As the title of this article points out, in terms of functionality, sometimes all you need is a transistor. This is true enough, and the TPL7407LA from TI provides this functionality. However, I don’t want you to get the impression that this IC is just seven MOSFETs. Even if it were, you would still save board space compared to a discrete implementation, but the device offers various benefits precisely because it incorporates more than merely the output transistors.
Diagram taken from the datasheet.
As you can see, each channel has overvoltage protection, specialized gate-drive and regulation circuitry, and a flyback diode. And of course a transistor—more specifically, an N-channel MOSFET. The idea here is that each channel provides the basic functionality of a FET switch but with additional features that eliminate complications and external components that would be involved if you were using discrete transistors.
I have to admit that I like this approach. There is something satisfying about driving things with ordinary transistors instead of highly integrated, multifunction ICs. It is also true, though, that I’d rather not divert time and mental energy to the irksome details associated with optimizing a low-side driver.
So with the TPL7407LA and other similar ICs, you get an old-fashioned NMOS transistor with additional features that simplify implementation and improve performance. And of course it could be argued that devices of this kind provide not just nostalgic benefits but also flexibility. As the datasheet points out, the TPL7407LA can be used in various applications—motors, relays, LEDs, and so forth. So you have one device that might find a place in numerous projects. Another aspect of this flexibility is that you can use just one IC in a project involving different types of loads; for example:
Diagram taken from the datasheet.
The TPL7407LA is described as a pin-to-pin replacement for Darlington arrays. This is good to remember if you have designs based on “old-fashioned” Darlington parts and want a simple method of upgrading to a higher-performance solution. According to TI, the key advantage offered by the TPL7407LA compared to a Darlington array is power dissipation.
When the output transistor is active and sourcing current, some nonzero voltage is present between the output pin and ground. In the case of a MOSFET, this is the drain-to-source voltage, but more generally it can be referred to as VOL (“output low voltage”). Power, as always, is current times voltage, so the power dissipation of the drive transistor is directly proportional to VOL. TI claims that the TPL7407LA can be much more efficient than a Darlington array because its typical VOL is much lower, though the datasheet implies that this benefit is more pronounced at lower load currents (i.e., less than 250 mA; the max current rating is 600 mA). The following plot shows VOL vs. current, but I don’t know how this compares to a Darlington implementation.
Plot taken from the datasheet.
At first glance 600 mA might seem a bit restrictive, but in reality it’s a minor issue, because the outputs can be paralleled in order to increase the max current. You can see this in the example circuit given above. If you have to control seven loads and they each need 1500 mA, yes, you’re out of luck. But many applications will not require all seven outputs, and overall, parts like this one offer pleasant versatility in terms of maximum current. As far as I can tell you could parallel all the outputs and drive a 4 A load, or you can control seven low-current loads, or anything in between.
You might recall the drive and regulation circuitry mentioned toward the beginning of the article. What’s going on there? Well, the TPL7407LA accepts a wide range of input (i.e., control) voltages. The extra circuitry ensures not only that the lower voltages turn on the output FET but also that the FET will be adequately enhanced.
If you’ve read this article, you know that higher gate-to-source voltage is needed to achieve lower on-state resistance. The TPL7407LA is designed to allow a logic-level device (e.g., a microcontroller) to switch high-current loads, and it accepts input voltages as low as 1.8 V. The “new” (according to the datasheet) regulation and drive circuitry ensures that lower control voltages do not result in degraded switch performance.
Do you have any thoughts on the FET vs. Darlington issue? Feel free to share your thoughts in a comment.