DSPRelated.com
Forums

Are there any LMS IC's?

Started by Tom August 25, 2004
I saw almost 15 years ago tha there was to be an LMS IC available but it
never materialised. I was wondering if there are any readily available
Least-mean Squares ICs on the market. There must be many custom ones of
course and many papers on the subject.


Tom


Tom wrote:

> I saw almost 15 years ago tha there was to be an LMS IC available but it > never materialised. I was wondering if there are any readily available > Least-mean Squares ICs on the market. There must be many custom ones of > course and many papers on the subject. > > > Tom > >
Yes, they'll require a bit of configuring for your application though. The most popular ones are made by Xilinx and Altera, although Atmel and Actel also make them. And yes I'm being a smart-ass, and yes I mean FPGA's. I doubt that a purpose-built LMS IC would be commercially viable given the current state of the art in FPGA's and ASIC's. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com
> Tom wrote: > > > I saw almost 15 years ago tha there was to be an LMS > > IC available but it never materialised. I was wondering if there > > are any readily available Least-mean Squares ICs on the > > market. There must be many custom ones of > > course and many papers on the subject.
I agree, I had one at HP Labs in 1981 and later, for an application at some 200 Megasamples/second, prototypes at a major firm, detailed in the ISSCC 1996 digest and one or two patents. These were moreover mixed analog-digital implementations so that they did not require the power consumption of textbook (digital) versions and they operated efficiently on analog input data (even if the ultimate objective was extraction of digital information). On the other hand, software was nonstandard ... (By the way, the 1996 version had a flexible _continuous-time_ discrete-tap analog FIR filter. Historically, the original FIR filters operated in continuous time, as technology-history students know.) Cheers -- Max
Max Hauser wrote:

> ... Historically, the original FIR filters operated in > continuous time, as technology-history students know.)
Some of them, mostly just for show, had astoundingly long impulse responses. I remember a quarter-second audio delay line (which could be tapped but wasn't when I saw it) at the 1939 Worlds Fair in New York. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
There are some Motorola DSPs that have a hardware coprocessor for adaptive
filtering:
http://www.freescale.com/files/dsp/doc/app_note/APR39.pdf

"Tom" <somebody@knowherex.netgx> wrote in message
news:1093488713.669585@ftpsrv1...
> I saw almost 15 years ago tha there was to be an LMS IC available but it > never materialised. I was wondering if there are any readily available > Least-mean Squares ICs on the market. There must be many custom ones of > course and many papers on the subject. > > > Tom > >
Tim Wescott <tim@wescottnospamdesign.com> wrote in news:10iqn3abb7dvfe1
@corp.supernews.com:

> Tom wrote: > >> I saw almost 15 years ago tha there was to be an LMS IC available but
it
>> never materialised. I was wondering if there are any readily available >> Least-mean Squares ICs on the market. There must be many custom ones
of
>> course and many papers on the subject. >> >> >> Tom >> >> > Yes, they'll require a bit of configuring for your application though. > > The most popular ones are made by Xilinx and Altera, although Atmel and > Actel also make them. > > And yes I'm being a smart-ass, and yes I mean FPGA's. I doubt that a > purpose-built LMS IC would be commercially viable given the current > state of the art in FPGA's and ASIC's. >
They are also available from Analog Devices, TI, Motorola, etc. Of course, these are usually known as DSPs (and a little code). -- Al Clark Danville Signal Processing, Inc. -------------------------------------------------------------------- Purveyors of Fine DSP Hardware and other Cool Stuff Available at http://www.danvillesignal.com
"Jerry Avins" in news:412df52c$0$19716$61fed72c@news.rcn.com...
> Max Hauser wrote: > > > ... Historically, the original FIR filters operated in > > continuous time, as technology-history students know.) > > Some of them, mostly just for show, had astoundingly > long impulse responses. I remember a quarter-second audio > delay line (which could be tapped but wasn't when I saw it) > at the 1939 Worlds Fair in New York.
Good stuff, Jerry! (At a few kHz, that would still be a time-bandwidth product around 1000, not bad.) Do you actually recall the 1939 fair, trylon, perisphere, and all? (I am only old enough to have seen the 1964.) A popular 1939 exhibit was said to be the General Motors Futurama, envisioning a far-off world of 1960. "In high-rise cities, inhabitants would be suffused with culture." From the same era is an encyclopedia I have, picturing an artist's conception of a "bridge of 1960" spanning distant mountains. That summer of 1939 was a pregnant moment in the history of the western world. While the nations showed off at New York, British society threw parties, and von Ribbentrop and Molotov put the final touches on secret protocols for dismembering eastern Europe. Before the summer was done, Poland faced invasion from both sides and the British public heard a world-weary, terminally-ill Prime Minister on the radio: "I have to tell you now ..." -- Max
Max Hauser wrote:

> "Jerry Avins" in news:412df52c$0$19716$61fed72c@news.rcn.com... > >>Max Hauser wrote: >> >> >>> ... Historically, the original FIR filters operated in >>>continuous time, as technology-history students know.) >> >>Some of them, mostly just for show, had astoundingly >>long impulse responses. I remember a quarter-second audio >>delay line (which could be tapped but wasn't when I saw it) >>at the 1939 Worlds Fair in New York. > > > Good stuff, Jerry! (At a few kHz, that would still be a time-bandwidth > product around 1000, not bad.) > > Do you actually recall the 1939 fair, trylon, perisphere, and all? (I am > only old enough to have seen the 1964.) A popular 1939 exhibit was said to > be the General Motors Futurama, envisioning a far-off world of 1960. "In > high-rise cities, inhabitants would be suffused with culture." From the > same era is an encyclopedia I have, picturing an artist's conception of a > "bridge of 1960" spanning distant mountains. > > That summer of 1939 was a pregnant moment in the history of the western > world. While the nations showed off at New York, British society threw > parties, and von Ribbentrop and Molotov put the final touches on secret > protocols for dismembering eastern Europe. Before the summer was done, > Poland faced invasion from both sides and the British public heard a > world-weary, terminally-ill Prime Minister on the radio: "I have to tell you > now ..." > > -- Max
My Father helped build some of the fairground buildings, so he knew the grounds quite well when he took me around. (He was a foreman with an HVAC firm.) We visited several times. Most of what I remember probably happened in 1940 (the fair had a two-year run). One of the highlights of the GM exhibit was the audio that accompanied the diorama. It described the cars in it as "going around curves at 50 miles an hour." Astounding! The talking lumbering robot at the GM exhibit didn't impress me, and the cut-away view of a turning-over V-eight engine in the Ford exhibit showed up my lack of tact when I piped out, "You see, Daddy: I was right about the firing order." At nearly 8 years old, I should have been more sensitive. One of the more impressive exhibits was Westinghouse's. It included "artificial lightning", A pair of large Van der Graf generators throwing 20-foot sparks through air. Somewhere, I still have a green Heinz pickle with a brass safety pin embedded in it, a souvenir of the Heinz exhibit. Inside the Perisphere, there was a circular moving walk with a view of the diorama it encircled. The Trylon wasn't open to the public, but my father got us in. There's more to see inside the Washington Monument. Oh yes: the audio delay was at the Bell Telephone exhibit. The trick was to put on earphones that played the output of the line (Bell was good at making artificial lines for test sets) while talking into a microphone connected to the input. At a quarter second delay, it was merely disconcerting, but coherent speech was nearly impossible at what my hazy memory estimates at around 50 ms. I saw myself on television at RCA's exhibit: 441 lines, image orthicon. The man behind the camera explained what was in the box when I asked him. When all my questions had been graciously answered and a little more filled in, I thought I really understood, (trapezoid correction whose need I had deduced) and all. That my father had the patience to stand there through it all still boggles my mind. Jerry -- Engineering is the art of making what you want from things you can get. &#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;
Al Clark wrote:

> Tim Wescott <tim@wescottnospamdesign.com> wrote in news:10iqn3abb7dvfe1 > @corp.supernews.com: > > >>Tom wrote: >> >> >>>I saw almost 15 years ago tha there was to be an LMS IC available but > > it > >>>never materialised. I was wondering if there are any readily available >>>Least-mean Squares ICs on the market. There must be many custom ones > > of > >>>course and many papers on the subject. >>> >>> >>>Tom >>> >>> >> >>Yes, they'll require a bit of configuring for your application though. >> >>The most popular ones are made by Xilinx and Altera, although Atmel and >>Actel also make them. >> >>And yes I'm being a smart-ass, and yes I mean FPGA's. I doubt that a >>purpose-built LMS IC would be commercially viable given the current >>state of the art in FPGA's and ASIC's. >> > > > They are also available from Analog Devices, TI, Motorola, etc. Of > course, these are usually known as DSPs (and a little code). >
Y'know, I was assuming that if one is looking for a purpose-built chip one would have data going too fast for a plain-ol' DSP. But I should have tossed it in anyway. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com