Tripath patents list

[thomaseliot]

I found 37 patents assigned to Tripath Technologies Inc. on the new http://www.google.com/patents.

Justs ordered the list by patent number including a brief description.

I find interesting the US Pat. 6970123 - Inductorless architecture for a switching amplifier . Filed in 2004 I don't know if they ever used it. What do you think?

US Pat. 5777512 - Method and apparatus for oversampled, noise - shaping, mixed - signal processing
US Pat. 5909153 - Method and apparatus for compensating for delays in modulator loops
US Pat. 5974089 - Method and apparatus for performance improvement by qualifying pulses in an oversampled noise - shaping signal processor
US Pat. 6107844 - Methods and apparatus for reducing MOSFET body diode conduction in a half - bridge configuration
US Pat. 6127893 - Method and apparatus for controlling an audio signal level
US Pat. 6169681 - Power supply topology to reduce the effects of supply pumping
US Pat. 6229390 - Methods and apparatus for noise shaping a mixed signal power output
US Pat. 6246283 - Power efficient line driver
US Pat. 6281747 - Power efficient line driver
US Pat. 6316992 - DC offset calibration for a digital switching amplifier
US Pat. 6329876 - Noise reduction scheme for operational amplifiers
US Pat. 6348836 - Dual independently clocked analog - to - digital conversion for a digital power amplifier
US Pat. 6351184 - Dynamic switching frequency control for a digital switching amplifier
US Pat. 6362679 - Power device driver circuit
US Pat. 6362683 - Break - before - make distortion compensation for a digital amplifier
US Pat. 6411165 - Active common mode feedback for unbalanced input and feedback signals and methods thereof
US Pat. 6414560 - Loop delay compensation for a digital power amplifier
US Pat. 6515604 - Mixed signal processing unit with improved distortion and noise characteristics
US Pat. 6518849 - Dynamic delay compensation versus average switching frequency in a modulator loop and methods thereof
US Pat. 6549069 - Self - timed switching for a digital power amplifier
US Pat. 6566946 - Noise reduction scheme for operational amplifiers
US Pat. 6577189 - Scheme for reducing transmit - band noise floor and adjacent channel power with power backoff
US Pat. 6577194 - Resonant gate drive technique for a digital power amplifier
US Pat. 6580322 - Dynamic switching frequency control for a digital switching amplifier
US Pat. 6603355 - Active common mode feedback for unbalanced input and feedback signals and methods thereof
US Pat. 6617642 - Field effect transistor structure for driving inductive loads
US Pat. 6621339 - Methods and apparatus for facilitating negative feedback, providing loop stability, and improving amplifier efficiency
US Pat. 6628166 - RF communication system using an RF digital amplifier
US Pat. 6630899 - Scheme for maximizing efficiency of power amplifier under power backoff conditions
US Pat. 6693491 - Method and apparatus for controlling an audio signal level
US Pat. 6737713 - Substrate connection in an integrated power circuit
US Pat. 6781458 - Providing DC isolation in switching amplifiers
US Pat. 6785392 - Methods and apparatus for enabling or disabling an amplifier
US Pat. 6798288 - Receive band rejection for a digital RF amplifier
US Pat. 6940703 - Overvoltage protection circuit
US Pat. 6970123 - Inductorless architecture for a switching amplifier
US Pat. 7027501 - Methods and apparatus for adaptive equalization

Regards

Thomas

[Pano]

Originally posted by thomaseliot
I find interesting the US Pat. 6970123 - Inductorless architecture for a switching amplifier .

A number of other companies are doing this. Works well as long as the speaker leads are very short. Otherwise RF is broadcast all over the place.

Because the speaker sees only the difference between the square waves, no filter is needed as the difference is only AF. But if you send the high power RF down speaker lines, they will broadcast the RF noise. The air does not see the difference signal, it sees the RF going down the speaker wire.

[fokker]

basically, the speaker itself is an inductor.

the downside with this type of design is that it assumes certain "inductance" characteristics of the speakers the amp is designed for. this is not an issue for an OEM however it may be difficult for aftermarket service or diyer.

[Pano]

Look at patent # 6693491. Crikey! How can you patent that? In 2004????

[koolkid731]

Tripath's too many so-so patents is one of the reason they go under. Let's see: at say 20K each for filing and prosecution, 37*20K = 740K. That's for US applications only.

I have followed Tripath's patents from its beginning. IMHO they didn't have anything fundamental except maybe the "bit pattern alteration" for reduction of bit rate of the sigma-delta modulation.

Their many "garden varieties" of feedback loops were granted patent with no fundamental concept being introduced. Myriads of feedback loops are mentioned in old text books on feedback control text books.

[thomaseliot]

Originally posted by panomaniac
A number of other companies are doing this. Works well as long as the speaker leads are very short. Otherwise RF is broadcast all over the place.

Yes I know the way to use speakers to filter audio band. The problem is that they must bear the switching energy so you can do that only with small amps.

Tripath approach in patent 6970123 is different and can be applied to all amps:

http://www.google.com/patents?id=tRA...Ajnwpw1M&w=800

quite difficult to understand for me.

Originally posted by koolkid731
I have followed Tripath's patents from its beginning. IMHO they didn't have anything fundamental except maybe the "bit pattern alteration" for reduction of bit rate of the sigma-delta modulation.

Always asked myself what is their "Patented Digital Power Processing - DPP". Can you explain this "bit pattern alteration"?

[koolkid731]

Sigma delta modulation with noise shaping results in very high bit rate. So Tripathi's patent looks at the bit pattern say 10101010 at 1Mhz and changes it to 11001100, which translates into 500kHz output switching rate.

Potential problems with that are:

inherent delay in the closed loop feedback
audio artifacts that some audiophile may hear

IMHO there's nothing better than analog feedback of self oscillating amps. With analog feedback, 0.001% THD is doable. That's 60 times better than Tripath's sigma-delta, and no HF noise vs harsh sound of Tripath.

Indeed, although human can't actively hear f>20khz, we can FEEL the difference when a 20kHz brickwall LPF is inserted into a good amp. Cymbal's harmonics go past 150kHz. We rely on them for directivity of the sound (soundstage)

[koolkid731]

patent 6970123 is no different than other inductorless implementation of ternary modulation (3 states)

[Pano]

Originally posted by thomaseliot
Yes I know the way to use speakers to filter audio band. The problem is that they must bear the switching energy so you can do that only with small amps.

No, not really. As long as the two legs are switching in sync and at the same voltage, no current will flow. How could it? Both sides of the speaker are at the same voltage. So there is no current flow.

But if one side has a pulse that is slightly longer than the other, current would flow for that short time. The speaker still does not see the rest of the RF, because it is the same on both ends of the voice coil. It only sees a difference. In this case the difference is caused by the difference in the length of the pulses between the two sides.

If you use a switching amp where one side of the speaker is tied to ground, then of course the RF energy would flow thru the speaker coil - because it would be higher on one side than the other. That's why the inductorless designs are bridged.

[phase_accurate]

Their pataent that I like most is also the switching reduction one - although this could be achieved more easily by several methods.

I still don't understand how their first patent could be issued since it is 1.) obvious to me and 2.) has been discussed before they even filed.

But I don't think that they are having a bad time currently, or do they ?

Regards

Charles