EMI Denial
I headed over to the TI site on audio class D they have a "beta" forum. I see a lot of people complaining about the noise. The responses of the TI employees is almost hilarious.
Complaint too much noise Ans. solution turn it up
its still there Ans. its supposed to be that way you'll never hear it
http://e2e.ti.com/forums/t/7560.aspx
Berates a customer and no technical help
http://e2e.ti.com/forums/t/7705.aspx
I headed over to the TI site on audio class D they have a "beta" forum. I see a lot of people complaining about the noise. The responses of the TI employees is almost hilarious.
Complaint too much noise Ans. solution turn it up
its still there Ans. its supposed to be that way you'll never hear it
http://e2e.ti.com/forums/t/7560.aspx
Berates a customer and no technical help
http://e2e.ti.com/forums/t/7705.aspx
I'd just say that all TI forum "audio apps engineers" are limited full stop. Their EVM designs supposedly meet commercial EMI requirements according to their data sheets. Which I doubt since reading the TI forums and looking at the schematics. They need some analog design consultants because signal processing experts and newbee digital engineers are not real worldly experienced enough. They really don't have a clue, so there is massive denial permeating through the whole TI organization. read this for a sample of organizational denial http://e2e.ti.com/forums/t/1767.aspx
Rotel may have the same thing going if they keep their engineers seperate in the product line orginzation.
infinia said:
True enough. And, sadly, I've heard Class D "beat noise" in finished products--even at CES demos and in other situations where it's more than a little embarrassing to the manufacture. The spread spectrum clocking I mentioned earlier in this thread is often considered a cure for beat problems, but in reality it just alters the nature of the noise to something that's completely different but can still be audibly annoying.
Designing high-end Class D products is tough because it adds several challenges over more conventional amps. In battery powered portable audio products, and other lower end applications, Class D is fairly plug-and-play. But making it work at levels suitable for high-end products is tricky--especially for something like a 5 or 7 channel home theater amp where beat issues, and overall noise, are multiplied.
And TI is a Giant Public Company that is run like most Giant Public Companies these days which is to say it's all about short term quarter-to-quarter growth and profit and little else. And, to make things worse, the lion's share of the net profits go into the pockets of a very select few. A lowly engineer being told he can't buy any new equipment, the department can't hire anyone, and having his already meager salary frozen due to budget cuts gets to watch the senior execs taking home several million each, getting huge bonuses, etc. It's no wonder they don't seem to care much about doing a good job!
The worst part, to me, is Giant Public Companies like TI buy up the small, nimble, really innovative, well run, employees-who-like-their-job, kind of companies. And it usually doesn't take long before the "Giant Borg" kills just about every ounce of innovation and enthusiasm among the employees of the former small company. It's really sad.
The problem is defined and history repeats once again
Taken from an old Fairchild CMOS application note.
Test Fixtures vs. Real Systems
Because ground bounce is so dependent upon the load the
device is driving, it has proven to be one characteristic of
CMOS devices that does not correlate well between results
taken on standard test fixtures and results seen in actual
systems. This occurs for several reasons. First, the AC
loading presented by standard text fixtures is not the same
as the AC loading generated by a system load, and second,
the standard test load creates a LCR tank circuit that
tends to oscillate during edge transitions.
For these reasons, ground bounce data taken on test fixtures
is useful for comparative analysis, but is not valid for
predicting actual system performance
Taken from an old Fairchild CMOS application note.
Test Fixtures vs. Real Systems
Because ground bounce is so dependent upon the load the
device is driving, it has proven to be one characteristic of
CMOS devices that does not correlate well between results
taken on standard test fixtures and results seen in actual
systems. This occurs for several reasons. First, the AC
loading presented by standard text fixtures is not the same
as the AC loading generated by a system load, and second,
the standard test load creates a LCR tank circuit that
tends to oscillate during edge transitions.
For these reasons, ground bounce data taken on test fixtures
is useful for comparative analysis, but is not valid for
predicting actual system performance
I had a recent ICEpower(ed) alpine car ampifier that wiped out AM completely. It might be the nature of the beast. This may be as quiet as they were able to make it. Unfortunately the Alpine also let its own power supply noise come through the speaker outputs - ensuring a short installation period in my car.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.