Just because you are flowing more amps doesn't mean you are more likely to have a fire as you will increase the wire cross section to accommodate for the increase in amperage in fact the higher voltage means a drop of water or piece of conductive debris is more likely to start a problem with the higher voltages.jonescg said:
214HP e-motorcycle built from scratch!
- Thread starter team_EMUS
- Start date
liveforphysics
100 TW
zombiess said:
I'm not advocating the CREE part or the pointless use of HV for power.
zombiess said:
Rather than looking up things about RMS, I have direct experience using them and tuning them, as my DP1e uses an RMS controller. I'm not going to say it hasn't generally functioned adequately, but It's definitely one of the things I won't miss when I upgrade the car.
zombiess said:
To create a sensible future of EV's rather than live in the legacy of old industrial motor drive tech.
jonescg said:
From all analysis perspectives, copper cross section is radically more reliable than HV BMS related BS. How many times have you been riding and said, ohh, my appropriately sized and properly terminated busbar/wire failed? Copper cross section isn't a moving part, isn't vulnerable to code glitches, doesn't involve hundreds of sketchy fire-hazard balance taps, has an inconsequential weight for a a motorcycle, and is many orders of magnitude lower cost as well as radically improved life safety.
zombiess
10 MW
You guys are totally missing the point. They already own the stuff and it's proven to work (except the batteries). They are not looking for suggestions on how to change it, but it sounds like they still need batteries and will hopefully get better ones as doubt their current choice will sustain the continuous amp load, but I don't know as I haven't seen them pushed that hard.
LFP you have good points and I don't disagree that lower voltage, high amps can be done, but they are not starting from scratch and I have not seen a low voltage high amp controller/motor that is commercially available. I haven't really looked hard either since I'm not in the market. Right now mid/high voltage appears to be the answer. As you already know, one has to carefully contain the flux fields at high power as stray inductance will eat it up and wreak havoc on nearby electronics.
flathill, Tesla went with what was cost effective for mass production, not what was best from an engineering stand point. It's a trade off that must be made when bringing a product to market. That picture is from a Tesla 200kW Roadster, not sure what is in the Model S but I'd be interested to see. I also doubt those are SiC devices hybrid or all.
Don't forget that I'm a big fan of paralleling devices myself (I have a layout with 12 TO-247s in parallel designed), but if I had a limited time frame I'd forgo the DIY method and look at off the shelf solutions. What this team is doing is already a large enough challenge without adding additional engineering work load. Power Electronics and custom motor design is far from an easy engineering task.
It would be better to support the team in what they are doing with the parts they have vs playing arm chair quarter back on what should be ideal or better from where you sit. This thread is going sour, let's turn it around
They are dealing with HV, let's provide suggestions to make it safer and better with what they got.
I would suggest keeping the battery leads as short and as close as possible. Twisting them would be ideal as the goal is to keep the loop area as small as possible. +200A at 600V is going to produce a big EMF field which needs to be managed in the cramped space of a motorcycle. Running a separate battery for auxiliary electronics might be a good idea to minimize noise on their power feed.
LFP you have good points and I don't disagree that lower voltage, high amps can be done, but they are not starting from scratch and I have not seen a low voltage high amp controller/motor that is commercially available. I haven't really looked hard either since I'm not in the market. Right now mid/high voltage appears to be the answer. As you already know, one has to carefully contain the flux fields at high power as stray inductance will eat it up and wreak havoc on nearby electronics.
flathill, Tesla went with what was cost effective for mass production, not what was best from an engineering stand point. It's a trade off that must be made when bringing a product to market. That picture is from a Tesla 200kW Roadster, not sure what is in the Model S but I'd be interested to see. I also doubt those are SiC devices hybrid or all.
Don't forget that I'm a big fan of paralleling devices myself (I have a layout with 12 TO-247s in parallel designed), but if I had a limited time frame I'd forgo the DIY method and look at off the shelf solutions. What this team is doing is already a large enough challenge without adding additional engineering work load. Power Electronics and custom motor design is far from an easy engineering task.
It would be better to support the team in what they are doing with the parts they have vs playing arm chair quarter back on what should be ideal or better from where you sit. This thread is going sour, let's turn it around
They are dealing with HV, let's provide suggestions to make it safer and better with what they got.I would suggest keeping the battery leads as short and as close as possible. Twisting them would be ideal as the goal is to keep the loop area as small as possible. +200A at 600V is going to produce a big EMF field which needs to be managed in the cramped space of a motorcycle. Running a separate battery for auxiliary electronics might be a good idea to minimize noise on their power feed.
Off topic a bit but even modules are using parallel semiconductors inside So you get your choice do you parallel them in your design using To220 To247, To 264 or what ever you chose or do you pay for another engineer to design something with them parallel inside in a package that might not work well for your needs....
Only reason I have the big modules I have is I got them cheep. Its not the best way but it is the easiest way to make something reliable.
Only reason I have the big modules I have is I got them cheep. Its not the best way but it is the easiest way to make something reliable.
Yup don't mean to go offtopic.
I actually suggested they keep the same system voltage but use higher quality cells. It is shocking how poorly the solder job is on these turnigy packs. Open the packs up and see for yourself! Anything but hobby grade packs in a lethal voltage pack! Hobby grade cells fine, but not prebuilt hobby grade packs. Disaster waiting to happen. They think the bullet connectors make the pack better which is a total joke.
I quote: The fact that it also come prepackaged with bullet connectors was a great advantage regarding the assembly process and packaging.
Jonescg uses hobby grade prismatics but he was smart enough to build his own pack.
I actually suggested they keep the same system voltage but use higher quality cells. It is shocking how poorly the solder job is on these turnigy packs. Open the packs up and see for yourself! Anything but hobby grade packs in a lethal voltage pack! Hobby grade cells fine, but not prebuilt hobby grade packs. Disaster waiting to happen. They think the bullet connectors make the pack better which is a total joke.
I quote: The fact that it also come prepackaged with bullet connectors was a great advantage regarding the assembly process and packaging.
Jonescg uses hobby grade prismatics but he was smart enough to build his own pack.
zombiess
10 MW
flathill said:
Yeah, I wouldn't want to be sitting on top of 24kW/Hr of those cells. I know what you means about the soldering job on them. Some of them can be sketchy and that's asking for problems.
Seems like we keep coming back to the same issue over and over, batteries are an expensive problem...
If it was me using those packs I would be unwrapping all of them performing a visible inspection, testing each cell, then re packaging them under compression. Long tedious job, but with those quality cells one has to be very careful. I had a Nanotech pack vent on me while it was in storage at 40% charge, the pouch swelled up and split, no fire, no smoke, just a dead cell and ballooned pack. Don't know what would have happened if the pack would have been fully charged.
jonescg
100 MW
liveforphysics said:
I'll let you know when I have a problem. The 168s battery management system is performing very well for us. I definitely don't recommend using CAN for everything on the bike - way too many places for things to stop working and it creates a massive headache on the racetrack.
The copper isn't the bit that would suffer, rather the connection between it and something else (bolted, soldered, resistance or laser welded).
I agree that 700 is too many volts / complexity for a road going bike, but 360 V is perfectly manageable.
zombiess
10 MW
I spun off a new thread to discuss the topic of voltage and drive system power.
http://endless-sphere.com/forums/viewtopic.php?f=30&t=67455
http://endless-sphere.com/forums/viewtopic.php?f=30&t=67455
Luther Blissett
10 mW
- Joined
- Mar 5, 2015
- Messages
- 22
Just want to say gosh (wow in the US). :|
...Best of luck guys!
...Best of luck guys!
team_EMUS
10 mW
- Joined
- Feb 27, 2015
- Messages
- 22
liveforphysics said:
Our scope of work was mainly Mechanical and Systems engineering. The core goal of the project was to design a MOTORCYCLE, not its components. Thanks for the hint though. We might design some power electronics in the future, but so far, we are going to stick to the Rinehart PM150DZ drive. The cost of going back on that decision is too high.
team_EMUS
10 mW
- Joined
- Feb 27, 2015
- Messages
- 22
liveforphysics said:
This is not a production bike. It is an insanely fast and nimble electric motorcycle. We have no plans of selling it.
liveforphysics said:
As you might know, electrical power equals to:
P=V*I
And Joules losses are given by:
Pj=R*I^2
Hence, reducing the current to reduce losses and increase the voltage to maintain power is only common sense.
Since most of the losses occur in the battery pack, due to internal resistance of the cells, it is easier to design a high voltage battery pack that respects the maximum C-rate of the cells and fits inside a motorcycle.
I do agree however that there is no perfect solution to this problem, but rather many correct solutions. The Best configuration depends on the designer and the design constraints.
Safety is at the core of the design, with our modular battery pack, the multiple safety circuit break points and other apparatus. Running at a high voltage is not an issue when dealt with professionnally and carefully. The rider will not be able to access the high voltage component during races. Motorcycle cannot be operates without the tank.
There is no going back on the motor and drive voltages as the Enstroj EMRAX 268 and Rinehart PM150DZ are such a perfect match (and already bought)
Coronna effects are to be considered in industrial applications as bearing wear due to HV will become an issue. However, for a hobby project or a racing vehicle, these will not be a problem, as the mean time between maintenance will likely be 1 race (30 minutes). If ever that problem showup, ceramic bearings can be easily swapped.
liveforphysics said:
Your concerns have been heard. We are now looking at 18650 cells for our next design. These cells are common and standardized, meaning that no matter what happens with one supplier, we can turn arround and use something else without changing the design. Power and energy density, once packed into a module is similar to th HK solution. http://endless-sphere.com/forums/viewtopic.php?f=14&t=27707&start=100#p1040216
team_EMUS
10 mW
- Joined
- Feb 27, 2015
- Messages
- 22
zombiess said:
You are totally right, we have made some choices we have to live with from now on... Luckily for us these are not so bad !
The key issue right now is batteries.
We are indeed moving towards better cells.
That is equally true for the AC asynchronous motor and the Panasonic cells wich are not that great performance wise, but are great engineering tradeoffs for a production vehicle.zombiess said:
zombiess said:It would be better to support the team in what they are doing with the parts they have vs playing arm chair quarter back on what should be ideal or better from where you sit. This thread is going sour, let's turn it around
We are not looking for confrontation here... We only wanted to share about what we did...
zombiess said:
Indeed, EMI is a problem right now. It's black magic!
team_EMUS
10 mW
- Joined
- Feb 27, 2015
- Messages
- 22
zombiess said:
Hell yeah!
zombiess said:
Indeed!
zombiess said:
Thanks!
zombiess said:
We have about 3kWh right now, wich is enough to do the tests and debug the EMI but not enough to actally ride it in a race... But like I said, we are looking at a new design using cheaper, better, more available cells.
The HK configuration is 162S3P
The stock wiring is big enough. We are aware of the flaws of the HK... I think we got the message
We did study the cooling using air passing through the slots at the front and back of the pack. Even without cooling (adiabatic) the batteries won't heat up to their critical temperature during a race. For our next design we are looking at phase chage material (PCM) cooling. 8)
team_EMUS
10 mW
- Joined
- Feb 27, 2015
- Messages
- 22
We have cosidered that issue and that is why we asked Doctorbass for help on cell matching ant testing. Anyways now, we are moving toward 18650s,jonescg said:
Great folks! Great products!jonescg said:
The electrical design is 100% safe and has more ways to make the bike not work than it has to make it run! On the other hand, no one is going to get injusred from running the bike!jonescg said:
jonescg said:
Our motor draws 240Am peak. We are looking at the same sort of top speed.
I would say that the more amps the more heat, so the more chances to catch fire.
jonescg said:Nice work guys, and I look forward to seeing Voltron racing against you in 2016
Thanks! Hopefully we are going to be ready by then!
liveforphysics
100 TW
team_EMUS said:
There is nothing to be scared of in drawing 4-digit currents. It's the same path to success with any current levels, size the conductor right, and cold-forge crimp the terminals on, done.
The number of time's I've had a power cabling failure is zero, including bikes drawing >1,300A. In the datacenter industry I used to work in, we had hundreds of connections pulling >5,000A continously 24-7. Never saw a failure on any joint in the 5 years I worked around them, and they never get any time off of drawing >5,000A.
The motivations to go HV are simply entirely lost the moment you package the battery, motor, and controller in some system less than 10-20ft apart from each other. It becomes all loss at that point, just swallowing the Vf drop on the IGBT's is more inefficiency than a well designed low voltage system's entire conduction losses.
Glad to hear you dropped the hobby pack idea, that should save your build a lot of headaches and possibly lives.
team_EMUS
10 mW
- Joined
- Feb 27, 2015
- Messages
- 22
team_EMUS said:
It's just a general principle... Of course the copper conductors are not going to catch fire if dimensionned porperly... The cells are the problem. If you respect the current limit of your cells and cool your battery properly, it shouldn't catch fire. However to avoid thermal runnaway a proprer BMS with temperature monitoring must be used.
The reason why we chose to go at higher voltage has not much to do with the size of the wires. It came down to the choice of motor, drive and cells that were commercially available. We tried to match them to get the highest power rating. Thats all.
For more questions abour power-electronics and motordrives,please refer to Rinehart Motion Systems. We had nothing to do with the internal design of the drive.
