I'm designing a class D amp similar to the UcD type of amp. It is self-oscillating and uses a portion of the audio output after the HF filter fed back to the input comparator. The ideas came from the original AES paper by Bruno Putszeys and from a Philips (NXP) application note UM10155 from 2006.
Since I am using HV gate drive ICs, I can choose whether I use 1 main positive supply voltage, like some of Tripath amps, or 2 main supplies, one plus and one negative, like the Hypex amps, at the same voltage. In either case, the output will be a full bridge with (4) MosFETS and the speaker output will be bridged, which means one side won't be tied to ground. Yes, there will be (2) sets of HF filters on the output to the speaker. I do realize that this is not the cheapest output implementation. The main advantage to a full bridge output is higher output power, as I see it.
Other than being able to make more audio power from the dual supply, is there any advantage to one approach over the other approach? I see the "pumping" of the dual supply version as a negative point against it. However, The full bridge version should not have this problem, I think, because you are drawing the same current from both supplies at the same time. The single supply is an easier supply to build and if you buy a cheap (Meanwell??) SMPS, you save money by only having to buy one of them. I guess if I can achieve the required power with the single supply version then it would be the preferred solution. Anybody else have any thoughts on the subject??
Since I am using HV gate drive ICs, I can choose whether I use 1 main positive supply voltage, like some of Tripath amps, or 2 main supplies, one plus and one negative, like the Hypex amps, at the same voltage. In either case, the output will be a full bridge with (4) MosFETS and the speaker output will be bridged, which means one side won't be tied to ground. Yes, there will be (2) sets of HF filters on the output to the speaker. I do realize that this is not the cheapest output implementation. The main advantage to a full bridge output is higher output power, as I see it.
Other than being able to make more audio power from the dual supply, is there any advantage to one approach over the other approach? I see the "pumping" of the dual supply version as a negative point against it. However, The full bridge version should not have this problem, I think, because you are drawing the same current from both supplies at the same time. The single supply is an easier supply to build and if you buy a cheap (Meanwell??) SMPS, you save money by only having to buy one of them. I guess if I can achieve the required power with the single supply version then it would be the preferred solution. Anybody else have any thoughts on the subject??
There's no advantage to a split rail for bridged outputs. Assuming all N output devices, your driver signal would be referenced to the negative rail anyways, so it makes perfect sense to have a single supply. The only argument for a split rail is the analog input ground reference would not need a decoupling cap. However, most designers don't realize that the power supply caps act as the decoupling cap too, so you might as well just do the single supply and spec a good quality cap for the decoupler.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.