Vladimir Vassilevsky wrote:> > Dirk Bell wrote: > >>I worked for a while at a major defense contractor that was working with 10 >>year old DSP code. No one left understood the DSP code, they just got >>follow-on contracts to make more copies run and to improve the GUI. They >>should have made the DSP code work right. Amazingly, since the customer was >>dumb and happy, the contractor was not real interested in fixing the code to >>improve performance or even in bringing it to the customer's attention. >>What a shame. >> > > > That is life. Once ago I had the experience similar to yours. I tried to > change the things and I only learned three lessons out of it: > > 1. Nobody cares what is inside as long as the overall product > performance is satisfactory. > 2. If the damn thing works somehow, leave it alone. > 3. You are not going to gain much for yourself by trying to improve > somebody's lousy piece of work. > > Vladimir Vassilevsky > > DSP and Mixed Signal Design Consultant > > http://www.abvolt.comActually, succeeding can hurt. While I was in personnel filling out papers on the first day in a new job, my boss, his right-hand man, and the owner were conferring about running out of ROM room with feature creep. A new feature had put them over the top in an 8-K chip, going to 16 K was an unhappy thought, and there it was. My boss had squeezed out all he could, but the code was still six bytes too big. He proposed eliminating a seldom-used feature, but that was vetoed. Modifying the board to take a bigger ROM was expensive; lots of units were in the field. The owner made an unpopular decision: "There's someone new coming on board. Let him have a go at the code before we decide." I knew none of the background. My new boss simply asked me to shrink the code as much as I could. I was new to the 8031, but it's a simple if ugly machine. By the end of the day, it was comfortable, and I wrote a thread tracker -- what in Forth would be called the inner interpreter -- that worked, then proceeded to factor the code. I had to get a feel for it overall, so nothing much good happened at first (which pleased my boss, but he didn't say so). Then it all came in a rush. After five full working days, I had enough to report progress and ask for guidance on how hard to compress; I had reached what I thought was the point of diminishing returns. (We had this conversation in the owner's office. I learned later that my boss wanted him to hear first hand how little could be done.) I was told that there were at least 200 bytes to be shoehorned in if possible, and congratulated on my having found the six bytes needed. I reminded my boss that he had instructed me to shrink the code as much as possible. There was now, even with the new feature, just short of 1 K to spare. I didn't get along with the boss after that. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
AM demodulation
Started by ●October 7, 2003
Reply by ●October 10, 20032003-10-10
Reply by ●October 11, 20032003-10-11
Jerry Avins <jya@ieee.org> wrote in message news:<bm748b$c59$1@bob.news.rcn.net>...> Actually, succeeding can hurt.It sure can. For reasons unclear to me, I apparently had gained a reputation as some sort of "signal processing wizard" amongst the people around me when I wrote that PhD thesis of mine. As I have commented upon before, I never did anything I don't expect any other DSP engineer to be able to if given similar time and resources. Anyway, after my dissertation I have not been able to find a decent, workable project. All those more or less weird things, based more on wishful thinking than anything else, seem to come my way. "If anyone can do it, Rune can." I am not sure if I ever heard it spoken out, but it seems to be the general attitude. It's not fun when I have to tell people that I'm only an engineer who uses the rules of maths and physics, and not some sort of sorcerer. Rune
Reply by ●October 12, 20032003-10-12
Rune Allnor wrote:> ... It's not fun when I have to tell > people that I'm only an engineer who uses the rules of maths and physics, > and not some sort of sorcerer. > > RuneBut aren't those the same thing? Many think so! Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●October 12, 20032003-10-12
Jerry Avins <jya@ieee.org> wrote in message news:<bmamiu$e2d$3@bob.news.rcn.net>...> Rune Allnor wrote: > > > ... It's not fun when I have to tell > > people that I'm only an engineer who uses the rules of maths and physics, > > and not some sort of sorcerer. > > > > Rune > > But aren't those the same thing? Many think so!Unfortunately, you're right. The scary thing is *where* I find these kinds of thoughts: - The 40+ years (of experience) senior scientist who wanted to specify "60 degrees phase lag between the 20 Hz component and the 80 Hz component" in a steady-state multiple-lines narrow-band source. I managed to wave that one off. - The 10+ years (of experience) seismic processor who found the sound velocity in the bedrock *immediately below* a high-velocity geologic layer by means of refraction seismics. A trivial property of such analysis is that the sound speed must be strictly *increasing* with depth. I found a possible explanation for this measurement which, if correct, could be the key to understand the acoustics of that whole geographical region. - The head of the Sig Proc department at an international research facility who asked me why I bothered with all that high-res analysis when a simple zero-padding + FFT of my data would resolve the features I was looking for. I was very glad *not* to be offered a job on that facility after that meeting. - The internationally renowned scientist who did not understand why I was so conserned about symmetry properties in the 2D DFT since "it's only the positive frequencies that go into the inverse transform". I was very cautious in how I read his journal articles after that. - The scientist who was overly concerned about not being able to make computational models that were able to generate as strong Rayleigh waves (ground roll) as measured in real life, when analyzing the muzzle blast of heavy artillery. I knew the Rayleigh waves well enough to see other possible excitation mechanisms in that experimental setup. No basic text on Rayleigh waves fail to mention these excitation mechanisms. - The 40+ years (of experience) scientist who asked for my array processing codes for analyzing arrays with logarithmic element spacing. "If you can't give it out, I'll just pick equidistant elements in the array and use standard FFTs". Anyone are free to try to pick an equidistant subsequence of three or more elements out of a series that goes as D_n = D_0*K^(n-1) n=1,2, 3,... where K =/= 1. All these are different people at different facilities, but make mistakes I find utterly trivial. "Why bother?" you may ask, "Everybody make mistakes. It's not unheard of that even a certain Rune comes with corrections to his own posts on comp.dsp". Well, it's a consistent pattern: People who really should know better make the most trivial mistakes. The real problem, though, is that they evidently are not open to neither accept the mistakes or correct them. I can only guess, but what I think I may be doing differently that these people (that may explain my ridicilous "wizard" reputation), is paying attention to detail. To not take anything for granted, not accept something as "easy" or "irrelevant" until I've cheked things out. To try to think the problem through. To try to find out what others have tried and how those attempts went. But that's what basic engineering training is all about, isn't it? Rune
Reply by ●October 12, 20032003-10-12
Jerry Avins <jya@ieee.org> wrote in message news:<bmamiu$e2d$3@bob.news.rcn.net>...> Rune Allnor wrote: > > > ... It's not fun when I have to tell > > people that I'm only an engineer who uses the rules of maths and physics, > > and not some sort of sorcerer. > > > > Rune > > But aren't those the same thing? Many think so!Unfortunately, you're right. The scary thing is *where* I find these kinds of thoughts: - The 40+ years (of experience) senior scientist who wanted to specify "60 degrees phase lag between the 20 Hz component and the 80 Hz component" in a steady-state multiple-lines narrow-band source. I managed to wave that one off. - The 10+ years (of experience) seismic processor who found the sound velocity in the bedrock *immediately below* a high-velocity geologic layer by means of refraction seismics. A trivial property of such analysis is that the sound speed must be strictly *increasing* with depth. I found a possible explanation for this measurement which, if correct, could be the key to understand the acoustics of that whole geographical region. - The head of the Sig Proc department at an international research facility who asked me why I bothered with all that high-res analysis when a simple zero-padding + FFT of my data would resolve the features I was looking for. I was very glad *not* to be offered a job on that facility after that meeting. - The internationally renowned scientist who did not understand why I was so conserned about symmetry properties in the 2D DFT since "it's only the positive frequencies that go into the inverse transform". I was very cautious in how I read his journal articles after that. - The scientist who was overly concerned about not being able to make computational models that were able to generate as strong Rayleigh waves (ground roll) as measured in real life, when analyzing the muzzle blast of heavy artillery. I knew the Rayleigh waves well enough to see other possible excitation mechanisms in that experimental setup. No basic text on Rayleigh waves fail to mention these excitation mechanisms. - The 40+ years (of experience) scientist who asked for my array processing codes for analyzing arrays with logarithmic element spacing. "If you can't give it out, I'll just pick equidistant elements in the array and use standard FFTs". Anyone are free to try to pick an equidistant subsequence of three or more elements out of a series that goes as D_n = D_0*K^(n-1) n=1,2, 3,... where K =/= 1. All these are different people at different facilities, but make mistakes I find utterly trivial. "Why bother?" you may ask, "Everybody make mistakes. It's not unheard of that even a certain Rune comes with corrections to his own posts on comp.dsp". Well, it's a consistent pattern: People who really should know better make the most trivial mistakes. The real problem, though, is that they evidently are not open to neither accept the mistakes or correct them. I can only guess, but what I think I may be doing differently that these people (that may explain my ridicilous "wizard" reputation), is paying attention to detail. To not take anything for granted, not accept something as "easy" or "irrelevant" until I've cheked things out. To try to think the problem through. To try to find out what others have tried and how those attempts went. But that's what basic engineering training is all about, isn't it? Rune
Reply by ●October 12, 20032003-10-12
Rune Allnor wrote:> Jerry Avins <jya@ieee.org> wrote in message news:<bmamiu$e2d$3@bob.news.rcn.net>... > >>Rune Allnor wrote: >> >> >>> ... It's not fun when I have to tell >>>people that I'm only an engineer who uses the rules of maths and physics, >>>and not some sort of sorcerer. >>> >>>Rune >> >>But aren't those the same thing? Many think so! > > > Unfortunately, you're right. The scary thing is *where* I find these > kinds of thoughts: > > - The 40+ years (of experience) senior scientist who wanted to specify > "60 degrees phase lag between the 20 Hz component and the 80 Hz component" > in a steady-state multiple-lines narrow-band source. I managed to wave > that one off. >[more scary stories]> > I can only guess, but what I think I may be doing differently that these > people (that may explain my ridicilous "wizard" reputation), is paying > attention to detail. To not take anything for granted, not accept something > as "easy" or "irrelevant" until I've cheked things out. To try to think the > problem through. To try to find out what others have tried and how those > attempts went. > > But that's what basic engineering training is all about, isn't it? > > RuneMaybe we need to be wizards to get it right, but I think all it takes is recognizing how east it is to be wrong, and finding error embarrassing if not shameful (depending on circumstances). I visited friends last night who anxiously awaited my arrival because the CD player wouldn't work. The drawer wouldn't stay closed. Why me? I had never seen the thing. Of course, I fixed it easily. The TV remote was lying next to the player and stuck on. Moving it out of the way so I had more room to peek was all the fix needed. Do they believe I fixed it by accident? Of course not. I went to high school with these people, and they know me to have been doing that sort if thing for a long time. Once I walked into a room filled with strangers, and was immediately accosted: "You're an electrical engineer. Please fix the television." I did. I replaces the battery in the remote with the cells in my pocket flashlight. If course I carry a flashlight; don't you? Get used to it. By the standards of most of the boobs out there, doctors (of all fields), lawyers, CEOs, you ARE a wizard. Grin and bear it. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●October 12, 20032003-10-12
Jerry Avins <jya@ieee.org> wrote in message news:<bmbq5f$fs9$1@bob.news.rcn.net>...> Once I walked into a room filled with strangers, and was immediately > accosted: "You're an electrical engineer. Please fix the television." I > did. I replaces the battery in the remote with the cells in my pocket > flashlight. If course I carry a flashlight; don't you?In winter, yes. The pockets in my winter coats are large enough to hold the flashlight as well as the drawing pad and my pencils. My summer jackets only have room for the drawing kit. And where I am there's the midnight sun during summer, so I don't need the flashlight... Rune
Reply by ●October 12, 20032003-10-12
Jerry Avins <jya@ieee.org> wrote in message news:<bmbq5f$fs9$1@bob.news.rcn.net>...> Once I walked into a room filled with strangers, and was immediately > accosted: "You're an electrical engineer. Please fix the television." I > did. I replaces the battery in the remote with the cells in my pocket > flashlight. If course I carry a flashlight; don't you?In winter, yes. The pockets in my winter coats are large enough to hold the flashlight as well as the drawing pad and my pencils. My summer jackets only have room for the drawing kit. And where I am there's the midnight sun during summer, so I don't need the flashlight... Rune
Reply by ●October 13, 20032003-10-13
Rune Allnor wrote:> Jerry Avins <jya@ieee.org> wrote in message news:<bmbq5f$fs9$1@bob.news.rcn.net>... > > >>Once I walked into a room filled with strangers, and was immediately >>accosted: "You're an electrical engineer. Please fix the television." I >>did. I replaces the battery in the remote with the cells in my pocket >>flashlight. If course I carry a flashlight; don't you? > > > In winter, yes. The pockets in my winter coats are large enough to hold > the flashlight as well as the drawing pad and my pencils. My summer jackets > only have room for the drawing kit. And where I am there's the midnight > sun during summer, so I don't need the flashlight... > > RuneI use my penlight mostly for looking into and under things, although it was handy down in the depths of Mammoth Cave when a storm knocked out the electricity powering the lights*. Jerry _________________________________________ * Before anyone asks, the guides all had flashlights, but we were strung out in a narrow serpentine passage, and no light from either end of the line came to us near the middle. The little light saved a claustrophobic mother from freaking out in front of her pre-teen son. -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●October 13, 20032003-10-13
Ren� wrote:> Steve wrote: > > Having looked through countless websites and books, I've learnt loads, > > but not what i'm looking for: An algorithm for AM demodulation... Can > > anyone help? > > demodulation > > output = lowpass filter ( abs (input) ) > > Ren�A better way is to recover the original carrier and multiply it back into the AM and then lowpas filter. This is so-called synchronous demodulation.To recover the carrier you could use a PLL or more crudely a hard limiter but the hard limiter is not a great idea at low SNRs. Tom
