Class D amplifier scalable in power from 25w to 1200w

Hello everyone, my name is Johan, in this document I want to present my project, which is a class D amplifier with only N channel mosfets, scalable from 25Wrms (or less) to 1250Wrms) ...! - This time, we are taking advantage of the fact that at least, here in Argentina, the IR2110 is being obtained at a good price, and that makes it possible not to go crazy with discrete mosfet drivers, and to use only N-channel mosfets ... step, the amp goes at one!

Let's clarify that this is a proof of concept in which many of the desirable characteristics in high power amplifiers have been omitted, such as protections and fuses that will be left to the reader. The main objective of this project is to bring a Class D amplifier closer to those who do not have experience with this type of circuitry and one of the most important rules in its design was to simplify it as much as possible and use only common components, low cost and easy to install. get for everyone.

I have specified IRFP250 as mosfets ... They are not the only possible ones ... For lower powers, much cheaper N-channel mosfets could be used. For example, if the supply voltage will never exceed +/- 20 volts, you could use the IRFZ44, or for 200Wrms over 4/8 ohms, the IRF640 ... Simply, smaller mosfets will heat more.

In relation to the heatsink, really, you have to put it ... And the reason is very simple: Although this amplifier has a theoretical efficiency of 97%, at 1250Wrms, that implies 37W of dissipation in the mosfets, so they require a heatsink. However, for example, for 200Wrms output, the power dissipated in the mosfets will be 6W in each one, so a few simple clips could work well. Simply, place an aluminum foil as a heatsink ... And remember that this time, both mosfets have to be isolated from the heatsink with micas and insulating washers ... But the TIP MUST NOT be isolated, because that transistor puts the heatsink to ground to suppress noise emission

The last issue I have left to name is the output inductor issue: It must be able to withstand the peak output current of the amplifier without saturating or burning out. Personally, I use 1mm² of section for every 4 ampere of current (that is, for 8A, I use a wire with 2mm² of section) ... Preferably, instead of using a single wire, use 2 or 3 in parallel whose section added give the required wire section. This will improve the performance of the inductor (I would use, to obtain an equivalent section of 2mm², 3 wires of 0.66mm² section put in parallel)

PD1: For those who want to improve the sound even more, I recommend lowering the gate resistors as much as possible, but ... carefully ... Because if they are too low, the mosfets can start to heat up or directly burn. You have to lower the values ​​of both resistors slowly, and test if it heats up or not, using a current-limited source, and with the amplifier without load and without an audio signal.

PD2: If possible, I would like to make this song a "Featured" ... There are very few super high fidelity class D amplifiers posted, and I really think this project could be a very good one for posterity, and you don't get lost in the midst of the huge number of songs that are opening up from standard class AB amplifiers or onboard, with the TDA series, which although they are really useful for beginners, do not add anything to the design technique or have nothing innovative ... Anyway, I hope it happens!: rolleyes.

PD3: Thank you very much to all the people who have been talking to me lately about this topic of the class D amplifier ... Many of the things discussed were taken into account for this design ... I especially name ricardodeni, who was one of the first in trying the combination IR2110 + LM311 !! --- And there are very many more people, who to a greater or lesser extent, have collaborated .. Sorry if I do not name you all, but thank you for all the collaboration that you have selflessly given!

PD4: Yes, technically, changing the mosfets for ones that are even more powerful, and with more voltage, it would be possible to easily exceed 1250Wrms. The limit of the IR2110, is +/- 250 power supply, and even, using an IR2113 (compatible pin to pin), it could be taken to + -300v. Of course, it would be necessary to change the 2n5401 transistors for mpsa92, and recalculate the resistors that dissipate power so that they do not burn out. Neither the plate nor the diagram itself would vary. Thus, as a curiosity, with that voltage of +/- 300v, over 4 ohms, 11250Wrms could be obtained ... Of course, I assume that anyone who starts to modify the values ​​to obtain such a high power knows what he is getting into, And do not ask me for advice to reach those values, because I consider them too dangerous for someone amateur to start playing with it, not to mention the source of power that would be needed!

without more to say, I say goodbye, thank you very much for your attention.