The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: firstname.lastname@example.org (David)
Subject: Re: Help needed - H-bridge, ClassD output stage IC
Date: 16 Jan 2003 21:53:51 -0800
NNTP-Posting-Date: 17 Jan 2003 05:53:52 GMT
> but be prepared to put more hours into it than the parts salesman or the
> thesis writer is likely to admit are needed.
You are right on that...It took a year to understand what the thesis
writer was describing in two simple pages!
I understand that not using feedback after the bridge can cause a lot
of problem. However, I really don't see how feedback could be
implemented. In the fpga, the quantizer output is feed back to the
input and is also used to control the bridge. If I understand what you
say, the output of the bridge should be fed back into the fpga instead
of the quantizer output to take into account the analog imperfections.
How could this be possible?
First, the voltages are not 'fpga compatible' anymore. This is not too
complicated to solve though.
Second, the fpga works only on rising clock edges, so even if the '1'
coming out of the bridge is fed back instead of the quantizer output,
the fpga would always see the previous value because of delay caused
by the bridge.
Maybe I'm totally wrong here and I'd really like to have more details
on what you meant.
Thank you very much
email@example.com (R.Legg) wrote in message news:...
> > I was planning to do the same thing as TI since my modulator is very
> > stable. I ran a lot of tests in matlab and I don't really see where it
> > could go wrong. However, I'm not really sure this is what you meant
> > since you said is is 'somewhat naive' to you. Do you mean that the
> > output stage would require feedback for proper operation?
> If the bridge timing and recovery delays were insignifigant in
> generating the intended volt-seconds, to be averaged by the output
> If the pwm was always used the full supply rail voltage, so that
> energy recovery did not modify output volt-seconds,
> If the output filter components were linear under large signal
> If the load impedance were resistive and linear with respect to the
> applied signal,
> If the noise generated by the switches were sufficiently partitioned
> so as not to affect any other circuitry (or test equipment)in the
> neighbourhood in an unpredictable manner,
> Then you could expect the output to reflect the intended result of the
> digital modulator.
> TI is confident that their evaluation circuit is capable of achieving
> this, under specified conditions and using the recommended parts,
> equipment and test methods. Some commercial manufacturers will sell
> you something in a box, which you must assume does what it claims,
> over the range of temperatures indicated.
> I wonder what it takes to guarantee that a breadboard will do the same
> thing, after it's been made to fit somebody's idea of a practical
> production line device?
> Considering the distortion generated by most audio transducers,
> differences may be irrelevant to the end user and it is hardly likely
> to be used in critical applications, without testing.
> The application of digital power modulation has it's advantages and
> it's disadvantages. Use it where it makes sense, but be prepared to
> put more hours into it than the parts salesman or the thesis writer is
> likely to admit are needed.
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup