Under the DUT – Electric Vehicle (EV) Power-train – Ep.4 Part 3

Under the DUT – Electric Vehicle (EV) Power-train – Ep.4 Part 3


hi it’s under the DUT and I’m Ken Hyatt we were looking at the e V power train now we’re gonna dive a little bit deeper into some of the system components and some of the testing challenges that we have Shashank what do we have setup here and we talked about how important testing is for this industry right and we have a battery charger here but we’ll we’ll talk a little more about what goes into testing right to pick that example of the battery charger battery charger essentially is an AC to DC converter right like any other power supply that we’ve seen in the past right and the circuit diagram looks quite similar as an AC to DC on a DC DC stage what’s different about automotive is really the performance angle of it and the power density they are really trying to push it in terms of performance and power density as we talked about to get there you really want to get the most efficient design as you can right what that means at the circuit level is you want to time those switches very very accurately you want to make sure your gate drive circuit is performing the best way possible and you want to make sure that the losses in all these components are you know very very low right what that also means is you want to try and measure all the switches simultaneously if you can especially for that deck time measurement know which is been right it’s very important in the past you know what has been kind of well understood and easy is the low side measurement right so if you look at this circuit the low side measurements are usually well understood because they’re ground reference right and you can easily make those measurements right what’s been really tough is the high side measurement and what’s even tougher for automotive is the high switching frequencies as we talked about right so the high voltage in presence of high switching frequency and a floating voltage extremely difficult we use differential probes or probing techniques to do it go ahead and do that what is the what is the reason why it’s different today the problem is this proves this differential problems were never designed for very high frequency of operation all right most of this probe exactly right so it’s not just high voltage and the floating voltage when you add high switching frequency rode it all big strong right think about this just to get the leads on this thing it’s crazy how long those leads are right and if we know RF if you know high frequency anything that’s that long no matter how much you twist them they’re going to act as an anthem yeah we’re just gonna pick up the latest edition so whenever you have a a circuit that switching really fast it just picks up all this noise and all you see is junk right so you don’t really know when the switch is actually turning on if the ringing is real if those oscillations are real or not right and what that does is adds a design time right because you don’t know if this is real and you’re trying to fix it right so engineers end up or designing their designs to compensate for this problems they end up spending more time chasing problems that are probably not even there what we want to do is run a measure exactly my point right so you if you really want to get those design to be optimized to the exact level you knot or design them adding cost right you really want a system that is very very accurately telling you exactly where the switches are turning on they’re turning off and the losses they’re having right if you’ve been working quite a bit with this engineer’s to solve this specific problem and that’s when we essentially came up with this isolated probing system right and you’ll notice very quickly right so we can’t read all the leads first of all right so the leaves are gone this is a nice shielded cable to actually take care of that problem so that nothing gets picked up right down to the connector right so you want connected to be nicely shielded really nice tight so nothing gets out of that signal and radiates this isolation part of it right most of this probes were electrically isolated meaning that they were still attached to a scope there is a parasitic sand inductance really here we actually got an optical cable so you have complete isolation from the scope which essentially means a truly floating system right and that gives you the performance you need for this high switching frequency systems so we really minimize the RF interference here at that and you can actually reduce all the common mode voltage right the floating voltage we talked about so you can really cancel that out right so you have a very high common mode voltage and a really good bandwidth response when something like this you know if you take the same measurement with ISO view probe again you’ll see the difference you can actually see the gate measurements and the Miller plateau right that’s pretty incredible what you can do that’s awesome yeah so that’s real interesting yeah and really significant what other challenges are we going to be facing I think the other one that I I think is still kind of partly solved is the current measurements current measurements are really really tough and I’ll give you an example so people have been trying to measure current on the drain and the lower side of highest switching frequency converters for a while and usually they use a loop right so they’ll put a loser how do you do it under something as small that’s that’s one part of it because one is really small the second part of it is the bandwidth right the switching price is so high this becomes an antenna right so you’re adding inductance to your circuit and that might end up blowing the circuit that is the extreme effect right and but the measurement sucks right because you’re trying to use the current probe it doesn’t have the bandwidth and this just adds more problems to your circuit right so this doesn’t work anymore for most cases right a good option they have left is with shut we started experimenting with this and as you can see you know shunt measurements are probably the most accurate you can get out of high frequency measurement and you essentially use a probe like this to measure the shunt as well so you can get the really low voltages across the shunt the drawback is the cost you know the put a shunt in right and then generally you’re adding inductance and stuff right so there is there is trade-offs as always but this is really the only method we have seen which works really well for that way so isolated probing system takes care of high frequency does it matter what kind of scope we have what really helps is having high resolution on a scope right and I’ll tell you why right so think of some of the measurements one of the very critical measurements people make is the RDS on measurement right there the on-time resistance measurement and what’s what’s interesting about that measurement is you’re trying to measure a really small voltage in presence of a high voltage right so it’s when the VD has the drain voltage actually turns off right so you have a 400 volts and then you are trying to measure millivolts in presence of 400 volts right so if you think about the vertical resolution you need it’s like pixels right how many pixels do I need to actually see details from that millivolts and that’s where high resolution really helps right so for example a scope like this which is a 12 bit scope helps quite a bit in terms of getting that resolution you need for another example is a vgs measurement the gate voltage is 5 volts 10 volts 15 volts in presence of VDS which can be 600 to 800 watts away so that’s where it really helps to have a high resolution at the end of the day we are trying to get the highest efficiency possible right and that’s where the component losses comes in that we talked about earlier right so it helps to have a system a complete system which essentially helps you measure the component losses so you know for example the power analysis package on this it gives you you know inductance losses magnetic losses component conduction as well as switching losses right so that that helps you quite a bit right it’s really about getting to market faster making your design process easier right taking inaccuracies out of it and making sure you’re really getting what you’re looking for right so that’s really what it is thank you so much for spending so much time with us today great explanation on the e V power train and diving into the details here we really appreciate you viewing this too and we’d like to get your comments and questions so please go ahead and join the conversation on our social channels and remember under the DUT is for engineers by engineers [Music]


Leave a Reply

Your email address will not be published. Required fields are marked *