TI Training home
Industrial TI Tech Days 2020
Industrial TI Tech Days - Power management sessions Increasing system robustness with integrated protection and diagnostics using TI high-side switches
Industrial TI Tech Days 2020
1.5 Increasing system robustness with integrated protection and diagnostics using TI high-side switches
OK, so I see the event is being recorded. So I'll take that as the good to go. Welcome to the TI Industrial Tech Day. This session is going to be increasing system robustness with integrated production and diagnostics by using TI High-side switches.
Thanks for joining virtually this year instead of in person. This is going to be presented by Shreyas Dmello. My name is Timothy Logan. I'm the applications manager for the power switches a group here in Dallas.
All participants, I just want to everyone a heads up. Everyone's going to be muted for the session. If you have a question to ask about the topic at hand, please use the chat function to ask it to address everyone. Also, if you're having any problems with video or audio, please feel free to use that chat, also, and we'll get someone to help you out with that. But with that, I'll hand it off to Shreyas to go ahead and get started.
Hey, everyone. Welcome to the Industrial Tech Day. I hope you all can hear me, but thank you for joining this afternoon. Today, we're going to be talking about system robustness and how TI High-side switches can improve that.
A quick introduction about myself. My name is Shreyas Dmello. I am the application engineer responsible for industrial High-side switches, load switches and power maxes here in the power switches group at TI.
So really quickly, our agenda for this tech talk is going to go over what a High-side switch is, and then we're going to talk about what differentiates a [INAUDIBLE] High-side switch from a quote unquote "smart" High-side switch. We're going to talk about these circuits in which these devices can be used, and why you would want to use them there in these circuits.
Rather than talking about a specific application since the industrial spaces is very varied and complex, it's not particularly fair that they got one application, and point to it, and say, we fit here. So that's why you should use it. But rather, we're going to talk about common design challenges faced by multiple customers in multiple applications, especially in the industrial space, and see how High-side switches, especially TI High-side switches, can fit the bill.
So when I say High-side switches, a quick overview would be to talk about the various points on a board, where a High-side switch could go compared to the variety of offerings that the power switch has. A quick way of-- there's a lot of content over here. But basically, it boils down to a High-side switch is those switches that are used to connect off board loads.
There's nothing much more to it that differentiates it from a competitive load switch or an ideal diode, except for the fact that it was built with the express purpose of driving off board loads. In the [INAUDIBLE], I would say that a good proportion of all loads are in some way considered off board. It's not necessarily possible that every load will have its own DCDC source or its own power board right next to it.
So when we're talking about loads, like solenoids and motors, even if they are either two meters away, this presents unique challenges that the switch has to survive that we will get into in time. So quickly, a High-side switch is just a device that goes between the power source and the load. This is not to say that low side switches are irrelevant or wrong to use. Because they definitely have their place, but the High-side switch is definitely beneficial in the case where a mechanical failure, like, let's say, dropping a wrench can occur.
Now, when you do this in a system with a low side switch, they're always going to be that risk of having a floating node. And therefore, a failure could inadvertently complete the circuit. On the other hand, a High-side switch won't complete a circuit, because there is no floating node. Simply, because we're disconnecting the source from the node rather than disconnecting the ground from the load.
Now, when we say smart High-side switch, there's not a lot of difference between a smart High-side switch and a High-side switch, except for the fact that this is more of a holistic product instead of a simple switch. We're talking about integrating not only the switch itself, but also, diagnostic features and protection features, as well as some sort of information and providing information to the users all into one holistic package.
So where would you use a smart High-side switch? As I mentioned, these devices are not going to be in a socket that is very different from where a normal High-side switch goes. In fact, you would put it exactly where you would put a High-side switch between a supply and the off board load. The difference comes in when we're talking about those specific loads that require a lot of power protection.
This is because we are-- as I mentioned earlier, we are integrating those protective features, like current limit, and diagnostic features, like current sense, into one singular device. Any system that powers an off board load runs a risk of a load failure, such as a short circuit event or a load current overload. This is just a factor of the long [INAUDIBLE] that connects the power supplies or the power board to the load. So the possibility of having the short circuit events or load current events is pretty high.
Another thing to take care of is because these are off board loads, the inductance of the wire itself is enough to [INAUDIBLE] for a lot of switches. Simply, because we are pushing a lot of current, in some cases, through a long wire. So inside the industrial space to summarize pretty succinctly, a lot of loads are going to be considered off board in some form or another. Either, because they're inductive in nature, or because they can't be very close to the power board.
But whether you're talking about factory automation, or motor drives, or building an automation, everything together can benefit from using smart High-side switches. We have devices that are built for the lower power output, as well as devices that are built for higher power. But what unifies them is the fact that all of these loads will benefit from the protection features, as well as the diagnostic features of this device.
Now, we come to the meat of the presentation. I think this is where we're going to spend most of the time while we're here, because we've gone through what is a switch and why you would want to use a High-side switch. We're going to talk about the specific design challenges that arise, especially in the industrial space that can be solved by using smart High-side switches.
Overarchingly, there are three major applications or load challenges that a High-side switch can solve. And we're going to go through each one of them in detail. The first one that I want to mention is load driving, and I say that this is load driving. Because we're talking about inductive load driving, as well as capacitor load driving.
Now, it is not possible to think of any load as a pure resistance. That just doesn't happen. It's something that we've learned in school. It's something that we've learned in college. But as soon as we start connecting any load with any [INAUDIBLE], [? ideologies, ?] like line inductance, come into play.
Now, when we're talking about inductive load driving, the issue does not arise when you're turning on a system. In fact, in some cases, the inductance can help slow the inrush current that flows into the load. But the inductor wants to refuse any change of current that flows through it.
The issue arises whenever you're turning off a switch. The inductor is going to hold the amount of current through it as much as it can, and this will cause a massive negative voltage spike on the output of the switch. Now, how a High-side switch helps over here is by integrating a voltage clamp that clamps the maximum voltage across the drain and the source of the FET.
These devices are all integrated NMOS devices, and by clamping the maximum voltage across the drain and the source, we do not allow an unsupported amount of voltage drop across the FET during the off state. This not only helps the device sustain its operation during the turn off events. But depending on the load, it helps dissipate the energy by turning on the MOSFET and providing a bath for this energy to either dissipate through the resistance of the MOSFET or through the rest of the load.
Of course, this is going to result in the MOSFET heating up. There's no free lunch over here, of course, and what that results in is a maximum amount of inductive energy that a certain switch can [INAUDIBLE]. I do want to reiterate that we're talking about energy over here.
So while we might talk about what switch can handle what inductance, do note that in some cases, a smaller inductor being driven at a higher current may have more inductive energy in it than a larger inductor being driven at a very, very low current. There is a maximum amount of energy that the switch can handle, and this is spect in all High-side switches datasheets. The next load that I want to talk about is another frequent load found in industrial systems, and that is [INAUDIBLE].
Compared to the inductive load, the issue arises during turn on and not turn off over here. During turn on, because of the effective resistance of a capacitor during turn on, an infinite amount of current-- it might even trigger a short circuit fault in some conditions. But an infinite amount of current is going to rush through the switch to fill up that capacitor. Now, this is not always going to be a concern for a lot of different applications, because some applications may be OK with handling this inrush current.
But in power limited systems, especially where the power source is extremely power limited perhaps by a backup battery or a very tightly specked DC/DC source, this could cause unintentional supply droop. Now, this supply droop could, in best case, trigger some sort of under voltage fault in the system. But in some cases, I have definitely seen this resulted in damaging the DC/DC source. Simply, because it could not supply the power requested during this turn on phase.
So how the High-side switches help over here is by holding the amount of current that can flow through it constant. This is done by having a maximum amount of current flow through the device known as the current limit. Now, most devices have some sort of internal card limit that is not a [INAUDIBLE] and usually has a wide range, let's say, between seven amps and 13 amps. But High-side switches from TI have a user set current limit.
This means that the user can set the current limit to whatever they want and hold it at that. So that the supply can fill up a capacitance, can be held from unintentionally drooping, and safely turn on the system. I'll get into how a user sets their current limit when we talk about protections. But do note that just like how there is a maximum amount of inductance that a switch can handle, all switches are going to have a maximum amount of capacitance that they can fill.
This is because when we are holding the current at a specific amount, this is done by modulating the on distance of the FET, causing the device to heat up. Eventually, the device will speed up enough that it triggers thermal shut down and turns off. So there could be a situation, where a very, very large capacitance could not be filled completely on the first try when you turn on the device.
This is a good segue into the next application that I want to talk about, and that is very much in protection for short circuits, as well as current limits and overcurrent events. A short circuit event is similar to an inductance or into a capacitive inrush event. Now, we define a short circuit as low impedance path to ground, but that's not necessarily always going to be the case.
Sometimes, there are multiple rails within an industrial system. Some having even negative voltage, if permissible, in the system, and there could be an event, where there is a [INAUDIBLE] shock that occurs. In all situations, this is a typical customer requirement that we don't cause any serious damage either to the load or to the device itself. More so than anything, if a supply is limited, this could unintentionally damage the supply to a point at [INAUDIBLE] recovery may not be possible, such as the case of a broken DC/DC source.
Now, how we get into this is interesting, but I want to talk about this by using a specific example. On the left over here, I have a very basic digital input output module. [INAUDIBLE] threshold for this module is one amp. Now, this has some sort of internal trigger to it.
But because you can have a user set current limit, this trigger does not have to be hit before a protection takes place in the system. Let's take an example of a one amp digital input output module. In this situation, at 24 volts, the supply power has to be upwards of 24 watts.
This is not always possible in this case. If we were to use a TI High-side switch with adjustable current limit at, let's say, 0.75, the nominal DC current is what we have to design for rather than designing for the short circuit threshold. If we were to clamp the amount of current that can flow through the system at 0.75 amps, all of a sudden, your loads, your components, your PCBs do not have to be rated to handle that 24 watts or that one amp.
Instead, it can be lower rated, because the High-side switch is going to hold the maximum amount of current that flows through the system lower than the short circuit threshold of some of these devices. In a short circuit event, a similar methodology occurs. The High-side switch is going to hold the current at this level, until the event is complete or until the device heats up and turns off, thereby, protecting the load and all the other components without necessarily requiring that these components be rated to very, very high currents and very, very high wattage ratings.
So when I talk about user set current limit, I'm not necessarily lying to you guys or anything like that. I really do mean users set current limit. By using an external resistor to ground, in most cases, the user can generate a reference current from the device, and this voltage is then used within the device to trigger current limit or not. That's it, one resistance either from the device to ground in most cases or in a few cases from the supply to the device.
It is one equation that is defined in the data sheet, and we have a multitude of High-side switches, where an entire family that are meant to drive different sets of loads. One device might be built for a higher set current limit, while the other might be better for a lower set. And this is fairly proportional to the on resistance of the device.
But in all cases, it is just one resistance from device to ground. Finally, I want to talk about the last application or last challenge that is faced by customers. And that is information, and I say information. Because I'm talking about fault detection, as well as agnostics within the system.
In any engineering system, information is key. If we know more about the system at hand, then we're going to be able to respond to any sort of event better and faster. By increasing the information provided to the user, we're increasing the safety and the reliability of that system itself.
Eventually, we're trying to build a situation, where our outputs are driven smartly. This helps, because it reduces system downtime since the diagnostics and the protection features tell the user what has gone wrong in the event of a failure. High-side switches integrate a variety of different sense options, like load current sense and open load detection functionality, to help meet these requirements.
Now, just like the user set current limit, the user can just use one resistor from the device to ground to set a analog signal, an analog voltage. Now, the resistance that you use over here is heavily dependent on what microcontroller or what ADC you are using. Because all of that will be handled external to the device. But at the end of the day, you can choose whether to see load current.
They can choose to see temperature, or they can choose to see the voltage that is being imparted onto the switch. And this variability is integral to a lot of different applications. Because in some applications, they may be very thermally sensitive. So the temperature sense might be what is necessary, rather than a load current sense.
We use one output, so all of these outputs are [? mixed ?] through one pin. So that the user has to make the choice whether they would want to see current, or temperature, or supply voltage. Still, values are output as an analog current on the SNS pin. And in the event of a fault, either a fault high voltages is output on the SNS pin, or a fault current is pushed through the SNS pin, resulting in a noticeable change from the linear region of operation.
Thus, the controller is able to distinguish between a linear region of operation, as well as a fault operation regime. The next diagnostic that's incredibly important in the industrial space is open load detection. High-side switches TI offer open load detection in both the on state, as well as the off state.
In both situations, the mechanism by which this is working is very similar. Since an open load would lead to a floating node in the system, figuring out when an open load occurs is incredibly important to the user. The device is going to be probing the input voltage and is going to be probing the output voltage at all times. And this is in on state and in the off state, assuming that the device is powered correctly.
If the difference between the input and output voltage is below the open load threshold-- and this is assuming that no load is there. So there's no current flowing through the device, and the device is on, meaning that there is going to be very, very little voltage drop across the FET itself, seeing as how these are very low on our own devices. The device is going to recognize this as an open load false.
In the off state, there's not going to be any path by which the input voltage can be passed to the output. So what we do instead is have a large pull up resistor, usually, around the range of 15 to 20 kiloohms that are not going to necessarily affect linear operation of the load, but what it's going to allow is a pass through for voltage during the open load state from the input to the output. Once the device notes that the open load exists, it passes this fault out through the SNS pin or through the fault pin, depending on the device upstream to the controller or to an MCU.
This sort of instantaneous feedback, both in the on state and in the off state, can help prevent any sort of issues that arise with having a floating four volt rail, or in some cases, 36 volt rail in the system. Now, I'd like to conclude this presentation by talking about this white paper that is available on ti.com. As I mentioned earlier on in the presentation, it is not possible to point to specific applications and say, because it works over here, it will work everywhere.
Rather, it is important to look at what the design challenges are faced by multiple customers in the industrial space and figure out how you are to respond to them. This white paper does just that. We're going to be talking about common design challenges, like current sensing, or open load detection, or inductive load driving, and see how TI power switches can help respond to those situations effectively. At the end of the day, what we are building and what we are offering to these systems is safe, robust, protected operation. Thank you. I'm going to open the floor to any questions now.
Yes, and I posted in the chat too. However, please just any questions that you have, put them in the chat, and then I'll relay them. And we can go back in the presentation if there's anything we needed clarification on or we're just curious about.
A little bit of a silent crowd today. We can hear your dog, Shreyas. So I guess I'll just leave it open here for a few-- oh, here we go. So any recommendations for low are on, but 200 milliamp configurable current limit? Shreyas, are you there?
Oh, I think I lost connectivity. I think it stopped sharing.
OK, I assume you guys can hear me. So for the 200 milliamp range, we do have devices that can have user set current limit. One of them is TPS 1H100. This is the automotive device, but we have an industrial spin of that called TPS 27S100. This is a low R ON device of about 100 milliohms or so, and it can meet the 200 milliamp current limit spec.
Actually, another device that I would say for this is a four channel device. TPS 274160, I believe, can meet a low current limit like this as well. Once we get into the very high current limiting settings-- oh, sorry. Once we get into the very low current limit settings below 100 milliamps, we're going to be working in higher and higher R ONs.
Because with the technology that these devices have, accuracy is something that matters. So if we have a very, very low R ON device, it is going to be impossible to hit those local limits. So I would say, definitely look at the high R ON offerings that we have going up from 100 milliohms to even one ohm.
OK, so the next question is, what is the max current of these devices?
That is an interesting question, because it depends on how you're going to be talking or what exactly you mean by max current. Most of these devices are specked for a maximum DC nominal current. But in the event of small spikes, I would say, we do have some that are rated than their normal current.
A good example of that would be TPS 27S100. This is a nominal four amp device, but it can handle small spikes higher than that. On the lower end of things, we have devices that have on resistance's as low as eight milliohms. And for this device, the nominal DC current through it is going to be in the range of 10 amps.
So it really depends on what exactly you are looking for, because there is a device in our portfolio that can be used in your system one way or another. Once we start getting into higher requirements, like 15 amps or higher than that, then we are going to be unable to fit that much current within that integrated system. And at that point, we're talking about using external power effects to handle that sort of current capability.
All right, cool. I will make one last call for questions here on the chat. I'll give everyone a couple of minutes just to see if anyone's typing anything out. All right, we got one. So what is the main difference between a High-side switch and E fuse?
That is a good question. An E fuse and a High-side switch are very, very similar devices, except that an E fuse has more integrated protection features that is built for input power protection. So let me see if I can scroll back up to there on this slide.
So an E fuse would be used in protecting your power board as such, while a High-side switch would be used when you're protecting your off board load or your power from an off board load. So an E fuse has a lot more controllability with features, such as flow rate control. That is not something that a High-side switch takes into account. We only do current limiting, but these are similar devices. They cannot be interchanged, though, because you cannot use an E fuse to drive an inductive load since the E fuse does not integrate any sort of clamp within it.
OK, last call for questions if there are any. And if not, one thing that I do want to just advertise here while I have the chance is, if you have any questions at all, we have a really nice online support forum called E to E. I encourage everyone to go to etoe.ti.com.
And if you post something on that forum, the question essentially ends up directly in our inbox. So just, if you have any sort of technical questions, any sort of just questions about the different products that we do offer, that is a really good way just to get a fast response directly from our team here. Thanks, Shreyas, for presenting, and thank you everyone for joining us.
Hopefully, the next time that we do this, it will actually be in person, so that we can be a little bit more personable here. Just as a reminder here, all the session recording and presentation will be available to view next week. You should get a link for that.
You'll receive an email with links for that on demand presentation and the post-event survey just to kind of let us know how we did if there's anything that you would want to change or see for the next one around. We really want your feedback. I mean, we really think these tech days and training events are going to be super important.
So if there's anything at all that you have as far as some feedback, A, if you want to be more technical, if you want to cover different aspects, if you want to go into just certain product lines, please, please, please be as candid as you can in those surveys. And we'll try our best to go towards that, and thank you again. I mean, just have a great rest of your day.
November 5, 2020
TI’s family of industrial high-side switches meets these requirements by offering various protection features like accurate current limiting and diagnostic features like accurate current sensing within the device. This presentation will discuss the many integrated features of industrial high-side switches and the advantages and benefits they provide to various off-board industrial loads.