high power dc-dc isolated bidirectional

[madnight]

now befor anyone says anything yes i know the risks about high voltage
hi all well this is a big one here i being looking round and struggleing to find any good info on how to go about doing this
basicaly i building a ev and want to take advantage of dc fast chargeing with out the need for a larger battery
so i was thinking could i use a high power bidirectional  (oviusly isolated)
but the issues i am haveing is finding info on building one
be able to support up to 100a out put
easily intagratable with a micocontroler
i would like side A (high voltage/charger side) to be able to support a minium of 270v dc
i would like side B (lower voltage but still high/battery side) to be able to support up to 135v dc
i would like the able to alter aide A voltage in coraspondence with side B (so as batter chargeing it shows to charger a higher over all voltage)

any questions just ask

[Zero999]

The basic topology is just a transformer with transistors both on the primary and secondary, which can act as an inverter or synchronous rectifier seamlessly.

[Slh]

I'm sure you know the dangers but you give the vibes of someone who's trying to run before they can walk...

But anyway, some thoughts:

1) why does it need to be isolated? Both sides are dangerous and the grid is isolated by the DC charger. You could just use an interleaved buck converter.

2) 135V * 100A = 13.5kW. Do you have a couple of power supplies for developing this? New you're probably looking around £20k for the pair of new.

3) the big problem is going to be charging standards. I watched people trying to implement chadamo and CCS - not fun.

4) at this level most people use digital control and DSPs. They tend to be a bit confusing to work with.

[madnight]

thank you for the infor
i have done suplys befor just nothing as powafull as this
and what would you recomend as turn ratio on the transformer

[madnight]

thank you slh and i just head isolated would be better for something like this
and as far as the dc fast charge protocol gose i useing a off the shelf pre made unit (thunderstruck MCU)
and i have some powersupplys a freind gave me from a closed down busness (up to 170vdc at 50a each)

[Slh]

i wouldn't bother with isolating it. There's no real advantages and no regulatory requirements for it. You've got to follow UNECE reg 100 as the battery is more than 60V so it all needs to be isolated from the vehicle and that's the bit that will keep the user safe.

Nice find with the Thunderstruck charge controller. From what I understand of CCS it sounds massively complicated and far too involved.

This might be a good application for an IPT (inter phase transformer) based buck converter.  The university of Manchester have a bunch of papers on them if you want to read more. It's basically two half bridges at 180 degrees feeding both ends of an equal winding autotransformer.. the middle point goes to the LC filter of a buck converter.
 If you keep the DC charger voltage at twice the battery voltage then there's essentially no ripple in the inductor or capacitor which could be good for losses. Basic control is the same as a two phase interleaved buck converter.

Regarding the power supplies, don't forget that you need a 13kW load as well. It can just be a big bank of resistors but cooling it is going to be a project in itself so generally people use electronic loads that return the energy to the grid. I wouldn't run it straight into the battery until I was confident that it's not going to blow up in my face.

Another point is it's unlikely this will be cheaper than buying something that's already out there in low volumes. How many are you planning to make?

Regarding the voltage - you're looking at an annoying range where you're probably going to have to use 600V/650V devices anyway. Are you thinking IGBT/superjunction MOSFETs or SiC?

[Psi]

Having a swimming pool is useful if you need to cool your dummy load

Failing that you can always have lots of 2kW oil heaters setup safely outside, 

The problem with developing really high power things is
- Everything is done on a large scale. huge dummy load etc.
- All the parts are expensive. 
- The parts tend to explode a lot during testing/development
- Fire risk is high.
- Shrapnel risk is high. (Less magic smoke events, more semiconductor detonations and arc flash events)

I recommend against testing with a raw unregulated battery as the power source. As tempting and cheap as it is.
You want some level of control. Ideally you want a system where it's monitoring the efficiency of your DCDC and will trip on unexpected low efficiency with high input power. ie. too many watts going somewhere it shouldn't be.

[jonpaul]

Bonjour,

Having worked in design. debug and mfg power electronics since 1970s, up to 12 KW, 300A, 70 kV, from hard expericne ,

A moment please:

The OP should consider:

Safety: Lethal voltages, possible arc flash, batttery fires.

Instrumentation: Special current prbes, meters, analyzers may be needed

Components: Many are critiacl to the operation, compliance or construction and are NOT stock eg the transformer and idnductor may be custom made. Expect to learn how to design and build magnetics!

The modern cycloconverters at these powers DO NOT use mains freq magnetis but run at high frequencey.

Spec: Even for a one off, safety and EMI need to be considered.

Chg level: Suggest to get the UK or other EV charge level spec 1,2,3.

Papers: See numerous papers and seminalrs on EV charging.

IEEE is a staring point.

Bon Chance


Jon

[madnight]

i wouldn't bother with isolating it. There's no real advantages and no regulatory requirements for it. You've got to follow UNECE reg 100 as the battery is more than 60V so it all needs to be isolated from the vehicle and that's the bit that will keep the user safe.

i have already had a skim throu that reg and know some of the basics and honestly to me it kinda common sence to totaly isolate battery from vhical in as meany ways as possible

Nice find with the Thunderstruck charge controller. From what I understand of CCS it sounds massively complicated and far too involved.

exsactly but doing research and what not about inplomenting reguler ac charges i came across a vidio about the thunder struck tms2500 wich in turn lead me to finding this controler

This might be a good application for an IPT (inter phase transformer) based buck converter.  The university of Manchester have a bunch of papers on them if you want to read more. It's basically two half bridges at 180 degrees feeding both ends of an equal winding autotransformer.. the middle point goes to the LC filter of a buck converter.
 If you keep the DC charger voltage at twice the battery voltage then there's essentially no ripple in the inductor or capacitor which could be good for losses. Basic control is the same as a two phase interleaved buck converter.

thank you i will have a look in to that design and that is my idea as i was thingking that might make it more easy

Regarding the power supplies, don't forget that you need a 13kW load as well. It can just be a big bank of resistors but cooling it is going to be a project in itself so generally people use electronic loads that return the energy to the grid. I wouldn't run it straight into the battery until I was confident that it's not going to blow up in my face.

and i know about a dummy load i going to use sevral heating elemants that way i can slowly adgust how much power i draw

Another point is it's unlikely this will be cheaper than buying something that's already out there in low volumes. How many are you planning to make?

well this is just a one off for myself and i asking about building one as i am raily struggleing to find one to buy that dose everything i need

Regarding the voltage - you're looking at an annoying range where you're probably going to have to use 600V/650V devices anyway. Are you thinking IGBT/superjunction MOSFETs or SiC?

i have not fully desided on that one yet but i was going to go well over kill on the parts anyways just to play safe ball

[madnight]

Having a swimming pool is useful if you need to cool your dummy load

Failing that you can always have lots of 2kW oil heaters setup safely outside, 

well that is my idae haveing meany meany heaters in a large body of water

The problem with developing really high power things is
- Everything is done on a large scale. huge dummy load etc.
- All the parts are expensive. 
- The parts tend to explode a lot during testing/development
- Fire risk is high.
- Shrapnel risk is high. (Less magic smoke events, more semiconductor detonations and arc flash events)

ho i know everything is sized up a lot but i already over size parts anyways for pice of mind and safty
i know they can get exspensive but building one over buying one meany you can buy parts when you can
and i already have planed for this i will keep sand and a CO2 fire exstinglisher ready at hand
and to provent shrapnal i going to do tests behind polycarbanate sheet and i will have safty glasses on to minimise eye damage from ark flash events

I recommend against testing with a raw unregulated battery as the power source. As tempting and cheap as it is.
You want some level of control. Ideally you want a system where it's monitoring the efficiency of your DCDC and will trip on unexpected low efficiency with high input power. ie. too many watts going somewhere it shouldn't be.

and i plan to moniter everything i can and have a air contactor conected to a foot switch so moment something gose wrong i can verry quickly disconect power and load

[NiHaoMike]

be able to support up to 100a out put
easily intagratable with a micocontroler
i would like side A (high voltage/charger side) to be able to support a minium of 270v dc
i would like side B (lower voltage but still high/battery side) to be able to support up to 135v dc

Just get a Prius inverter (about $100 on Ebay) and use the buck/boost converter inside it. The actual inverter outputs would likely be handy as well. There's a logic board that outputs PWM signals (from some ASICs), replace that with your own control board to make it do what you want it to do.

[mikeselectricstuff]

Some of the Vicor dcdc modules can work bidirectionally

[Faringdon]

Here is some LTspice on DAB converter...bidirectional
DAB means "Dual Active Bridge"......putting that into google should give you some tasty things.

[mzzj]

thank you for the infor
i have done suplys befor just nothing as powafull as this
and what would you recomend as turn ratio on the transformer

42:42 or fifty-fifty.

Your question pretty much reveals that this is not suitable level of project.

Btw: minimum voltage with Chademo is 50 volts and CCS 150v or 200v

[madnight]

Btw: minimum voltage with Chademo is 50 volts and CCS 150v or 200v

thats what the spec sheets say but from reading information and others exspariance not meany stations wock at those voltages

[max_torque]

What a silly thread.

if yhou do the math, then there is clearly a winning solution, which is to spend the money you were/are going to waste on this ridiculous dcdc on just buying some more cells for your vehicle


And of course, fundamentally you can't high power charge a "small" battery anyway, because the C rate of the cells becomes limiting, esp at high / low voltages and temperatures.  I'm going to guess you haven't got any significant BMS or BTMS experience as you are not asking anything like the correct questions

You also are foolish to connect something unproven, DIY, and completely uncertified at any public fast charger. The network operates are going to chase you for damages if you blow up their charger! And HV public liabilites insurance is expensive.


BTW I actually develop BEVs for a living, both OE and motorsport, and have power electronics experience to 1kV and 1MW in this field.


The "Best" route to DIY BEVs is simply to repurpose existing componentry, ie use a BMW CCS controller, an Leaf inverter/DCD or any number of Tesla parts that have all been hacked and provide a very cheap, very reliable, and importantly, VERY SAFE, architecture.  Even if you do all this, there are still hundreds of ways you can kill or set fire to yourself and others......

[Faringdon]

What a silly thread.

You're joking, right? This is a forum..a place to discuss and encourage this kind of thing....and by the way, your actual points are well received and a good contribution either way. so thanks for that.
Most Power Electronics in the West has now been outsourced to China so please  lets please not discourage some activity/discussion  in this area......the more the better.

basicaly i building a ev and want to take advantage of dc fast chargeing with out the need for a larger battery

This a great idea, and  well worth looking into.

if yhou do the math, then there is clearly a winning solution, which is to spend the money you were/are going to waste on this ridiculous dcdc on just buying some more cells for your vehicle

Thanks but i think they have a great idea.....maybe he wants to look into a cheaper, /lighter battery

You also are foolish to connect something unproven, DIY, and completely uncertified at any public fast charger.

Thanks, but i dont think they  are saying they are going to connect anything to a public fast charger at this point.

And to the OP, i confess that bidi chargers are amazing, but would a  "one-way_t'other way" type thing be out of the question?.......it may be a little bigger i would admit.
I mean, these things tend to be stationary and not mobile items. So does it really need to be "smallest"?
If so then the DAB  i state above is the way to go.

And of course, fundamentally you can't high power charge a "small" battery anyway, because the C rate of the cells becomes limiting, esp at high / low voltages and temperatures.

In Battlefield situations, you often take a risk and way overcharge batteries, because you need to take the risk.
In Formula E, they way overcharge batts......and yes they start smoking!...my ex gaffer used to be in charge of batts for UK Formula E.

[max_torque]

No, this thread is moronic and your reply is pretty stupid as well, sorry..

The OP said

" i building a ev and want to take advantage of dc fast chargeing with out the need for a larger battery"


Which tells me they have no actual understanding or expertise in power electronics, energy storage, battery electric vehicles or anything of any use.

Lets start at the beginning.  What IS the advantage of dc fast charging?  Obvious isn't it, it allows a high POWER charge to occur.  So when does one NEED to leverage that advantage?  Again, obvious, when one wants to rapidly charge a large capacity battery.


So fundamentally, it's stupid to want to leverage the only advantage of a system that is in no way applicable to the system the OP is saying they are going to install......


AC charging at up to 22kW (3ph) or 7kW(1ph) is going to be absolutely sufficient for a small storage battery. To accept the high power of a DC charger (50kW to 1MW) you absolutely need a large battery.

As result, half arsing some sort of converter together to allow one to plug a small (low voltage unless the cells they are using are tiny capacity, which would be idiotic as well) battery into a DC fast charger is stupid, dangerous and would be a legal nightmare, esp in the UK, where the OP claims to be from.



^^^ That's just the fact of it, it makes no difference what they do on a 'Battlefield' or in FormulaE or whatever... (btw i have designed powertrains and energy storage systems used in both those those environments to good (championship winner for FE)effect :-)

[Faringdon]

Thankyou this is excellent knowledge, i often think on this, since once i was terminated because i refused to implement a huge battery charger without a BMS...they insisted i just shovelled the output of our multi kW voltage output regulated power supply into it (this was a supply which we were using for something else). We had been told not to charge with over 70A, (it was another co's battery) but putting that supply on it would have meant way more than that, as it had really high current output clamp level, way over 70A. It was for a one-off demo for the MOD, so i was told the battery would survive it for that one go. I guess it wouldnt have from what you say. (im not harping on, but that co had all their actual sellable products made in China, and received the MOD grant based on the turnover they made from that operation)

[Zero999]

Yes. Silly thread. Ask a question about an advanced topic, then precede with beginner questions. I didn't bother responding when he asked what the turns ratio of the transformer needs to be. If you don't know how to work out something that simple, then you lack the knowledge and experience to design something as complicated as this. At least build a small scale power supply more complicated than an LM78xx circuit first.