirely reliant on the program sending out the correct on/off sequence to the bridge to prevent it blowing up. If at any point an upper and its lower are activated together, even for a microsecond, infinite current tries to flow from + to - via the pair. Boom. From the wiki, this is just one of a wide variety of things that could end up with bridge destruction. I'd even go as far as saying plan on it blowing up several times before you get it dialled in completely, particularly as you get more adventurous. At least do all your testing on low voltage and then start your HV testing on capacitors that don't have sufficient energy to blow the bridge.
Rectangle Azure Organism Font Screenshot


Industry standard is to have interlocks so that any issue with the code cannot propagate to the bridge
 

·
Registered
Joined
·
124 Posts
Also, not sure if you are aware but the openinverter board is quite limited in its protections, it is therefore very easy to destroy the power stage with incorrect settings
This worries me and is the main thing stopping me from using the openiverter board. Its not like you can just buy a new power stage, So if I killed the inverter, That's another $6500 straight up for a whole new motor. Unfortunately the alternatives seem to be very limited. Either its a control unit that the vendor will only sell complete with a motor, Or its an expensive unit on its own. The EVControls T2C unit seems to be the pick of them all, It interfaces directly with the Tesla control board and is a simple plug and play setup, but its quite expensive.
Does the stock Tesla control board have interlocks to protect the IGBT's? Surely it would.

Catphish, In regards to the openinverter board, They mention that there is some assembly required. What is it? Just soldering some plugs on? I guess if its that simple then once its installed its just a matter of following the same parameters as another 'known' running setup. I've seen some people post their parameters on the openinverter forum.
 

·
Registered
Joined
·
141 Posts
Discussion Starter · #44 ·
Catphish, In regards to the openinverter board, They mention that there is some assembly required. What is it? Just soldering some plugs on? I guess if its that simple then once its installed its just a matter of following the same parameters as another 'known' running setup. I've seen some people post their parameters on the openinverter forum.
The assembly is quite challenging for the small drive unit because some sensors need to be de-soldered from the board as it is removed. Watch this video for details:
It is very much my hope that copying the known good settings from https://openinverter.org/parameters/view.html?id=15 will give me a working (and not exploding) drive!
 

·
Registered
Joined
·
221 Posts
So if I killed the inverter, That's another $6500 straight up for a whole new motor. Unfortunately the alternatives seem to be very limited.
That is entirely the reason I rejected the openinverter board. It is a pretty amazing bit of kit in that it is done mostly by two guys in their spare time, however HVDC is merciless and it will find the weaknesses, it is only a matter of time. Even with interlocks, there are still plenty of other ways to kill the bridge

It is entirely possible to run successfully with this hardware, you just have to be very, very clinical in your approach. Any change you make could end in boom. Scope any signal wires for noise that could send incorrect information, sneak up on any changes with low voltage and then high voltage capacitors first before battery. Ground, shield, isolate everything that could put noise on a signal line. Fit a fuse that is a lower rating than full power for initial battery testing- it won't protect the bridge but it will limit the damage inside, bridge failure can be quite violent. Have a plan on how to replace the bridge so you aren't stuck when it happens.

The other thing is that if you are running a single accel pedal sensor then it only takes a glitch in the signal or a break in the 0V line and you have lost control of the rig. Redundant accel input is mandatory here, for a very good reason.

This gap in the market will be filled at some point by commercial operators. AEM are just now releasing the LDU control board and VCU, presumably at some point they will release an SDU one. Yes it is expensive but it is cheaper in the long run if it significantly reduces the chances of destroying the power stage.

I have spent several years developing/testing/blowing up inverters in a commercial application, the sound is burned into my brain
 

·
Registered
Joined
·
141 Posts
Discussion Starter · #47 · (Edited)
I'm starting to look more seriously at buying parts, and have a couple of questions about this charger.
1) Do you have the documentation, particularly the wiring diagrams and CAN command set? It doesn't seem to be published.
2) How much hassle is the water cooling? I'm thinking I likely already need to run 2 water cooling loops, one for batteries, and one for motor. I'll also have to think about running the cooling loop during charging, though this doesn't seem insurmountable.

This alternative part was also recommended to me, I'm somewhat unsure how to choose between these Chinese devices: Dilong New Energy Technology Official Website EV OBC DCDC On-board Charger Power Supply Manufacturer
Interestingly, they also do an air cooled version, but I'm unclear on the pros and cons of air cooling, apart from seemingly simpler installation: Dilong New Energy Technology Official Website EV OBC DCDC On-board Charger Power Supply Manufacturer

I assume (since they're rare) that 6.6kW probably isn't a good match for air cooling. I could also go with a 3.3kW charger instead. Less impressive, but potentially simpler, and still capable of fully charging my (approx 27kwh) pack over night.

Any advice welcome!
 

·
Registered
Joined
·
124 Posts
I'm starting to look more seriously at buying parts, and have a couple of questions about this charger.
1) Do you have the documentation, particularly the wiring diagrams and CAN command set? It doesn't seem to be published.
No, It didn't come with much documentation at all, but im not at that stage in my build where ive even installed it yet. Maggie and Frank at Ovartech are very prompt and helpful in their responses, I have no doubt that when I request the documentation on the CAN messaging, it'll be no issues. The wiring is pretty straight forwards. It's listed along with a whole bunch of specs in the datasheet, available on their website. It comes pre configured with either 250k or 500k baud rate, you need to know which one your using before you order it. The Orion BMS I was planning on using is 500k, as is the Tesla motor to my understanding, It seems simple enough to just make any other CAN parts match those critical components.

2) How much hassle is the water cooling? I'm thinking I likely already need to run 2 water cooling loops, one for batteries, and one for motor. I'll also have to think about running the cooling loop during charging, though this doesn't seem insurmountable.
It's not a high heat load item. Yes, it needs to be water cooled but my understanding is that the peak heat load will be during charging, so its simple enough just to add the inverter loop inline with the battery loop, which would ideally be running circulation when charging anyways. Setting it up to do that is easy enough, you just used a switched output to activate the cooling loop when the system is registering a charging state, Basically the same switched output to activate the OBC when its connected to the wall supply. Id say that if you have the capacity to use a liquid cooled unit, you should, especially if your in a hot part of the world. That's between you and your build though.

The only thing I didn't like about the unit is that the DC-DC output is not adjustable, Its fixed at 14.4v. Its not a big thing at all and for user simplicity this is best and not something ive ever really seen as an adjustable setting in any other model, But if I was able to adjust that output up to around 18v, I could have replaced my 12v battery with a Li-ion pack. Pretty much everything electrically on the 12 side is voltage regulated at each unit, and all the lights are LED so it would have been nice to use a better battery, more compact, better tech, sits idle for longer without issue etc, I can probably still figure out a way to manage it though, but that's specific to my build, its by no means a show stopper.

6.6kw is essentially level 2 charging, Here in Aus, 240v is standard at the wall, so it makes no sense for me to use a smaller derated charger. It also gives me the ability to charge at a faster rate elsewhere away from home, where time may be more of a concern. For my situation, it was a no brainer. I would like to add DC fast charging ability to the build, but im not there yet and that'll be its own charging system anyways.
 

·
Registered
Joined
·
141 Posts
Discussion Starter · #49 ·
It's not a high heat load item. Yes, it needs to be water cooled but my understanding is that the peak heat load will be during charging, so its simple enough just to add the inverter loop inline with the battery loop, which would ideally be running circulation when charging anyways.

6.6kw is essentially level 2 charging, Here in Aus, 240v is standard at the wall, so it makes no sense for me to use a smaller derated charger. It also gives me the ability to charge at a faster rate elsewhere away from home, where time may be more of a concern. For my situation, it was a no brainer. I would like to add DC fast charging ability to the build, but im not there yet and that'll be its own charging system anyways.
Thanks, I think I'll give the water cooled 6.6kW a try. As you say, it makes sense to cool the batteries during charging anyway, so putting the charger in the battery cooling loop is no trouble at all.
Mains power in the UK is also 240V. A normal socket is 13A (3KW), but there are lots of 30A (officially 7KW) chargers around, and it's easy to install one at home too, so it's probably worth it.

I think there are 2 competing chargers I need to look at, the ovartech version, and a similar "dilong" version:

Ovartech seems to be more widely known, but I've had recommendations for both. I'll try to get documentation for both.
 

·
Registered
Joined
·
13 Posts
I don't have workshop space to begin taking the VX220 apart yet, but started work on the Tesla SDU today, installing openinverter board...
Do you know of a step by step on disassembly of the SDU inverter side like you have just done? Youtube or blog?

I'm sure I could muddle through but would be nice to know where I am going.
 

·
Registered
Joined
·
141 Posts
Discussion Starter · #51 ·
Do you know of a step by step on disassembly of the SDU inverter side like you have just done? Youtube or blog?

I'm sure I could muddle through but would be nice to know where I am going.
Basically, it's very easy apart from one part (desoldering the current sensors) which is quite difficult and delicate. I'd suggest you don't attempt to remove the board without watching and understanding this video:


The only other thing to mention is that you need to remove the bolts from the motor cables (though a small plastic access panel) before removing the inverter itself.
 

·
Registered
Joined
·
13 Posts
Yeah, I've watched that video a few times and even managed to stay awake once. But I was more interested in the disassembly of the motor/inverter assembly (such as disconnecting 3 phase and whatever that blue wire is).
 

·
Registered
Joined
·
141 Posts
Discussion Starter · #53 ·
I don't have a video, but there are 4 steps:
1) Remove the small orange plastic cover that probably says VOID all over it.
2) Deep behind that cover are 3 bolts. Reach in with a long enough tool, unscrew and remove all 3 bolts.
3) Unscrew the 10 bolts that hold the motor on. Support the inverter when you remove the final bolt. It will probably take a couple of taps with a hammer to actually come away, but you don't want it to fall and damage the blue wire.
4) As you pull the inverter away, you will see the blue wire is quite tight. There are 2 connectors to unplug. You ideally need 3 or 4 hands to hold the inverter body while unclipping the cables.

Once you've done that, you will have the detached inverter as below (plus a bracket that holds the motor wires).

5) Remove the plastic bracket that holds the motor wires. This is fiddly but you'll figure it out.
6) Unclip the remaning wires.
7) Unscrew the metal tray from the inverter, and remove the tray and board together. Do not unscrew the board from the tray until you are sure you know what you're doing.

Sorry I don't have more photos or video.

Passive circuit component Circuit component Hardware programmer Electrical wiring Electronic component
 

·
Registered
Joined
·
141 Posts
Discussion Starter · #54 ·
Bonus photo showing the motor cable bracket. The following explanation may or may not make sense, but I'll try: To slide the cables out of the bracket, you need to slide them further in, so that the insulation is between the clip. This is easy if you unscrew the bracket and rotate it so that its screw mounts are out to the left.
Circuit component Passive circuit component Hardware programmer Electrical wiring Computer hardware
 

·
Registered
Joined
·
141 Posts
Discussion Starter · #56 ·
Small update: I now have basic BMS functionality ((cell voltage reporting and balancing) working on my LG cells using their built-in slave modules, yay.

I am now trying to decide between purchasing an Ovartech 6.6kW charger, or attempting to hack an OEM tesla charger. The Ovartech option is certainly easier!

Still waiting for workshop space before beginning serious mechanical work.
 

·
Registered
Joined
·
141 Posts
Discussion Starter · #57 ·
This week I've bought a 10kw Tesla on board charger. I'm waiting for an open source logic board to get this up and running. I'll hopefully get a Tesla DC-DC converter to complete my set!
Circuit component Passive circuit component Electronic instrument Hardware programmer Electronic engineering

Rectangle Font Gas Electrical supply Screw
 

·
Registered
Joined
·
141 Posts
Discussion Starter · #58 · (Edited)
I have now received 18 x 1.5kWh 16S LG battery modules, which I intend to run in a 6S 3P configuration, if space allows. These came with a big stack of 250A and 150A relays and fuses, and a couple of smaller 40A fuses from the original battery pack.

I have also tested the Tesla charger, which is faulty (I got it for next to nothing). 2 of the 3 modules work so it can achieve my desired 6.6kW, but I've decided it's too bulky for my build. Instead I've ordered the Ovartech 6.6kW charger and DC-DC converter.

I have also tested my new BMS which is able to automatically and accurately balance strings of these LG modules.

So I now have the following ready to go:
  • Vauxhall VX220 - currently unmolested
  • Tesla Small Drive Unit - Openinverter board installed and tested, HV cable
  • Tesla driveshafts
  • 18 x LG 16S battery modules
  • 3 x DIY BMS controller for LG modules
  • Ovartech 6.6kW charger + 1.5kW DC-DC converter
  • 150A and 250A relays for general use
  • 175A, 225A, and 40A fuses for general use
I anticipate still requiring the following:
  • Type 2 charge socket
  • Main contactor
  • 2 x radiators + 2 x coolant pumps
  • Lots of cable, connectors and custom fabricated parts
  • Current sensor / charge counter

I hope to have my VX220 in the workshop by newyear!

Batteries and charger testing:


Relays and fuses from a mystery OEM battery pack:


GEN2 charger openinverter board ready to go:


My custom BMS for my LG battery strings, balancing and CAN bus monitoring
 

·
Registered
Joined
·
141 Posts
Discussion Starter · #59 ·
Small update. I've now developed a charge controller to connect to the EVSE and Ovartech charger.

I've decided that I will use the 3 x 175A fuses independently on each of my 3 parallel battery strings, and the 250A relays will be sufficient as main contactors, as they can handle 500A for longer periods than the Tesla SDU can.

I've had word that my workshop space is available, so the real work can begin soon!

I've ordered a 100A CAN isa current sensor, so the only major parts still to source are the charging socket, heating + cooling, and everything I've forgotten. I hope to follow up with a complete wiring diagram, and in the new year, photos of the beginnings of disassembling the VX220.

Passive circuit component Circuit component Hardware programmer Microcontroller Audio equipment
 
41 - 60 of 126 Posts
Top