Designing Cleaner Cooling Plants: A New VFD Topology for Data Centres

0:00 Catherine Giles: Hello. Good afternoon, and thank you, everyone for joining us today for the live webinar, Designing Cleaner Cooling Plants, a new topology for data centers brought to you by Smart Technologies. I’m Catherine Giles, digital media coordinator with Canadian Consulting Engineering, and I’ll be your moderator for today’s session. Before we get started, just a few housekeeping items. Today’s webinar is being recorded. A link to the recording will be emailed to all registrants twenty four hours after the session. The audience is in listen only mode with cameras and microphones turned off. We do encourage your participation during the presentation. If you have any questions, please submit them using the Q&A button found at the bottom of your Zoom window. If you don’t see it, you’ll need to click on the more dot, dot dot button first in a pop up will appear and you’ll see it there. We’ll do our best to address all questions during the Q&A session at the end of the presentation. There will also be a post-webinar survey that will pop up on your screen once the webinar ends. So please look out for this. We look forward to your feedback. Now I’d like to introduce your speaker. We’re pleased to be joined by Simon Leblond, CEO of Smart Technologies, a Montreal based designer and manufacturer of advanced power electronics for motor control and the company behind the Clean Power VFD. Smart drives our UL certified and deployed across North America through national distribution. Simon brings more than twenty years taking building automation and industrial control technologies from engineering through to certified field proven products, and today leads a team of over fifty. Please join me in welcoming Simon.

1:48 Simon Leblond: Thanks, Catherine, and my turn to welcome you to our Canadian Consulting Engineer sponsored session on a different approach to VFD topologies for data centers. As Catherine mentioned, I’m Simon, I’m the CEO of Smart Technology. Um, I will be presenting for roughly thirty five to forty minutes so that we have a full fifteen minutes for engineering questions. And let’s get started by introducing to you who we are at smart D. Uh, we’re a Montreal based manufacturers of VFD and no humility here. Uh, we’re on a mission to revolutionize motor control. Uh, we design, we assemble in Canada for the North American market. Uh, our products are ul certified CSA, uh, compliant for Canada, of course, and our manufacturing partner is ISO nine thousand and one and fourteen thousand and one at the factory. We also have a full ownership of our technology stack. So from everything, I’m going to show you silicon carbide topology firmware. The support is also in-house, and we’re available coast to coast through different national electrical distributors Netco, Rexel, to name a few. So, um, let me show you what, uh, revolutionized motor control means specifically for the data center cooling specifications. Our agenda will be, uh, pretty straightforward. We’re going to cover six topics. So, uh, why data center cooling is forcing the power quality conversation right now, where vfds live in the cooling plant and where and how that gets complex. We’ll have a as short as possible harmonics in IEEE five nineteen discussion. Uh, we’ll keep it to, uh, the minimum because I’m sure a lot of you have heard it before. We’ll talk about what kind of mitigation there is and what are the trade offs for those? What the different topology smart DDI introduces actually means, and what that means for you in terms of specification and single line. Um. At the end of it, um I’m going to show you how you can get your spec reviewed. So stay to the end and we’ll show you that free specification review offer. And before I jump into the meat of it, just, I am staying, uh, more at a high level. So no mathematical equations, no deep dive electrical, uh, really trying to make it a clean engineering argument. And we’ll see how our topology can really impact you on a practical, uh, way for data center and goal really is for you to understand by the end of the presentation what it means for your system. So to set context, you know, why are we having this conversation now? Um, I think, um, really there’s so much discussion around AI and data centers. It’s kind of self-evident. But really the force and function here is the IT load growth. Um, the compute is just scaling faster than the mechanical plant and that cools it. And so the cooling is where the energy goes, uh, thirty to forty percent of the total facility energy. It is the primary driver of the power usage effectiveness. So the p u e, that’s the total facility power over the IT power, um, a measurement of how effective the data center is. So, you know, to give you an idea of best in class hyperscalers run at about one point two PUE industry average around one point five. Um And what that implies in terms of money. If you have a ten megawatt IT facility at a p e of one point five, you’re drawing roughly five megawatt just for cooling. So four point four million dollars a year at ten cents a kilowatt. So definitely, um, makes it quite interesting to address variable loads with, uh, vfds and the opportunity to save money twenty and fifty percent less energy using those Vfds. Um, that’s, you know, roughly one and a half, two million dollars a year for that ten megawatt facility. That being said, it’s not because you’re saving money that you want to compromise on uptime. That remains your number one priority. And it’s how do you reconcile those two facts together that we’re going to be looking at in terms of landscape, uh, the market for vfds in data center has been growing. So last year, twenty twenty five, over twelve percent year over year growth. And we are seeing a trend to using low harmonic drives. So I don’t, you know, full disclosure, I don’t have, um, specific data of low harmonic drive in data center, but we are seeing that market grow with about eleven percent penetration, uh, in North America. And, and there’s really two things or two, um, structural driver here. Uh, one is a top down push from the grid, the utility, um, I triple e five nineteen compliance that we’re going to talk about. So you know, that top down down pressure and then the bottom up where, um, as a critical facility, you want to manage your harmonics, uh, ensure that harmonics are not impacting your uptime, you’re not generating more heat. So all in all, you have these two pushes coming together. Driving this extra adoption of low harmonic solution. Um, and that’s kind of the, the frame for the rest of the discussion. If we’re talking specifically of data centers and why, um, harmonics are important for data center. Um, we’re going to talk about it later. But the number one consequence of harmonics is heat. And so, um, as you have harmonics, uh, you’re generating more heat, uh, which requires more cooling, more vfds, more harmonics. And so you have this kind of, uh, heat cascade that you want to manage after that. And this is true for any facility, uh, a standard transformer de rates about twenty percent under a heavy harmonic load. So, uh, you’re going to have to upsize your transformer. Um, more cost, less efficiency, uh, due to harmonics and then more specific. And we will take some time on this during the presentation, uh, generator ups. So given the criticality of uptime for, um, data centers, you’re going to have, it’s not an edge case. You’re going to have ups and or generators to ensure uptime. And they also get impacted by harmonics. So they lose effectiveness. But they can also create an unstability. Um, because they’re a weak source. So, uh, finally, uh, we’re seeing it more and more with, uh, AC motors. So, uh, electronically commutated motors where they’re creating fan walls with a lot of these AC motors are typically unfiltered, although we are starting to see some low harmonic AC motors. Um, so they aggregate and they have a horrible, uh, harmonic signature and really, uh, they’re having quite an impact on the data center as well. So really, um, the headline here is that harmonics hit availability, not just efficiency for the data center. And really it applies across the cooling, uh, plant in the data center. You’ve got so many variable torque loads, uh, in there. Um, not necessarily that I want to go, uh, in details at every level, but whether you’re talking air handling units, chilled water pump, um, you really have a lot of opportunity to take advantage of the affinity law, which, uh, you know, I’m sure everyone knows about, but really cubic relationship between the energy you provide and the flow, that’s where your biggest energy savings come from and your dollar value. Um, so, you know, if you’ve got a rough number ten megawatt facility, you’re running thirty to fifty Vfds, um, anywhere between ten and sixty horsepower. Um, if you don’t have a harmonic strategy and you aggregate all of those harmonics together, your total harmonic distortion at the point of common coupling can reach twenty to forty percent. So you’re not compliant, but you’re also going to have impact in your facility. Um, the limit mandated by IEEE five nineteen is below five percent. So you’re, you know, you’re four to eight times higher than what you should be. Um, and that’s really why you need to think about it before the design, not as an afterthought. Now, um, you know, if we look at your list of priority in a data center. So obviously the first one is uptime and redundancy. Then you need to make sure that, uh, your, your motor control is, um, laid out and proper, uh, you need to make sure that everything’s maintainable and that your commissioning goes smoothly. But, and this is really the underlying challenge is the energy bill is the owner’s bottom line. So you need to put those vfds. And as I showed just before, those now make power quality and harmonic performance your problem and the thing you need to consider. So if we go into the heart of the problem here, um, first, you know, let’s all recognize that vfds have been growing and are performing superbly well because yeah, they, they Really offer large energy energy savings for all of these variable speed, uh, loads. So that’s right for them to put it. They also offer soft start. So you’ve got, uh, no electrical shock to your motor. You, uh, extend their life. Um, and you’re also having much better, um, airflow control. So all of these reasons make it and make vfds the right choice and your right to put them in the cooling plant. So it’s not, should we put vfds or not? Is does the traditional VFD, uh, offer the best architecture for the data center? Um, and so if we jump and look into that architecture, um, on the line side, what is actually causing harmonic is the rectifier. So six pulse diode bridge, which charges the DC bus by drawing current impulses from the grid, and you can see it on the right side of your screen. You’ve got these little humps about six per cycle, and between those humps there’s no current draw. So you really having a pulsing current that and if you decompose that, you’re getting integer multiple of the sixty hertz fundamental in this case, the fifth, the seventh, the eleventh, the thirteenth, and so on. And those harmonic are then flowing back towards the utility. Um, and for a typical, uh, VFD, uh, which has a DC choke, you’re going to have around thirty five or forty percent harmonic content. Um, some, you could have a VFD without a DC choke that would be over seventy percent, but I doubt any of you have installed those in a very long time. So, um, really that’s what’s causing the harmonics on the, um, on the grid and what those harmonics are. I mentioned it quickly. There are the distorted wave, the integer multiple of the sixty hertz fundamental. And when you sum them up, they come into this black, uh, distorted wave. And that’s how you see harmonics on the grid and the VFD because it’s, uh, drawing the current and these pulsing, um, waveform, that’s what we call the non-linear load. And you have that also with, uh, other examples would be like LED lights driver versus a linear load, like a resistive heater or incandescent fixture, where you can really see that the voltage in line current is drawn, um, symmetrically as a sine wave. Um from a, you know, your facility perspective, what’s the consequence of harmonic um the main one and by far is heat. Every time you have harmonic, this is energy that needs to be dissipated. Uh, that’s not usable. And so it really transforms transfer as heat, uh, whether it’s skin affected cable Eddy current losses in transformer. Um, those, uh, harmonics are being translated, uh, in heat in your facility. Just a key factor. Uh, twenty transformers sees roughly twenty times the stray losses heating of a standard unit. So really it counts. And that’s true regardless of if the harmonics are on the grid side or the load side. So a motor exposed to harmonic will also run hotter, uh, by having extra copper and iron losses. So it’s not just something that affects the grid. It really affects any of your equipment in your system exposed to those harmonics. Uh, you do also have other impact outside of heat. So you do have a nuisance impact. So breaker nuisance strip, uh, you can have a, um, a capacitor bank. If they’re present, they can resonate with the harmonic source and overheat. Um, but generally, you know, the main consequence, the one that you can always count on is heat. And then, you know, I mentioned it before the generators, um, there’s another impact because they um, have roughly three to four times the source impedance of a utility transformer. And so that higher impedance means that there is more voltage distortion from the same harmonic current. You know the Ohm’s law. So I promise no equations. So I’m just going to mention Ohm’s law. Um, but basically, uh, at that point with, uh, that generator, you, um, generally need to, uh, derate it and you have an instability risk, um, when you go on backup power. So and this rule will also apply to UPS system, uh, which also have these elevated impedance. Um, so I’ll talk about it in the, what, what the cycle is there in a moment. Now, you know, we’ve talked about the bottom up, uh, the, uh, impact from your equipment. Uh, but the top down is really this, um, the I triple E five nineteen standard, which is what you designed to in North America. Um, I’m not Uh, as I promised at the start. Uh, I’m not going to do a deep dive on this. Um, I’m just going to highlight a few quick things. Uh, first. You know, I triple E five nineteen is really assessed at the point of common coupling. So it’s not a device level, um, rule. Um, so even if you would have a drive without a DC choke that is generating seventy percent, uh, t h d um, you could still be five nineteen compliant if the rest of the system is clean enough. And the inverse, if you have a low harmonic drives everywhere, um, your system could still fail if you have other source of, um, of noise and your point of common coupling is small enough. So really when you think I triple e five nineteen harmonic study, you do need to think at the complete installation. Second thing I want to highlight is that the current version of IEEE five nineteen, which is twenty twenty two, only goes up to the fiftieth order of harmonic. Um, there is talk that future revision will go higher, but for now, there’s kind of this blind spot between the fiftieth order and EMI standards where anything goes. Um, and maybe the last thing I want to mention about IEEE five nineteen is without going details table, just if you want to think about practical target, you know, less than five percent current, uh, TDD and less than eight percent voltage THD. Um, so current usually is what you care about. Voltage distortion is what the utility cares about. Uh, but both are important, as we’ve discussed before. Now we’ve talked so far only about harmonics on the grid, uh, as I mentioned, they have the same impact on the motor and they do exist on the motor and they are caused by vfds on the motor. The mechanism is not exactly the same though. Um, in the case of the motor side, it is the igBT inverter that causes the harmonic and this is the switching. So typically a igBT inverter and VFD will be switching at two, four, maybe six kilohertz uh to produce a pulse width modulated waveform, uh or p w n. So it’s not a sine wave. You’ve got these, um, you can see it on the right side of my screen. You’ve got these voltage pulses and that pulse width modulated waveform for the current. And what that does is that the motor sees these uh rapid voltage transition. And this causes DV dt stresses. So that’s when you get into these reflected wave phenomenon for long cable runs. You also get common mode currents that flow through the bearing. Um, and so this is why you will get special rated cables for those long runs. You’ll put extra filters, you’ll get an inverter duty motors to face that extra um voltage stress. Um, you get shaft grounding rings and insulated bearings for the common mode currents. Um, but really here you’ve got one phenomena on the grid and one on the motor side and both really get addressed through, uh, mitigation. So, uh, you’re adding band aids to that problem through, um, filters on the grid, um, and adding filters on the motor motor side, expanding the size of your, uh, panel, adding potential, um, Uh, heat losses. And so it’s rarely just a drive. It’s a drive plus a kit of external part very often put into a panel, um, for one drive, one install. It’s manageable when you have thirty to fifty to manage into your data center. It can get, uh, it becomes a compounding problem really. So what does typical mitigation look like? Uh, if we go up the ladder of the solution, um, as I said, most modern, uh, vfds today will have a DC choke. So they’ll bring that seventy percent harmonic down to a thirty, forty percent. Um, you could also put an external line reactor on top of that, bring it down to twenty five or thirty percent. If you want, you can go to a passive filter, which will get you to the five percent mark. Um, but there is a risk that isn’t always discussed. So passive filters are tuned to a specific impedance and at specific loads. So as loads come on and off, uh, system impedance can change and the resonant frequency calibrated for the full load, uh, can become actually a resonant amplifier at partial load. So you’re potentially worsening the distortion at certain frequency rather than suppressing them. And I’ve actually seen this in some hospital settings, um, as you see some, uh, sweets coming on and off throughout the day. And that’s also a phenomenon you’re seeing in data center where, uh, your IT load will shift over the course of the day, but also over the course of the week. So, um, you know, think of a passive filter as a snapshot, a snapshot solution that’s really trying to address a moving target. If we continue up the mitigation ladder, um, twelve and eighteen pulse rectifiers, um, they enable you to reach ninety four percent THC. Uh, they’re quite effective, but they are large, they’re costly. And really importantly, right now they require special isolation transformers, which, you know, we know lead time on transformers are quite critical. So probably not a favorite solution. And then we come to active front end, uh, drive or active filters that will typically get you the best harmonic performance. And, um, we’ll talk about them, um, uh, in a second, but I’ll just highlight here that none of these harmonic solution actually address the motor side. So they’re really just that, uh, line side, uh, solution. If we talk about the active front end, they are getting more and more prevalent in the industry. So we are seeing more and more manufacturers roll them out. Uh, you’ll have varying degrees of performance. So some will state three percent, five percent. Uh, typically at full load. Um, which is not necessarily what you’re looking at, um, on your day to day, because if you were running at full load, you wouldn’t have a VFD. So, uh, important to also look at what’s the performance at different, uh, you know, at that fifty, seventy percent mark where you’re actually going to be running your VFD. Um, and, um, really given the prevalence of these offer, if you are addressing low harmonics in your data center, you’re probably already installing active front end drive in your data center today. So the question is really, what does the athlete leave on the table with the current technology? And there are three things I’m going to highlight here. Uh, we’ve talked about IEEE. IEEE five nineteen and how it measures only up to the fiftieth order. Um, and typically with an A F, you have these spikes and you can see it here on the capture, uh, beyond the fiftieth order clustered around the a f switching frequency. So these are hidden harmonics, um, that will not be captured by studies that are just looking at the first fiftieth order. That doesn’t mean they don’t exist. That doesn’t mean that they don’t produce heat and potentially other interactions. Uh, they’re just not captured by the current standard. Um, if you’re only concerned by the, uh, compliance to IEEE five nineteen. Then you don’t need to consider them for now until the next revision. If you’re concerned with the total harmonic distortion, then those are real. Now, second thing I did mention it harmonics also exists on the motor and all of the solution we’ve looked at. And that includes a phase only look at the grid side. Um, and you can see it’s interesting here, you can see that the motor side harmonic signature is much more, uh, dispersed because it’s really noise from that high frequency switching. Um, and so all of your AF are causing and causing that harmonics on the motor side. They’re causing the other issues I’ve talked about before. So the d v d t, the common mode current, and actually many of the active front end have higher common mode current than standard drives. So they’re actually potentially Amplifying the issue on the motor side of bearing flooding in early failures on that side. And finally, I mentioned generators as a weak source. It could be also UPS. Um, you know, the active front end is actively monitoring the grid and has a control loop to compensate those harmonics. Harmonics. And so, uh, when you have that weak slow source, the control loop can be too slow to damp the interaction with the generator. And so you can generate, uh, you can create an oscillation. Um, and it’s a potential reliability issue that you don’t necessarily think of because you’re designing for the grid, but really with the uptime front and center, that generator should be addressed, um, from the start. So, um, you do have a cleaner grid, but there are stuff that is left open and really, um, here’s where the clean power VFD topology shines because we address all of them. So from the motor side, um, there is no harmonics. Uh, and actually there’s no, uh, dv dt. There’s no common mode, uh, bearing because the motor really sees a pure sine um, directly. So both sine clean both sides, clean both the grid and the motor. Um you don’t have these higher order harmonics either. So uh, there’s nothing hidden and you’re already compliant to future revision of I triple eight five nineteen. So you don’t need to worry about that. And finally you don’t get destabilised by a weak source um such as a generator or a UPS How do we do this? Um, really at the core of it, um, I mean, we are an active front end, uh, with bidirectional, uh, inverters, but we’re using silicon carbide or sic. This is what you’re seeing on the screen. So silicon carbide MosFET are a new material that have been introduced about ten years ago, but they’re now present in pretty much every electric vehicle on the road. They’re also present in solar chargers and fast chargers for electric vehicles. And the reason why they’ve been adopted so massively by this, these industry, uh, is not only do they enable faster switching, but they also have higher power density and much lower losses. Um, so when you take these, uh, specification, um, you really get the potential for a whole new level of performance. And we’ve taken that and combined it with a multi-level, uh, architecture and with, um, from an algorithmic perspective, with harmonic cluster cancellation. So you put all of these, uh, silicon carbide, high switching frequency, the multi-level, uh, architecture and the, uh, harmonic cluster cancellation. And you get with a solution that has, um, two hundred times smaller filters directly built into, uh, the drive. So from your perspective, you’re getting a self-contained drive enclosure with no filter cabinet outside. So, you know, filter list is the shorthand here because in effect, we do have filters, but they’re just integrated in the product. And from your perspective, when you’re specifying, you don’t need to consider that. Um, and that’s really a symmetrical architecture. So it’s on both sides. We have those integrated filters, uh, with that technology. So, um, double stage filters and EMI filters on the grid and sine wave and EMI filters on the motor side. So as I mentioned, not only do you get the, uh, low harmonic on the grid, you also really get the pure sine wave to the motor. Uh, that removes entirely any, um, cable length limit, removes the, uh, common mode, um, uh, choke requirement or, uh, bearing flutter risk. And, you know, I know most of the data centers are new build, but, uh, you also get a great opportunity to retrofit a non inverter duty motors or standard cable, uh, applications. Um, all in all, also because the motor is seeing a sine wave and not a Pu, it’s running nine to twelve percent cooler. And that sine wave is true at any speed and load profile. So it’s not just like full load as you’d have when we’re specifying harmonics content. Um, you’re also getting an efficiency boost. Uh, not to say that you’re competing with the affinity law that remains your main, uh, efficiency driver, but as you remove those, um, harmonic impact both on the grid and the motor side, you’re, um, you’re removing that harmonic tax if you want. Um, and that harmonic tax is quite visible. If you put a filter to remove the harmonics, um, then it’s the filter that captures that lost efficiency. Um, but even if you don’t have that filter, Well, then it becomes the transformer. It becomes the motor that needs to dissipate that harmonics content and that becomes your efficiency drain. So really with smart D, the drive becomes the system and you get your efficiency right up front with no hidden losses. Um, I’ve talked about it. So, um, you get applicability of the solution both in, uh, new construction where you can simplify your specification by not requiring any external filters. So whether it’s an active filter on the grid side for harmonics or choke on the motor side. Just simplify one product, you know, you won’t have surprise. Um, but also great in retrofit scenarios. So, uh, if you have installed, for example, passive harmonic filters, and you’re getting these resonant issue and you need to retrofit because of the the problem they bring. Then you have a one self-contained drive that can really replace that whole filter cabinet and the drive cabinet, so much smaller footprint for these retrofit projects as well. Um. When you, uh, write the, uh, specification, so the document you’re working on, uh, really simplifies, you have less mitigation device to define and coordinate makes for a simpler bill of material, but also simpler install simpler commissioning. Um, and so the, the whole sequence from start to finish, uh, becomes much simplified because you only have one device to worry about. I’ll go quickly through this. I mean, really, you can use the Clean Power VFD anywhere where you have variable loads or even a constant torque load. It is a VFD as any other. From that perspective and usability perspective, where it really shine is that any place where. Of course you would have that mitigation, you don’t need it anymore. But also any place that would be constrained by cable length is suddenly something that you can reconsider. So you can rethink your architecture. Where, um, do I need my drive? Not where does the cable length dictate? I put my drive. So that opens up possibilities that you didn’t necessarily have or you were constraining yourself to based on traditional limitation. Um, and you know, as I mentioned quickly on retrofit. If you have an existing facility and for any reason you have places where you’re not running vfds and you maybe don’t have the right cables, um, type or you don’t have the right motor type, they’re not inverter duty. That becomes a easy energy efficiency win by plugging in the Clean Power VFD. And you’re ensuring that the motor really still sees a sine wave even with the variable speed control. And finally, um, you know, this is where with data centers you need to consider the generator, the UPS, uh, with the, um, being a really weak source. Um, because it’s a weak source. Your active front end doesn’t necessarily have the speed, the bandwidth in its control loop To hook on to that weaker, weaker source. So that really can lead to an interaction grow into an oscillation. And, you know, either simple derating or instability and then nuisance strip because we are running a control loop at one hundred and five kilohertz. Um, you’re really having much more bandwidth and you can maintain that stability in a much better way with these, uh, generators and ups than with a standard, um, active front end. So for data center design, this really becomes at the center of your design phase and your consideration because they’re not, um, edge case. Um, we have been deployed in, uh, real data center and, uh, meeting not only the low harmonic harmonic requirement requirement specified at the point of common coupling, but also eliminating the load factor to protect the motor and cables. So with that super simplified, um, uh, panel, as you can see, there’s almost nothing in that, uh, so less wiring, a smaller footprint footprint and much simplified versus the standard, uh, filter based approach. Um, and, you know, concretely, we met the spec, uh, we met the motor side compliance and had a very easy commissioning using our mobile app. So not the focus of today’s talk, but I’m going to plug that mobile app in there because it’s really enjoyed by the installers as they’re deploying this. Where does our product lineup, uh, stand today? So we have drives from fifteen to seventy five horsepower in both four hundred eighty volt and six hundred volt. We are extending that range in Q4. So, um, projects are open with the certification bodies. Um, talking of certification. So, you know, we certify our products to, uh, us and Canadian standard. Um, our products are typically rated for fifty K for thirty hp and less, one hundred K for forty hp and above. Um, could do higher with coordination studies. We have embedded uh, communication. So Modbus, TCP, Bacnet IP, Ethernet I, p are all native, so you don’t need any add on modules or hidden cost. Um, it is a Nema one wall mount package. Uh, so you don’t necessarily need a panel and they are rated up to fifty degrees C without Derating so you can go in quite hot environment and potentially lower your cooling requirement in that VFD room. And our standard warranty is eighteen months, um, with a potential to extend that if uh, the install is done by a certified partner. You know, I talk quickly about the, uh, Nema one just on the previous slide. So, uh, you know, this is an example of install where you really maximize your wall space. Uh, just put the drive directly on the wall. Um, remove the cabinet entirely. Really reduce your cost because you don’t have any filters to add around it. Um, you know, we do get the question quite often, you know, where are you install? So you can go to our website, we have, um, a page with the case studies on it. And um, we’re adding some, uh, every month almost. And then we also have, uh, lab validation. So we do publish, uh, on our downloads section, uh, our mtbf. And if you reach out to us, we can share with you, uh, accelerated aging tests. We also do on the drive, you know, we do design these products from the ground up for, uh, to be long lasting and being these kind of reliable, um, environments where you depend on them. Um, so forty three minutes and I’m going to speed up. I’m almost done. Um, really as you go through your spec, uh, go through the check and checklist of, you know, what are the loads that are variable speed? Uh, what are your harmonics limit specified at the PCC? What, um, you know, are you working in a retrofit or an existing situation or a new build with existing motor motors or are new. What are your cable length constraint and do I need some filters here? What kind of space do I have in my electrical room? So with that, let’s close with some of the typical objections I get and then the offer. Um, you know what? If you already use standard vfds. That’s good. Um, really, you can think of using the Clean Power VFD when you need filters. Uh, or you need long cable support or harmonic mitigation, uh, go where the complexity or the problem drives you. What about just adding filters? Well, that works. And, uh, it’s really a question of do you rather bolt on a filter or have everything embedded and not think about it later? Uh, sometimes going back on site and tuning the filter, making sure you have the right solution ends up costing more, uh, due to, labor and trips and servicing afterwards. Do you guarantee I triple e five nineteen? Well, really, that’s a misnomer because, uh you’re not guaranteeing I triple e five nineteen at the device level. And I still see a lot of specs that ask for a product that is triple a five nineteen compliant at the device level. This is a system level, um, uh, standard. So what about the ROI? Is it only energy? Well, of course, the energy saving from the VFD drives a lot of the decision. But in the case of the Clean Power VFD, you need to think about the full install cost, the life cycle cost. Uh, so of course, the filters, the space, but also, um, potential save downtime coming from that and should you install it everywhere? Well, I’d really like to say yes and specify it everywhere. Uh, but the reality is you can start by pushing it where the complexity is the highest and you get the best ROI. So the more filters, the more potential problems than the more we shine. Okay. And now in terms of conclusion, maybe just three things to take away here. Um, when we talk about harmonics and compliance two I triple e five nineteen. Really need to think about it at the system level, not at the device level. They do exist on both, uh, the motor side and the line side. So talking about harmonics only for the grid is actually, uh, cutting half of the equation out of the problem. And finally, you know, the silicon carbide based filter list topology that we’ve developed removes any kind of bolt on mitigation solution. So it gives you a simpler spec, easy retrofit and compliance that holds up in a generator and UPS situation as well. So that’s it for me. Um, I’d say don’t drown in harmonics. You can surf the wave with our Clean Power VFD. Um, here’s the offer for all of you. Completely free. Um, no sales call required to trigger it. You just email your spec to specs at smart dot tech. You can also scan the QR code. Um, and within two business days, we’ll analyze that spec and get it back to you as to where there’s IEEE five nineteen uh, potential risk. What are the filters or other mitigation device that could disappear? And, um, what exact wording change you need to, uh, make the spec either technology neutral or to call out the Clean Power VFD solution however you prefer it. So as we open it up to question and I’m sorry, we’re going to have twelve minutes of questions, not fifteen as I promised. Um, I’m going to leave on the, um. Uh, the QR code on the screen so you can scan it. And you also have the address.

47:33 Catherine Giles: Excellent. Thank you. Simon, I have a few questions here for you. Great. If anyone has any, please enter them in the Q&A box and I’ll read them out to Simon. So, uh, first is the drive compatible with all motor types such as Pmsm?

47:50 Simon Leblond: Uh, so the drive is designed to be compatible with all types of motors, but I need to be transparent that we don’t yet support the algorithm for permanent magnet. So we mostly do induction motors today. Um, this is something we plan to roll out later this year. So, uh, if you do have a pmsm application, definitely reach out to our team and we’re happy to look at it, but we won’t be able to address it right away.

48:16 Catherine Giles: Excellent. Another one here is is there any cable length limit?

48:23 Simon Leblond: Um, in theory, yes. In practice. We haven’t reached it yet. Um but I’d say um so if you go to um our spec sheet it says four point five kilometers. Uh, we’ve never done a deployment at four point five kilometers. Happy to try it out with anyone and give a money back guarantee on that one. So, um, you know, challenge us with, uh, with your cable length.

48:50 Catherine Giles: Okay. And another one here. Uh, why is the drive not producing higher order harmonics like other AFEs?

49:03 Simon Leblond: Um, okay, so, so this is really, um, related to, um, how, how we switch. So, um, and maybe I can go back to that slide. Um, so, uh, on here, you see a typical active front end harmonic harmonic cluster coming from the switching at six kilohertz. Um, I don’t have the exact slide for, uh, the hours, but it would be similar to this one on the right. Uh, where because we switch at much higher switching frequency, um, the, all the higher orders are filtered out by our integrated filters. So you’re basically putting your harmonics and your EMI in the same bandwidth and filtering them at the same time. So you get a really clean signature. Uh, not exactly that one. I don’t want to misrepresent, but basically absolutely nothing above the thirty fifth order.

50:11 Catherine Giles: Okay. Excellent. I think that’s all we have for questions here. Um, some came in through the Q amp A that your team was so, so great to be able to answer directly. So I think that’s the end of the webinar. And I’d like to thank Simon for sharing his expertise with us and smart D technologies for sponsoring today’s session. Oh, sorry. Here I lied. There is another one here, Simon. So could you touch base on the use of drive cable such as Nexen R?

50:44 Simon Leblond: Um, so I think those are, um, uh, reinforced cables. I don’t know exactly that, uh, that type, but typically, you know, there’s two things that, uh, you’re trying to address with, uh, reinforced cable. There’s the extra insulation due to the voltage spike in your PWM and that we remove entirely. So you don’t need any kind of. Of reinforced insulation. And you can have just standard insulation cable. Uh, because you don’t have those voltage spikes, you really have a sine wave. Um, now the other thing that, uh, sometime what you have with a shielded cable is, uh, EMI, um, suppression due to the filter cabling. And here, um, I can’t say we don’t do any EMI. We actually, we reduce greatly the EMI because we have these embedded EMI filters. Uh, but even if you have a contactor, um, a cable remains, uh, an antenna, so you’re always going to have some EMI effect. So depending on the, um, environment you’re in, you can have completely standard cable and you’re not going to have any issue if you’re in very, very sensitive environment for EMI, you might still want to put a EMI shielding on your cable. So hopefully I’ve addressed the question by going for both angles of VFD cables.

52:22 Catherine Giles: Excellent. And uh, so we’re talking another one came in. This one’s a bit long, so, uh, bear with me. Here in our area, some vfds are really affected by the voltage that is supplied by the utility. Most vfds are designed for five seventy five plus minus the standard tolerances, but the utility is sending at six hundred volts plus minus the standard tolerances. And there are instances where vfds will lock out without or sorry, with overvoltage. Would the smart DDI be vulnerable to this issue as well?

52:59 Simon Leblond: Well, I can’t say we can withstand any voltage. Um, we are definitely design at six hundred, so our standard specification is six hundred plus ten percent. So we’re good up to six hundred and sixty volt minus fifteen percent. So down to five fifteen. So that does give you quite a big range here in which you can operate. Um, so I don’t, and typically to, uh, we do because it’s an active front end, we do correct for, uh, the grid distortion. Um, if you have specific use case, happy to look at it. You know, one thing I mentioned quickly at the start is because we are in Canada, if you call the factory for support, you’re getting the engineers that design the drive. So it’s very quick for us to get you with the active front end grid control expert. And we can even simulate your use case and tell you, will we withstand it or will we have an issue? Um, so not part of the, of the offer, but if you want to write at spec at smart dot tech, uh, we’ll take care of you. No problem.

54:17 Catherine Giles: Excellent. Thank you. Simon. Um, I think that’s all. Is there any more? Please put them in the Q&A. Okay. Oh. Someone’s here. Just giving you a thank you so much for the presentation. Uh, so unfortunately, that is the end of this, uh, the, this session. So thank you so much, Simon, for, for talking to us today and SmartD technologies for sponsoring. Thank you all for joining us. And please keep an eye out for the survey that will pop up on your screen shortly and watch your inboxes for a link to the recording from today’s session. I hope you all had a great day and uh, enjoy the rest of your day. Thank you.

55:01 Simon Leblond: Thanks, Catherine. Thanks, everyone. All right.