Holy moly! This is no small improvement!! I doubt the non-engineers out there can really grasp what just happened! No, really... this concept will truly be a huge benefit... better low-end torque AND better high speed efficiency! They really don't have that much of a cost disadvantage either, as the Inverter transistors are serving dual purpose... drive the motor AND boost conversion for charging. SO COOL!!!
These EV tear-down videos are always my favorite. I love seeing the manufacturers coming up with clever new engineering solutions as EVs grow in the market. Thanks again!
When I was a little boy (70 now) I lived next to a small bakery beaking very nice brekfast bread - he had a machine to work with the dough that machine had a spesial oilfilled Y - Delta switch, he started in maby the Y config. and when it up in speed, he just switched to the Delta config. There was 2 on the same switch, and ofcorse the oil propably killed the sparks. No electronics, just meccanical and electrical parts in say 1964.... Nice memorys, and very very good both the smell of bread, an technics 😎👍
Where I work, We still use Delta-Y switch for our very big industrial mixer (Around 600 hp each). We still have 3 of them still working today. No spark and near direct switching because those use (Electronic Switching) Those drive are called Soft Start. Like 30-40 years later, still working well. 🙂 But of course, those must cost so much in those time compared to simple contactors used in those time. But we talk about 600hp working with 2400 volts, not 240 volts. Then those need very big Cabinet because those drive are so big and need so high voltage.
AugustaWestland had several 11kV 3,500 hp motors in 1982 that used a Wye-Delta starter. The changeover happened at about 30 seconds after startup, made a thump.
Some 30 years ago i learnt to scuba dive. The dive club had an old air compressor for filling the dive tanks. Sometime during the filling of a set of tanks the D-Y switch has to be operated to get up to the pressure we needed.
@@sergiomomesso1590 Interesting to hear 😎 . Edvin Olssons "dough moover" was maby about 100 litres and the motor was on 3 phase 380volts So in Y configuration the windigs get 220 volts, and in Delta the windings get 380volts. I guess the fuses in the house would blow out if it started in Delta position. Best wh. from Sweden.
Thanks for this fantastic tutorial on the Kia EV9 motor innovations. This is the type of information that would influence my first EV purchase. Brilliantly done!
You may need to correct the inaccurate information in this video. Based on the description on page 12 of the patent application US20240039448A1, the two configurations are close-end winding (CEW), which is essentially a star or wye connection, and open-end winding (OEW), which is NOT delta connection. With open-end winding connection, the two inverters feed PWM voltages to both ends of the the stator winding. In summary, the Delta connection is not utilized with this dual-inverter design.
That is interesting, but is the purpose of OEW if not to enable direct application of full PWM voltage to the winding? Effectively this is what delta configuration achieves.
I really wish to understand these matters more thoroughly as I’m looking for an engine to (cough) hybridize my Mazda RX-8, the best road tripping car I’ve ever had. Sorry, my Tesla 3P & YLR, though that reserved CT is looking awfully damn good, but maybe not quite on the ergonomics side. Wish me luck!
I have read this patent, it is a fantastic idea from Hyundai. OEW can implement a three-level inverter, and the two inverters use different power module --- SIC and IGBT. The star connection even retains the inverter's boost function. The only unknown is how much this will increase the cost with so many power devices.
Agree, I think that this additoinal connection is used to step up the voltage (as KIA Motor ENG said in the motor presentaiton video) when the car is connected to 400V charger to charge the 800V battery (using motor as equivalent transformer/indicive component) and has nothing to do with delta or anything related to motor operation.....
I did this Delta WYE on my very first diy ebike on the hub motor back in 2007. I called it the "electrical transmission". I was using a 3DPT relay to switch between the winding config. It worked very well but when switching back to WYE there was some care required to not blow the inverter mosfets with high bemf at high speed on WYE. I published all the info on Endless sphere forum. Was it me who tried this very first setup idea on EV then😊?
Heavy industrial vehicles use it as well, nothing special just not worth the trouble of implementing in EV's so far. Delta works fine there, consider how fast even lightly powered EVs can accelerate.
Great video detailing the new innovation of EV motor design. Industry is moving towards electrification and Tesla is not the only one pushing the EV technology forward anymore. Thanks for posting. :)
I feel the need to interject but the Renault Zoe has been using the motors as charging inductors for probably 10 years now. People have been complaining about the whining noise while charging until the reason was disclosed. Similarly, I might be wrong, but in some old electronics books I read about big motors being started in Y and switched to delta once up to speed. I guess it's an innovation in the powertrain industry but IGBT prices are now very small compared to battery prices. I'm sure thousands of people thought about this already. Sure, cost-effective implementation is a big step so I have to commend them on that.
With battery chemistry getting better and discovering that there's still so much stretch in electronics application. i love to watch more of this in the future !!
@@markplott4820 - depends who designed it, oems outsource their parts production for these types of specialist parts to raise quality levels. Will soon tell if the motor and inverter/charge-booster are third-party parts then they should soon start appearing in non-Hyundia/Kia/Genesis cars as well.
Looks more like a 2x3-phase system to me rather than a wye-delta. Did you tear down the connection ring to confirm the 3 additional leads are indeed the output terminals (i.e. open star point)? Are there relays in the inverter? You would have to be at no-load and outside your field-weakening region if you wanted to switch wye-delta and vice-versa without detrimental impact on drive dynamics (and avoid blowing your switches). I would be surprised if Kia went down that route, but please prove me wrong.
Kia and Hyundai's next gen Electric design with the improved heat pump is going to be the one to get. It's shocking how quick they've come with their EV's already. The big thing to solve is that frunk space.
Looks like a very high quality motor, and persumably with a significant volume of manufacture too, which can be tricky to achieve! I'd love to see a full tear down of the power electronics, and how the 400V and 800V DC links get connected to the inverter? Is there some hard switching upstream to either send 800V direct to the battery and 400V to the inverter and then to the battery, or do they always route the DC charging cables to the inverter and simply run in continuous conduction, non boosting mode for 800V charging? (and accept the small additional loss?)
This is very interesting! Thanks for the video. I would like to see the inverter teardown in the next one. I'm curious if there really happens switching between Δ and Y or if there are just two 3-phase H-bridges there, connected to 3 terminals of the motor, i.e. completely independent driver for each coil. The reason why Δ connection is not used in EVs is the variable inductance of the coils due to saliency of the rotor. I've measured it on the “Hyundai Ioniq Electric 28 kWh” motor. The inductance of the two coils in series in the Y configuration varies from 440 μH to 1100 μH depending on the angle of the rotor. As the motor controller regulates the current thru the coils, the Y configuration is better because just 2 current sensors outside of the motor are enough the know the current thru all three coils. In the Δ configuration, the current splits according to inductances and if these are not the same and varying one would need the current sensor on each of the coils to be able to control it. That's why a motor with Δ configuration would probably have 6 terminals as well, because the current sensor is naturally housed in the inverter and the Δ connection would be realized there. Switching between Δ and Y would allow them to use higher inductance of two coils in series in Y configuration in lower speed to be able to control lower currents more efficiently. At higher speeds they could switch to Δ-configuration, alleviating the need for field weakening. But this would be at the cost of additional Rdson of the transistors making up this switch. On top of that these would have to be actually 2 NMOS SiC in series (IGBTs would need additional components as we need it to switch the current in both directions). Permanent 6-wire connection to two 3-phase H-bridges, on the other hand would have higher switching losses at lower speeds, so I am really curious which architecture they actually chose. Other automakers like Honda use step-up converter (350V -> 700V) in their hybrid vehicles to shift the need for a field weakening towards higher speeds.
Very informative comment. The Munro live community never ceases to amaze me with their insights. We did check the winding pattern and physical connections to the inverter to determine whether this was set up as two 3-phase machines or a wye-delta arrangement. We also did a patent search and found a Hyundai patent covering this circuit and control method.
The description and the appearance of the inverter case suggest that the inverter has six power outputs using three of them for delta and the other three for wye.
@@brianb-p6586 That's highly unlikely. The motor has 6 terminals where each pair of terminals is connected to the coils of each phase, respectively. Having a stator with two sets of windings one for Δ, the other for Y would be wasteful as most of the time only one set of coils would be conducting current. This means -- if we made a single set of coils of twice a cross-sectional area of the conductors we would have 50% reduction in losses. Therefore I think it is better to have both terminals of each phase exposed and let the inverter to do the right connection. Besides, in Δ configuration, the controller needs to know the currents in each coil. It is not enough to measure to summed current in the vertices of the triangle and it does not seem that the current sensors are embedded into motor windings. The are usually found in the inverter.
@32bits-of-a-bus59 you're right - I had not had my coffee yet! The six inverter terminals, each driven by one power section, need to connect to the six winding terminals. All six are required at all times, driven to effectively make the two different connection topologies. In delta each phase would need to be driven by two inverter sections, with their net current equaling the net phase current, but current from one winding to the connected (in delta) phase passing through both sections instead of none. In wye, three inverter sections would need to serve as the three phase outputs, with the other three effectively shorting the other ends of the windings together.
I'm not too sure why the Munroe engineer is acting so surprised at this technology. It was introduced to Munroe by a Canadian company, Exro Technologies, about a year ago. Their Coil Driver technology is now certified to be used in passenger cars in the United States.
The idea of using the motor winding as a second application purpose ans avoid the need for additional coil is fantastic. Back in 2015 there was a company from Moscow that I represented at the Interbike in Vegas called adaptto E drive that was offering ebike motor controller with very similar idea but it was to use the motor as boost converter coil to recharge the battery from wide lower voltage range.
I don't fully understand using the 4th, extra winding's inductance(?) to boost 400V charger to 800V vehicle architecture...but it sounds just like what your describing with previous adaptto application?
the switch from wye to delta is due to large back emf spikes damaging the controller, so you just need to have no load on the motor when you do the switch, nothing complicated, but you missed the reason for motor off during switching, motor can still rotate of course just need to have low load or no load at time of switching. The time for the back emf to be absorbed would be very short, you wouldn't notice power was off for a miniscule time period.
@@krashanb5767 When you disconnet a motor under load, as in disconnect the power wires quickly whilst accelerating or going up hill, since there is a huge amount of energy stored in the electric field, it collapses very very rapidly, that energy has to go somewhere, it can generate a very voltage potential and a large spark, which can damage mosfets/igbts etc. So the controller needs to absorb that energy rather than fry the igbts with high voltage spike. I'm not sure which method they use to absorb this spike, I'm guessing it will be diodes or capacitors or a combination of the two, but since it occurs over a very short time period, it would be too small a glitch to even notice any change in power if they do power off the motor load just before switching. I'm only guessing here how they do it, some electronics expert in the field would likely know the answer in detail. It would make for a rather interesting video.
As well as timing/load based control, I wonder if you could use a supercapacitor to dampen the back emf spikes, protecting the switches, and re-use the energy after transitioning
Thanks, so much! I really wish to understand these matters more thoroughly, as I’m looking for an engine to (cough) hybridize my Mazda RX-8, the best road tripping car I’ve ever had. Sorry, my Tesla 3P & YLR, though that reserved CT is looking awfully damn good, but maybe not quite on the ergonomics side. Wish me luck! Re: Inverter Versatility. Also, for bi-directional flows for possible doubling as power walls?
Excellent presentation! I don't know if what Kia did could be termed First Principals thinking, but it is definitely thinking outside the box. Couple of notes: 1) When Paul mentions "interior magnet" the full description of the term is "interior permanent magnet". 2) When he refers to the "switches" in the inverter, I'm pretty sure he is referring to the Silicon Carbide switches used by the inverter (not sure if they are MOS FETs or IGBTs). Interesting that Tesla has backed off of SiC based switches (as opposed to traditional silicon), as Kia ramps up their use.
Wow.. This is impressive. I was talking about doing this also -- but by having both sets of windings inside the motor--- so you could just switch from one to the other. but they have gone even further and gotten the same functions out of one set of windings.... And then like you're saying with the "boost" charging-- it's like they are passing extra charge through the same path as the regen current.. So like a simulated regen (as an analogy). Very cool. I wonder how many patents they have on all those design changes. lol..
This star-delta reconfiguration reminds me of industrial applications such as trains. I always wondered why they "shifted" or made a tone change from low to high.
In effect, the switch between the "Y" and "Delta" configurations acts like a 2-gear transmission. Will there be a clutch and gear shift lever offered to simulate a manual transmission? 😀
Delta configuration allows to circulate 3rd harmonic current components which in a triangular connection cannot do as they are zero sequence/ homopolar currents. Overall is a bad decition to go to delta connection (at least if motor is tracktioning) as long as this 3rd harmonic is not being control by SW which I do not think is happening. It would be great if you have the chance, to ask to Hyundai ENG how do they deal with this 3rd harmonic issue. Very Good Video as always!
Smart and right comment. Actually it seems that the machine is not physically delta connected. As spotted out by someone in this thread, the patent shows that they implemented an open end winding scheme. In virtual delta connection the homopolar machine current can be actively controlled (the drawback is a reduction in the maximum voltage limit for torque generation). It would be interesting to analyze the machine to see if they made design choices to reduce the third harmonic in the back emf by winding scheme or, better, rotor geometry.
This will make regenerative braking more efficient! Hope it becomes standard on all EVs ASAP. Electric bicycles could also use it. Monroe should look into Freegen and maybe help that new technology advance into production faster.
I wish you would draw the same graph in the "Getting Wound Up" video which compared bar and stranded motors for both Wye and Delta configurations. It seems this ability switch between why and delra really cements the strength of bar windings by pushing the AC losses out in the Delta wiring.
This wye-delta configuration is very common in industrial applications where induction machines need to be started from line frequency directly (without inverter). Wye connection reduces effective voltage by sqrt(3) thereby also reducing the inrush current considerably. After machine reaches certain speed it switches to delta for higher power. In this motor I assume it makes easier for inverter to control the current during the run up phase.
2 місяці тому+3
Does this re-wiring the motor mean all the current goes via power silicone twice? That could mean double the losses if so, in the inverter.
Yes, the way the "dual stage" inverter is set up the current passes through both sets of switches. Only one set has to pulse width modulate though so only the conduction losses from the second set are added. This allows Kia to smoothly transition between wye and delta unlike the contactor method that some commenters have mentioned.
Some current will need to go through two power stages instead of one, and some will go through two instead of none, depending on the configuration. Yes, that has an efficiency cost. The word is "silicon", not "silicone"... and the semiconductor may not be silicon.
I wonder if changing the operating mode from Wye to Delta requires very precise waveform synchronization. I think that the high RPMs and the fact that the engine works with magnetic slip anyway make the change at a random moment imperceptible. I would like to hear what people smarter than me think.
I have a 38kW Ioniq and a 24 Leaf..on the same path, same speed, same temperature etc, to get 14,5kW/100km of consumption on the Leaf, I get 9.9 on the Ioniq.., with the AC running (not on the Leaf). The efficiency of these Hyundai/Kia motors is quite extraordinary.
Safe to assume this is a 2024 EV9 not the forthcoming 2025? Just want to make sure it's not a coming advancement but what's already out there. Great video! We love our EV9 and learning more about the tech behind it.
I don't know how he can feel comfortable making this statement: "No one's been able to come up with a practical way to actually implement this for a traction machine in an EV." Exro came up with a more advanced application of this concept, and it has been publicly available for several years. Exro's technology is well known to Monroe & Associates. A few months ago, Sandy Munro called it "Magic."
I wonder if the EV6 and Ioniq 5 refreshes will get this new motor design. Also.. I know the EV6 GT and 5N use a unique dual inverter design on the rear primary drive motor where only one inverter functions during low power mode or both inverters provide power for high output mode, but I don’t know if the rear motor has the ability to switch from wye to delta like this. I’d be really interested if Munro did a teardown of an EV6 GT/Ioniq 5N rear motor to see how it compares to the regular EV6/Ioniq 5 and this new EV9 motor design.
I drove the ev9 recently, 30 miles, for such a large vehicle, the consumption was remarkably low. 16-18 KWh in mixed driving situations, mostly highway. At 100-110 mph it‘s weight is pretty hard to handle.
Been a while I heard about Y/delta switching. People say theres not much to innovate in EVs, well... Also from what I learnt in the industrial context, what he said about needing to switch off the engine to switch between modes isn't correct. You use one when starting the motor from 0 rpm to reduce the initial current spike, and then switch to the other mode once the more nominal speed is established.
In the HVAC industry they have used Why Delta starters for many years to run big centrifugal chillers. Before the advent of inverters they used a series of large contactors to make this switch, mounted in the power cabinet on the Chillers so its nothing new , just new to EV cars and honestly, I don't know why they haven't done it sooner, seemed like a no brainer with our inverter technology. I bet ABB had it and they just took their tech and made it compact and integrated it in the cars.
Not sure if there is inverter tear down video or not. If it is 6 phase leg inverter, the system is open winding configuration, allowing it to use low voltage dc bus to drive higher voltage machine.
When using the motor as an inductor, what affect does this have upon the permanent magnets in the rotor? Would this not subject them to excessive AC magnetic drive causing them to be demagnetized?
Thoughts on using Kia motor as boost converter to achieve 800V from a 400V charger, vs the Chevy dividing battery to be in series or parallel based on charger voltage (in terms of efficiency/cost/feasibility). Some initial reports on the Kia has charging limited at 400V to under 100kWh. I would imagine the SilveradoEV would have a better charging curve at both 400 and 800V, and cost less.
Great video, does this mean the EV9 can actually get more power through their boost converter than the prior EV6/Ionic approach? For example, can they use the star configuration to create three parallel centre tapped boost inductors or was this hwo the previous motor worked in any case with it's single additional tab?
I've read that it's not so much that a wye winding makes more torque, but that it improves the "torque constant" - i.e. the wye winding makes the same torque for less current than the delta configuration. Why is that?
Not clear from the video, is the the front or rear electric motor, or is it used in both? Also, is Kia still using the mechanical disconnect for the front motor like they did in the 1st egmp vehicles?
Plenty of equipment in the industrial industry that switches between Star and delta without stopping the motor , not sure why you would think otherwise, I bet the compressor in your workshop has a star delta starter in it.
I’ve seen ev9’s managing less than 100kwh speed at Tesla superchargers, do you think by the time Kia gets access to Tesla superchargers next year they’ll be able to bump up the charge speed since superchargers are meant for 400v ev’s
Holy moly! This is no small improvement!! I doubt the non-engineers out there can really grasp what just happened! No, really... this concept will truly be a huge benefit... better low-end torque AND better high speed efficiency!
They really don't have that much of a cost disadvantage either, as the Inverter transistors are serving dual purpose... drive the motor AND boost conversion for charging.
SO COOL!!!
Re: Inverter Versatility. Also, for bi-directional flows to double as power walls?
These EV tear-down videos are always my favorite. I love seeing the manufacturers coming up with clever new engineering solutions as EVs grow in the market. Thanks again!
I really want to see a tear down on the new Mach E and Chevy EV truck. Apparently both are quite good.
As a past ‘56 ‘vette owner, I’m hoping GM will request that they tear into the new ones.
When I was a little boy (70 now) I lived next to a small bakery beaking very nice brekfast bread - he had a machine to work with the dough that machine had a spesial oilfilled Y - Delta switch, he started in maby the Y config. and when it up in speed, he just switched to the Delta config. There was 2 on the same switch, and ofcorse the oil propably killed the sparks.
No electronics, just meccanical and electrical parts in say 1964....
Nice memorys, and very very good both the smell of bread, an technics 😎👍
Where I work, We still use Delta-Y switch for our very big industrial mixer (Around 600 hp each). We still have 3 of them still working today. No spark and near direct switching because those use (Electronic Switching) Those drive are called Soft Start. Like 30-40 years later, still working well. 🙂 But of course, those must cost so much in those time compared to simple contactors used in those time. But we talk about 600hp working with 2400 volts, not 240 volts. Then those need very big Cabinet because those drive are so big and need so high voltage.
AugustaWestland had several 11kV 3,500 hp motors in 1982 that used a Wye-Delta starter. The changeover happened at about 30 seconds after startup, made a thump.
Some 30 years ago i learnt to scuba dive. The dive club had an old air compressor for filling the dive tanks. Sometime during the filling of a set of tanks the D-Y switch has to be operated to get up to the pressure we needed.
@@sergiomomesso1590 Interesting to hear 😎 .
Edvin Olssons "dough moover" was maby about 100 litres and the motor was on 3 phase 380volts
So in Y configuration the windigs get 220 volts, and in Delta the windings get 380volts. I guess the fuses in the house would blow out if it started in Delta position.
Best wh. from Sweden.
신기하네요 ㅎㅎㅎ 건강하세요 할아버지!
Thanks for this fantastic tutorial on the Kia EV9 motor innovations. This is the type of information that would influence my first EV purchase. Brilliantly done!
Great to hear!
Which LLM did you use to write that response?
Look at watt-hours per mile as the true metric.
You may need to correct the inaccurate information in this video. Based on the description on page 12 of the patent application US20240039448A1, the two configurations are close-end winding (CEW), which is essentially a star or wye connection, and open-end winding (OEW), which is NOT delta connection. With open-end winding connection, the two inverters feed PWM voltages to both ends of the the stator winding. In summary, the Delta connection is not utilized with this dual-inverter design.
That is interesting, but is the purpose of OEW if not to enable direct application of full PWM voltage to the winding? Effectively this is what delta configuration achieves.
I really wish to understand these matters more thoroughly as I’m looking for an engine to (cough) hybridize my Mazda RX-8, the best road tripping car I’ve ever had. Sorry, my Tesla 3P & YLR, though that reserved CT is looking awfully damn good, but maybe not quite on the ergonomics side. Wish me luck!
I have read this patent, it is a fantastic idea from Hyundai. OEW can implement a three-level inverter, and the two inverters use different power module --- SIC and IGBT. The star connection even retains the inverter's boost function. The only unknown is how much this will increase the cost with so many power devices.
Power electronics are cheap compared to mechanical parts like alloy casting parts.. this should be no surprise and truly ground breaking
Agree, I think that this additoinal connection is used to step up the voltage (as KIA Motor ENG said in the motor presentaiton video) when the car is connected to 400V charger to charge the 800V battery (using motor as equivalent transformer/indicive component) and has nothing to do with delta or anything related to motor operation.....
It’s all the details. This is genius
Charge on!
That's what I call innovation😊
@@psulak02 - outsourcing is innovation? ......lol.
Outsourcing is great and Its lowering costs!
@@markplott4820 The technology itself is an innovation. Who does it doesn't matter.
@@EinzigfreierName - I dont see ANY Legacy OEM thats INNOVATING.
@@BrunoHeggli-zp3nl - then why RIVIAN , Loocid , and LEGACY auto , all are NOT profitable ?
Example of Excellent engineering
thank you folks. Great to see these units cracked open.
Brilliant, Thanks for sharing and explaining.:)
My pleasure!
I did this Delta WYE on my very first diy ebike on the hub motor back in 2007. I called it the "electrical transmission". I was using a 3DPT relay to switch between the winding config. It worked very well but when switching back to WYE there was some care required to not blow the inverter mosfets with high bemf at high speed on WYE. I published all the info on Endless sphere forum. Was it me who tried this very first setup idea on EV then😊?
Heavy industrial vehicles use it as well, nothing special just not worth the trouble of implementing in EV's so far. Delta works fine there, consider how fast even lightly powered EVs can accelerate.
The star-delta or wye-delta starter has been used in 3 phase AC motors at least as far back as the middle of the last century.
Great video detailing the new innovation of EV motor design. Industry is moving towards electrification and Tesla is not the only one pushing the EV technology forward anymore. Thanks for posting. :)
I feel the need to interject but the Renault Zoe has been using the motors as charging inductors for probably 10 years now. People have been complaining about the whining noise while charging until the reason was disclosed. Similarly, I might be wrong, but in some old electronics books I read about big motors being started in Y and switched to delta once up to speed. I guess it's an innovation in the powertrain industry but IGBT prices are now very small compared to battery prices. I'm sure thousands of people thought about this already. Sure, cost-effective implementation is a big step so I have to commend them on that.
The ACP Tzero and original design of the Tesla Roadster also used the motor for charging.
Wanted to say the same thing. The Zoé is 12 years old now :D. Interesting that they came back to it
This is not new tech and the advantages are significantly overstated. This whole video reeks of sponsored astroturfing.
With battery chemistry getting better and discovering that there's still so much stretch in electronics application. i love to watch more of this in the future !!
🙋♂️THX PAUL AND THE MUNRO TEAM ,😎🧐💚💚💚
So impressive! Awesome technology and build quality! Thank you for the video!!
Thank you for sharing this info. Munro is the place to go if you’re in the service and repair business and want to stay abreast of new tech!
Thats is some cool electrical engineering!
@@jake5993 - ZERO done by KIA ,done by others.....lol.
@@markplott4820 Maybe so... but somebody had to approve it.
German engineering Baby!
@@markplott4820 - depends who designed it, oems outsource their parts production for these types of specialist parts to raise quality levels. Will soon tell if the motor and inverter/charge-booster are third-party parts then they should soon start appearing in non-Hyundia/Kia/Genesis cars as well.
@@markplott4820Done by LG for EV9. Mobis supplies motors for other EV lineup.
Very informative and at the same time something new and different information regarding upcoming tech.
Wow, real innovation
Looks more like a 2x3-phase system to me rather than a wye-delta. Did you tear down the connection ring to confirm the 3 additional leads are indeed the output terminals (i.e. open star point)? Are there relays in the inverter? You would have to be at no-load and outside your field-weakening region if you wanted to switch wye-delta and vice-versa without detrimental impact on drive dynamics (and avoid blowing your switches). I would be surprised if Kia went down that route, but please prove me wrong.
In just five years the EV technology is going to be WOW.
Maybe. We keep hearing promised of super batteries and definitely progress has been made. Frankly, it makes more sense to produce "gen set" hybrids.
Excellent presentation, very informative and clear. Genius design and excellent execution
Kia and Hyundai's next gen Electric design with the improved heat pump is going to be the one to get. It's shocking how quick they've come with their EV's already. The big thing to solve is that frunk space.
They are killing it recently! And really trying to show quick business like tesla does!
@@jake5993 - NOPE, can't scale. cars too expensive & heavy.
We have lived without a fronk for decades no big deal.
Yeahhh, now if they can paint their trash.
@@ConstanceCox - Cybertruck needs no paint.
Thanks, Paul. Kia is ahead of everyone in this.
That's an amazing one! Thanks for the tear down.. enjoyed it
Looks like a very high quality motor, and persumably with a significant volume of manufacture too, which can be tricky to achieve! I'd love to see a full tear down of the power electronics, and how the 400V and 800V DC links get connected to the inverter? Is there some hard switching upstream to either send 800V direct to the battery and 400V to the inverter and then to the battery, or do they always route the DC charging cables to the inverter and simply run in continuous conduction, non boosting mode for 800V charging? (and accept the small additional loss?)
Another excellent episode with great insight into details 👍👍
Glad you enjoyed it
Fascinating. clear Expo. Thanks
This is very interesting! Thanks for the video. I would like to see the inverter teardown in the next one.
I'm curious if there really happens switching between Δ and Y or if there are just two 3-phase H-bridges there, connected to 3 terminals of the motor, i.e. completely independent driver for each coil.
The reason why Δ connection is not used in EVs is the variable inductance of the coils due to saliency of the rotor. I've measured it on the “Hyundai Ioniq Electric 28 kWh” motor. The inductance of the two coils in series in the Y configuration varies from 440 μH to 1100 μH depending on the angle of the rotor.
As the motor controller regulates the current thru the coils, the Y configuration is better because just 2 current sensors outside of the motor are enough the know the current thru all three coils. In the Δ configuration, the current splits according to inductances and if these are not the same and varying one would need the current sensor on each of the coils to be able to control it.
That's why a motor with Δ configuration would probably have 6 terminals as well, because the current sensor is naturally housed in the inverter and the Δ connection would be realized there.
Switching between Δ and Y would allow them to use higher inductance of two coils in series in Y configuration in lower speed to be able to control lower currents more efficiently. At higher speeds they could switch to Δ-configuration, alleviating the need for field weakening.
But this would be at the cost of additional Rdson of the transistors making up this switch. On top of that these would have to be actually 2 NMOS SiC in series (IGBTs would need additional components as we need it to switch the current in both directions).
Permanent 6-wire connection to two 3-phase H-bridges, on the other hand would have higher switching losses at lower speeds, so I am really curious which architecture they actually chose.
Other automakers like Honda use step-up converter (350V -> 700V) in their hybrid vehicles to shift the need for a field weakening towards higher speeds.
Very informative comment. The Munro live community never ceases to amaze me with their insights. We did check the winding pattern and physical connections to the inverter to determine whether this was set up as two 3-phase machines or a wye-delta arrangement. We also did a patent search and found a Hyundai patent covering this circuit and control method.
@@PaulTurnbull-EMfield Interesting! What did you find? Do they use Δ-Y switching or just 3 independent H-bridges, one for each coil?
The description and the appearance of the inverter case suggest that the inverter has six power outputs using three of them for delta and the other three for wye.
@@brianb-p6586 That's highly unlikely. The motor has 6 terminals where each pair of terminals is connected to the coils of each phase, respectively. Having a stator with two sets of windings one for Δ, the other for Y would be wasteful as most of the time only one set of coils would be conducting current. This means -- if we made a single set of coils of twice a cross-sectional area of the conductors we would have 50% reduction in losses. Therefore I think it is better to have both terminals of each phase exposed and let the inverter to do the right connection. Besides, in Δ configuration, the controller needs to know the currents in each coil. It is not enough to measure to summed current in the vertices of the triangle and it does not seem that the current sensors are embedded into motor windings. The are usually found in the inverter.
@32bits-of-a-bus59 you're right - I had not had my coffee yet! The six inverter terminals, each driven by one power section, need to connect to the six winding terminals. All six are required at all times, driven to effectively make the two different connection topologies.
In delta each phase would need to be driven by two inverter sections, with their net current equaling the net phase current, but current from one winding to the connected (in delta) phase passing through both sections instead of none.
In wye, three inverter sections would need to serve as the three phase outputs, with the other three effectively shorting the other ends of the windings together.
Outstanding lecture professor👌👏👍!! Always enjoy your tutorials!
I'm not too sure why the Munroe engineer is acting so surprised at this technology. It was introduced to Munroe by a Canadian company, Exro Technologies, about a year ago. Their Coil Driver technology is now certified to be used in passenger cars in the United States.
Thanks again, Paul, for the excellent explanations and historical perspective, and highlighting the monumental changes we are going through.
The idea of using the motor winding as a second application purpose ans avoid the need for additional coil is fantastic. Back in 2015 there was a company from Moscow that I represented at the Interbike in Vegas called adaptto E drive that was offering ebike motor controller with very similar idea but it was to use the motor as boost converter coil to recharge the battery from wide lower voltage range.
I was thinking about the adaptto when I watched the video and now I see your comment 😄
I don't fully understand using the 4th, extra winding's inductance(?) to boost 400V charger to 800V vehicle architecture...but it sounds just like what your describing with previous adaptto application?
Renault did it eons ago
the switch from wye to delta is due to large back emf spikes damaging the controller,
so you just need to have no load on the motor when you do the switch, nothing complicated, but you missed the reason for motor off during switching, motor can still rotate of course just need to have low load or no load at time of switching. The time for the back emf to be absorbed would be very short, you wouldn't notice power was off for a miniscule time period.
So it would work like a two speed powerglide on a drag racer 😊 Just shifting faster. Cool 😊
@@krashanb5767 When you disconnet a motor under load, as in disconnect the power wires quickly whilst accelerating or going up hill, since there is a huge amount of energy stored in the electric field, it collapses very very rapidly, that energy has to go somewhere, it can generate a very voltage potential and a large spark, which can damage mosfets/igbts etc. So the controller needs to absorb that energy rather than fry the igbts with high voltage spike. I'm not sure which method they use to absorb this spike, I'm guessing it will be diodes or capacitors or a combination of the two, but since it occurs over a very short time period, it would be too small a glitch to even notice any change in power if they do power off the motor load just before switching. I'm only guessing here how they do it, some electronics expert in the field would likely know the answer in detail.
It would make for a rather interesting video.
As well as timing/load based control, I wonder if you could use a supercapacitor to dampen the back emf spikes, protecting the switches, and re-use the energy after transitioning
Just brilliant 👏. Thanks, KIA!
Thankyou for that one guys, most enjoyable and enlightening.
excellent presentation
Thank you! Cheers!
Thanks, so much! I really wish to understand these matters more thoroughly, as I’m looking for an engine to (cough) hybridize my Mazda RX-8, the best road tripping car I’ve ever had. Sorry, my Tesla 3P & YLR, though that reserved CT is looking awfully damn good, but maybe not quite on the ergonomics side. Wish me luck! Re: Inverter Versatility. Also, for bi-directional flows for possible doubling as power walls?
Excellent presentation! I don't know if what Kia did could be termed First Principals thinking, but it is definitely thinking outside the box.
Couple of notes:
1) When Paul mentions "interior magnet" the full description of the term is "interior permanent magnet".
2) When he refers to the "switches" in the inverter, I'm pretty sure he is referring to the Silicon Carbide switches used by the inverter (not sure if they are MOS FETs or IGBTs). Interesting that Tesla has backed off of SiC based switches (as opposed to traditional silicon), as Kia ramps up their use.
re 2), cost most likely SiCs are a fair bit more expensive than traditional silicon.
Hasn’t the wye delta switch been around forever. I believe even the BART train from 1970 uses it.
Yep , don't know why he is ignoring that
I never thought this was possible Thanks for sharing
Brilliant explanation
That's brilliant engineering! I enjoyed hearing about this.
All the videos this guy does are amazing
Does anyone know, if Ioniq 5 Facelift also is getting this new motor?
Hyundai/Kia are killing it. It's a company you can tell enjoys making cars.
Some early electric bike motors have been using the wye/delta switch method for a long long time.
star/delta starters have been the norm in 3 phase industrial motors for like a century.
@@spankeyfish interesting
Wow.. This is impressive. I was talking about doing this also -- but by having both sets of windings inside the motor--- so you could just switch from one to the other. but they have gone even further and gotten the same functions out of one set of windings.... And then like you're saying with the "boost" charging-- it's like they are passing extra charge through the same path as the regen current.. So like a simulated regen (as an analogy). Very cool. I wonder how many patents they have on all those design changes. lol..
Love EV tear downs, learn so much. Hindustan south. Show the components more.
Brilliant genius implementation. Like dual cam VTEC engines
Good explanation. I'm waiting from 1st video of Munro.. Never disappoint in EV teardown Always best...Keep it up..
Renault has been using a 4 connection motors for many years in their Zoe, for the inductors of the charger boost.
Yes, I think few here know that.
This star-delta reconfiguration reminds me of industrial applications such as trains. I always wondered why they "shifted" or made a tone change from low to high.
In effect, the switch between the "Y" and "Delta" configurations acts like a 2-gear transmission.
Will there be a clutch and gear shift lever offered to simulate a manual transmission? 😀
Delta configuration allows to circulate 3rd harmonic current components which in a triangular connection cannot do as they are zero sequence/ homopolar currents. Overall is a bad decition to go to delta connection (at least if motor is tracktioning) as long as this 3rd harmonic is not being control by SW which I do not think is happening. It would be great if you have the chance, to ask to Hyundai ENG how do they deal with this 3rd harmonic issue.
Very Good Video as always!
Smart and right comment. Actually it seems that the machine is not physically delta connected. As spotted out by someone in this thread, the patent shows that they implemented an open end winding scheme. In virtual delta connection the homopolar machine current can be actively controlled (the drawback is a reduction in the maximum voltage limit for torque generation). It would be interesting to analyze the machine to see if they made design choices to reduce the third harmonic in the back emf by winding scheme or, better, rotor geometry.
This will make regenerative braking more efficient! Hope it becomes standard on all EVs ASAP. Electric bicycles could also use it. Monroe should look into Freegen and maybe help that new technology advance into production faster.
Thank you for another great video! I really enjoy the "nuts and bolts" of it all. Cheers
Great video
Thanks!
Best report
Thanks Paul
I wish you would draw the same graph in the "Getting Wound Up" video which compared bar and stranded motors for both Wye and Delta configurations. It seems this ability switch between why and delra really cements the strength of bar windings by pushing the AC losses out in the Delta wiring.
This wye-delta configuration is very common in industrial applications where induction machines need to be started from line frequency directly (without inverter). Wye connection reduces effective voltage by sqrt(3) thereby also reducing the inrush current considerably. After machine reaches certain speed it switches to delta for higher power.
In this motor I assume it makes easier for inverter to control the current during the run up phase.
Does this re-wiring the motor mean all the current goes via power silicone twice? That could mean double the losses if so, in the inverter.
Yes, the way the "dual stage" inverter is set up the current passes through both sets of switches. Only one set has to pulse width modulate though so only the conduction losses from the second set are added. This allows Kia to smoothly transition between wye and delta unlike the contactor method that some commenters have mentioned.
Some current will need to go through two power stages instead of one, and some will go through two instead of none, depending on the configuration. Yes, that has an efficiency cost.
The word is "silicon", not "silicone"... and the semiconductor may not be silicon.
I wonder if changing the operating mode from Wye to Delta requires very precise waveform synchronization. I think that the high RPMs and the fact that the engine works with magnetic slip anyway make the change at a random moment imperceptible. I would like to hear what people smarter than me think.
"Magnetic slip"? This is not an induction motor.
Great stuff to show to people who think EVs aren't well developed
I'm more concerned about the charging infrastructure than the vehicles themselves.
I have a 38kW Ioniq and a 24 Leaf..on the same path, same speed, same temperature etc, to get 14,5kW/100km of consumption on the Leaf, I get 9.9 on the Ioniq.., with the AC running (not on the Leaf).
The efficiency of these Hyundai/Kia motors is quite extraordinary.
This is the concept from a ship's architecture. We have a same design at Maersk Line used in "Bow Thruster" motor.
Safe to assume this is a 2024 EV9 not the forthcoming 2025? Just want to make sure it's not a coming advancement but what's already out there. Great video! We love our EV9 and learning more about the tech behind it.
Very cool; this would be great to see in trucks with V2L or V2G
I don't know how he can feel comfortable making this statement: "No one's been able to come up with a practical way to actually implement this for a traction machine in an EV."
Exro came up with a more advanced application of this concept, and it has been publicly available for several years. Exro's technology is well known to Monroe & Associates. A few months ago, Sandy Munro called it "Magic."
Is exro mass producing them for production vehicles?
Exro has been around since 2008 trying to sell their silly idea.
@MunroLive What is the model of the electric motor from Vitesco
The switching from y to delta is great, but what is the theoretical vs the real world efficiency.
I wonder if the EV6 and Ioniq 5 refreshes will get this new motor design. Also.. I know the EV6 GT and 5N use a unique dual inverter design on the rear primary drive motor where only one inverter functions during low power mode or both inverters provide power for high output mode, but I don’t know if the rear motor has the ability to switch from wye to delta like this. I’d be really interested if Munro did a teardown of an EV6 GT/Ioniq 5N rear motor to see how it compares to the regular EV6/Ioniq 5 and this new EV9 motor design.
ua-cam.com/video/xZoWRgDkZuE/v-deo.html switch
ev6gt / i5n / ev9 is same
gm and Hyundai are teaming up. I could imagine GM borrowing some expertise from the Ionic 5N to make an EV Camaro.
I drove the ev9 recently, 30 miles, for such a large vehicle, the consumption was remarkably low. 16-18 KWh in mixed driving situations, mostly highway. At 100-110 mph it‘s weight is pretty hard to handle.
Been a while I heard about Y/delta switching. People say theres not much to innovate in EVs, well...
Also from what I learnt in the industrial context, what he said about needing to switch off the engine to switch between modes isn't correct. You use one when starting the motor from 0 rpm to reduce the initial current spike, and then switch to the other mode once the more nominal speed is established.
In the HVAC industry they have used Why Delta starters for many years to run big centrifugal chillers. Before the advent of inverters they used a series of large contactors to make this switch, mounted in the power cabinet on the Chillers so its nothing new , just new to EV cars and honestly, I don't know why they haven't done it sooner, seemed like a no brainer with our inverter technology. I bet ABB had it and they just took their tech and made it compact and integrated it in the cars.
Not sure if there is inverter tear down video or not. If it is 6 phase leg inverter, the system is open winding configuration, allowing it to use low voltage dc bus to drive higher voltage machine.
Very informative.
Renault uses the motor windings of their ZOE and other Z.E vevicles as inductance for high power ac charging up to 43kW since more than 10years.
Great video! Has Tesla also used this technique in their latest motors?
This is awesome!
Munro Live 1M+
Very informative thanks
When using the motor as an inductor, what affect does this have upon the permanent magnets in the rotor? Would this not subject them to excessive AC magnetic drive causing them to be demagnetized?
Thoughts on using Kia motor as boost converter to achieve 800V from a 400V charger, vs the Chevy dividing battery to be in series or parallel based on charger voltage (in terms of efficiency/cost/feasibility). Some initial reports on the Kia has charging limited at 400V to under 100kWh. I would imagine the SilveradoEV would have a better charging curve at both 400 and 800V, and cost less.
Great video, does this mean the EV9 can actually get more power through their boost converter than the prior EV6/Ionic approach? For example, can they use the star configuration to create three parallel centre tapped boost inductors or was this hwo the previous motor worked in any case with it's single additional tab?
Hi Fidel Here, I love the video. Now if the winding of the motor is been used to charge the battery, doesn’t these shorten the life of it.
Thanks.
Great video as usual. Can you please provide more compute & sensing videos?
I've read that it's not so much that a wye winding makes more torque, but that it improves the "torque constant" - i.e. the wye winding makes the same torque for less current than the delta configuration.
Why is that?
Good job
Thanks
Thank you for nice Tutorials. Im not 100% shure but I mean that a part of this Innovation is from Rimac.
Not clear from the video, is the the front or rear electric motor, or is it used in both? Also, is Kia still using the mechanical disconnect for the front motor like they did in the 1st egmp vehicles?
Rear Motor
Plenty of equipment in the industrial industry that switches between Star and delta without stopping the motor , not sure why you would think otherwise, I bet the compressor in your workshop has a star delta starter in it.
Cool , innovation.
Excellent Intro !
Thanks!
3:30 My wind turbines (500 kW) of 1995 use this Y / delta switching between low speed/low power (Y) and high speed/high power (delta). 30+ year old.
what e great explantion
I’ve seen ev9’s managing less than 100kwh speed at Tesla superchargers, do you think by the time Kia gets access to Tesla superchargers next year they’ll be able to bump up the charge speed since superchargers are meant for 400v ev’s
German engineering baby!