Why big guys do not like the TPA chips?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am very curious why this chip is not popular among big amp makers. I have heard a lot of amps and although this is not the best sounding chip/amp among all type of amps including Class A, TPA sounds darn close to amps costing 50 times more expensive ones.

I ASSUME if they make amps with this chip, how can they sell if for $5K? That's the reason why they don't make it. I see no other reason.
 
True enough, I was thinking about the older TPA3116 series of devices.

I can think of another reason beyond your suggestion - if they were going to pitch classD chips to the audiophile market they'd need to have a cracking good story for why their amp beats the pants off a single digit $ amp PCB from eBay and they've not got the creative chops to come up with that.

Another fly in the ointment of your argument is that expensive amps don't sound hugely different from cheaper ones anyway so your 'darn close' isn't really going to be close enough. Which particular TPA implementation were you listening to incidentally?
 
Disabled Account
Joined 2002
And is that a stock TPA3116 board or something modified/tweaked? I listened to TPA3116 going back a couple of years now and while I felt it was impressive for the price/size it didn't beat a humble TDA1521-based chipamp.

That's a coincidence. I was going through old projects today and found quite a few unused TDA1521 chips and PCB's I designed myself for TDA1521. At that time I really liked it. Maybe I should try it again. TDA2616 is the same chip but with mute feature.
 
The TDA1521 had input transformers - used for isolation and for phase splitting to get them bridged.

I've not had much success at all with TDA1545, the ones I've built into DACs have a weird 'fuzziness' on certain music. TDA1387 doesn't have this problem though.

Never heard an LM4780, nor done a side-by-side between TDA1521 and LM3886 but I prefer the TDA for its ease of decoupling - the power pins are close together which lends itself to placing a lot of 1206 10uF ceramics between them to get a clean HF.
 
Output trafo on class D is nothing different to an output trafo on any other kind of amp - its there to transform the impedance of the speaker into something more optimized for the electronics part of the amp. I use step-down transformers on the TDA8932 not to improve the bass (which is already awesome) but get better dynamics overall and match the capabilities of the chip to the load. The bass does improve slightly though as a result of better dynamics.

The output transformer does improve the overall quality and reduces the maximum output power of the chip so that it can run at higher supply rails without overheating. But the biggest improvement in SQ comes from much better decoupling on the analog rail (pin8). The DS just shows 10R and 100nF feeding this pin which is woefully inadequate at filtering the noise from the main supply rail.
 
Boutique amp builders can get away with higher prices because of the ignorance of the consumer. Also at an absolute level, the chip amp performance can deteriorate appreciably higher up the spectrum. Hypex modules apparently do well in this regard.

Some manufacturers also focus on marketing technical nonsense. An example is how femto clock jitter is going to transform your system and you are going to open the wallet to the tune of $80,000 for a DAC.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.