Hello! This is my first time posting to any of the forums because I guess I suck. I would like to experiment with a 400V DC circuit and I want to find out how I can do this here safely, without crossing any rules or guidelines and without frightening anybody. Aside from basic safety, I figure I can also either cordon off the area when I’m working on or put up some “high voltage” signs or something?
This will be the first phase in developing an adjunct battery system for Nissan Leaf cars. I’m going to start with a battery array with an impractically low capacity for sheer simplicity and price – so far I’ve found a 3.2V 8Ah pouch cell I can assemble into a 3kWh array for about $600. This will be plenty for developing and testing the battery controller with bidirectional charger/discharger circuit, which I’ll be home-rolling. (This unit will only discharge through a current-limited, bucking converter.)
My testing for phase 1 will only involve charging and discharging with a dummy load and making sure that current both ways matches what the controller computer is dictating.
When i worked on high voltage. I can only the common stuff i pretty sure you know.
1: We used the one hand rule. Never touch live units with twos hands. Arking through heart will most like damage or kill you. Clip common ground lead in before you power up.
2: Never use ground straps in test. Arching grounding straps will hurt you
3: Volt test large capacitor after electrical testing unit under test. Discharge charge capacitor with any high voltage and large capacitance through on mega ohm 1 watt.
wear some thing over you face. pcb can explode if a lot of power burns out the traces.
Thanks. After I posted, I realized that the first step is writing out and implementing a safety plan, including all of your items above.
Yes, these are lithium batteries. I suppose I should include part of my include thermal management in phase one with at least the temperature sensors and a static (always on) fan for now. The controller will eventually run a liquid thermal management system.
My intention of posting here is also the “see oz or art” part.
A hazmat fire in the building will cause all kinds of grief. You might want to consider a friendlier lithium chemistry that will tolerate more abuse as you and your project get up the learning curve. Maybe a lead gel cell for your first set of “proof of concept” prototypes. The ELab does not have a class D fire extinguisher or bucket of sand. I’d recommend any load / torture testing be done out in the back parking lot - just in case things go badly. BTW what do you plan on using as a dummy load??
Do you have any design / build experience with high voltage? I think TI and others have app notes / guidelines on HV techniques. Over rate your components and insulation. You don’t need to cordon off the area your working in. Most members are respectful of leaving other members projects alone. Darwin may take care of those who don’t. Consider some kind of emergency disconnect switch - as in big red button…
Do not leave project on / live unattended.
IMPORTANT!! - Keep mind that some of our bench test gear will not tolerate a 400V signal. Blowing out channels / damaging test gear as you get up your learning curve is not acceptable.
When in doubt - ASK!
Do you have any switching pwr sply design / build experience? Consider looking for / buying an off the shelf DC - DC bidirectional switching pwr sply to study / play with. The DMS knowledge base has members who have hands on design / build experience with switchers. Network the membership and see if anyone’s interested.
I sense you have a long learning curve ahead of you. There is a plethora of app notes, tech papers, online tools, etc out there. Break your project down into small functional blocks and build upon each block. Lab notes, picts, screen shots - document as you go. I use this approach on my current project and it has worked extremely well.
Thank you for your thorough and thoughtful response!
I don’t think that limitation would cause any problems for the very initial stages, that’s probably a good idea.
I don’t think I would prefer to pursue this without acquiring appropriate safety gear, including a class D fire extinguisher (presuming they aren’t stupid expensive). And I’ll definitely get some sand.
This is indeed my first high voltage application! I’m pretty excited about it. Also why I might go “a little overboard” on safety. I’ve worked a little bit with switching power supplies in the past and designed an automotive transient protection circuit, but the vast majority of my EE is with digital circuits. My day job is software and firmware engineering, but I’m being forced to .net right now…
I’m still in the early stages and have been reading a few white papers. Mostly I’ve been gawking over the stats on these newest SiC MOSFETs, like the Infineon IPDQ60R007CM8 series 600V, 288A, with Rds(on) = 7mΩ and Vfsd = 900mV! Of course, you gotta keep them cool or those stats break down.
I’ll be doing all of my work in KiCad so I’ll post my work periodically for review. I also like to annotate my schematics with lots of notes about what each little group of components is for.
I’m looking at 600V components for a circuit that will be 420-ishV max. Think that’s a fair enough margin? As for amperage, I’ve got to have temperature sensors on those MOSFETs so I can keep current limited to some margin below their safe range. The one I’m looking at is 288A at 25°C, but that drops to 182A at 100°C, though in practice, I’ll have them cooled using the car’s cooling system.
Hrm, maybe that would be a good requirement for my phase 1. I don’t only want this for development, but I want the software to have a “panic button” and perhaps a physical panic button all throughout the development and testing cycle would be a good idea.
iiuc, you can’t just open a relay under a heavy load w/o fireworks, so that’s another thing I’ll examine. I believe that the Leaf has a power cable somewhere that can literally be detonated by the VCM.e
JESUS CHRIST! LOL! Darwin indeed!!
I think I would rather design the full thing and test it in SPICE first – that’s how I tested my transient protection circuit. The SiC MOSFETs are pricey ($25 each), though they thankfully have TVS diodes built into the gate-source and are supposed to be HMB class 2 rated. So considering component price, there could be some benefit to prototyping using cheaper parts first – lower voltage/amperage and hotter Rds(on) and Vfsd.
But I sure do have a lot to learn yet about power circuits. I’m not strong in RF type things and I’ve never used these types of MOSFETs before – they have a 4th “driver source” or “KS” pin that I need to understand. I’ll be doing a lot more reading for sure. One thing I learned from simulating my transient protection circuit is that all sorts of strange things happen in the very tiny time intervals of power circuits.
I also need to learn a lot more about the properties of these pouch cells and figure out how I want to do the physical part of the thermal management, how many temperature and voltage sensors I’ll want, etc. The unit will use a fairly complex liquid thermal management,so there’s a lot that will need to happen on the software end of things too. While I’m designing this specifically for the Leaf, it would be nice if it could be used on other vehicles as well.
Actually, I’m planning on implementing the thermal management system as an independent kit. There’s a company in the early stages of designing a replacement battery pack for the Leaf (40kWh NMC) and I’m trying to convince them to go liquid instead of air so that it will be compatible with my kit, or somebody else’s that’s similar.
For now, it’s enough to just have a simple radiator and fan on my adjunct battery system.
Re: Class D extinguisher. Depends on your definition of “stupid expensive”. I think we paid $400-500 each for the 2 we own. At least, I think we’ve got 2. Metal Shop definitely has one, and I thought that the fire marshall made us get one for Machine Shop too (he was hyperventilating about the situation).
That said, at that time I checked with my fire extinguisher expert, and she quoted around those prices for the standard D, and somewhere $800-900 for the one that’s designed to put out lithium. We didn’t go for the litium option…
Holy crap. I think I’ll call that one “expensive” and the $800-900 “stupid expensive.” I’m afraid that if I buy one I might not donate it … at least not initially…
I don’t know yet. Probably just a high current resistor with a big heat sink and fan in series with a fuse in case my circuit pulls more than intended.
My two cents from our work on the drift trike, have a plan for when you accidentally short something. It’s a bit scary waiting to see how hot they get (ask me how I know).
Here’s a cheap contactor option. I believe @michaelb has work experience with high voltage components and can maybe give you some advice if you drop by a motorsports meeting.
I think the bucket of sand is your best fire bet, as I understand it you can’t really extinguish them if they actually go, since they’ve got fuel + oxidizer. You’re just trying to deal with the heat.
As others have already observed, voltage is only part of the list of hazards.
One that I don’t see already mentioned is the energy hazard. Even it you don’t shock yourself or ignite the batteries, there is a LOT of power here, like melt a crowbar type power. This requires it’s own special care.
I think you are taking too big a bite. Work on the electronics first in the absence of the battery. Once that’s well behaved, then add the bomb battery pack. High current 400V dc supplies are not common, but they are available. Find a surplus unit. If I was doing this, I would not start with a battery stack, I’d start with a bench top power supply.
Deal with as few problems as possible with each bite.
And I’d avoid pouch cells. They require a lot of mechanical support, and unless you are qualified to provide that, avoid them. There’s a reason car manufacturers pretty much all use cylindricals and prismatics.
As someone who’s designed high voltage high power supplies, and works with high voltage DC as part of his job, my professional and personal opinion is “Don’t”.
I don’t want to gatekeep but you’re taking a non-insignificant risk to your life and DMS’s property for this one. If I’d not had multiple years of training under someone who had been doing it for decades then I’d have definitely ended up as an ink spot somewhere.
If you really want to continue, start working with 60V to 100V 1-2Ah where you have enough voltage and current to probably die instead of definitely die and build your setup with parts that can tolerate the higher voltages. Once you’ve got that working get your own place (not DMS) that has a sand vat and a fire suppression system and you can step up the voltages. You should be able to build a pretty cheap pack out of 18650’s and with 16-21s1p you will have plenty playground for balancing circuits and charge/discharge circuits to get your feet under you. Unfortunately this method won’t really get you safely working with high voltages but it can get you started on the engineering task journey at least.
Worth noting that contactors can’t be relied upon to disconnect batteries during an over current event (they weld shut and operate relatively slowly).
A common way to disconnect batteries under high current is a pyrotechnic disconnect (though this adds yet more complexity to your setup to provide current monitoring and control).
First of all, thank you to everyone for your input! I should perhaps mention that I’ve got to finish my “fix my apartment” projects before I can start on this one, so this won’t be happening very soon.
Thank you! These are $182 new on Mouser and have 200µΩ resistance on connectors, which is perfect! I’m planning on putting relays in parallel with the MOSFET I’m going to use to prevent unintentionally charging the main battery, but when it’s actually plugged in I don’t want to pay the efficiency penalty of it’s Rds(on), which is 7mΩ. So this is perfect! When a charger is plugged in, I can I’ll first activate the MOSFET and then the relay and disconnect in reverse. This has the added benefit of not doing bad things if my system accidentally tries to connect the battery when there’s already a charge voltage – which it should never do anyway.
I’m still in the early stages of this project and have more to learn, but from what I’ve been reading it seems like there are various stages of severity for Li+ battery fires and the initial ones can be extinguished by fairly mundane means. But once Li is actually burning it undergoes a thermal-runaway affect. Though it doesn’t have it’s own oxidizer, I presume that it can pull oxygen straight out of water or other materials. It seems that the extinguisher’s goal here is to reduce the heat below the threshold for that reaction to continue.
I’m probably going to look into training for working with Lithium battery fires.
Yes, PERFECT! I can do a great deal of the initial work with PSUs. Thanks. So no batteries for phase 1, just PSUs, dummy loads, fuses, fans, and lots of temperature and voltage sensors.
That’s probably sound advice, though I don’t want to use cylindricals in a liquid cooled system as it seems like a terrible waste of space to me. None the less, there’s no reason why the entire first version can’t use prismatic cells as long as measuring the external temperature can give me a descent idea of the cell’s core temperature when it’s also being cooled. I just need to know that the cell isn’t overheating when such a condition isn’t indicated from it’s surface temperature. I’ll want to learn a lot more about their physical and thermal characteristics.
But once I’m further into this project, if pouch cells seem to offer a large benefit, I can work on those details then. Thank you for your insights. I figured they would need physical support, but there’s surely a lot to know about using them.
I sure would love as much of your expertise as I can get! But I think the reasons that I would stop pursing would be along the lines of
somebody else already did it, so I don’t have to
something comes up in my life that makes it infeasible
insufficient funds
the project idea turns out to not provide as much benefit as it initially looked like or is otherwise infeasible (that’s happened a before!!)
I get rich and can pay somebody else to work it out.
Not to AT ALL dismiss the danger you are warning of, I take non-insignificant risk getting on these highways in Dallas! As this is my first high voltage venture, I intend to work out safety procedures and get the necessary safety equipment. I’m not one to rush into something haphazardly. I’m also not going to rule out a slight career change at some point, from firmware for embedded communications systems to EV systems.
But I like the idea of just using PSUs to develop the initial circuitry, which can also be done at lower voltages until I’m ready to torture test it.
I do hope I can get you to look at my schematics later on and get feedback from you.
I’m planning on using an ordinary EV automotive fuse for over-current protection. iiuc, pyrotechnic disconnect is for when your car’s computer senses a crash or other “very not OK” condition and you want to disconnect the battery from the rest of the system very quickly and without fail, even if it’s under load.
Whoopsie, you’re correct @Bizwacky! NMC batteries use a combination of LiCoO₂, LiMnO₂, and LiNiO₂ in layers for the cathode while the LFP has the oxygen right there in the name: LiFePO₄. So I presume that, much like thermite, after an NMC battery fire in a vacuum (yes, looks like it can burn just fine in a goddamn vacuum) we’ll have Co, Mn and Ni in their reduced forms with Li in an oxidized state. I’ll definitely need to learn the full chemistry here, it’s interesting!
Cathode chemistries vary, but some of them contain oxygen as well, like Lithium titanate (Li₄Ti₅O₁₂) which has a lot of oxygen, though I presume it’s harder to liberate that from Ti than other elements.
Your basic class D fire bottle is sodium chloride - ie salt. And yes it is used differently. You spray down from the top of the fire. The idea is that the salt melts on the way down and encases whats burning.
I would like to experiment with a 400V DC circuit and I want to find out how I can do this here safely, without crossing any rules or guidelines and without frightening anybody. Aside from basic safety, I figure I can also either cordon off the area when I’m working on or put up some “high voltage” signs or something?
But I think the reasons that I would stop pursing would be along the lines of