Ha - yes... Plumbers do not understand the flow of water either... I had a big shower pump installed, output was pathetic... Then I saw the tortuous plumbing - no wonder! Had it re done without loads of 90 degree bends, pipe diameter restrictions... Result pins you to the shower wall as desired! 🙂
The RF stuff is somewhat a distraction, I'm sticking to the dirrectionality and distortion at audio frequencies.
I was just the monkey that had to pull the stuff through the cable trays and conduits. They had real engineers there responsible for the connections.😉 so, I don't know what they did..
The LMR 240 I had to pull in groups of 5. We had to absolutely control radius of curvature and sectional integrity. When a coax is over bent, the shield currents will redistribute to maintain a common centroid, and if that happens the inductance will increase and that kills the prop velocity. Four of the cables had to be exactly the same length electrically as the real engineers (not me) use inter cable prop delays as a measure of sumptin 😕, and that extends down below the picosecond level. All the kids doing the pulls either had bikes or hot cars, and they all did their own work on their machines. Explaining prop velocity concerns using the brake line example was speaking their language, so the job was done perfectly. 17 thousand cables ranging from 535kcmil pairs, helix, lmr240, 400 foot single mode fiber, totally perfect. The kids were good.
And, all their matched length runs were single ended, so ground loop concerns were huge for me as was tray layout and conduit layout.
Yet the IEEE found it acceptable to retire STD-1050?? That was the freakin bible for my install design.
jn
"What hit me in the face, was the fact that the motion control engineers, even the with masters degrees, have absolutely no clue what a bode plot is, nor what phase margin is. NO experience with gears, springs, masses, electronics, nothing!! Never tuned a car, never cut a gear, never operated a lathe or milling machine."
----
Has a lot to do with the fact that universities are operating as businesses nowadays and not as institutions of learning. Some here will no doubt reject that characterisation. But, consider that many of the major universities in the US and in the UK are fabulously wealthy. Put making money ahead of educating youngsters, and pay university deans six figure salaries based upon how much money the institution makes and you have a recipe for disaster.
Back in my day, BTW, we had to spend 6 weeks in 'filing school': You were given a lump of metal (literally) from which you had to file various complex shapes. 'Hard labour' and frustration that resulted in cut knuckles, slashed thumbs but at the end of it, a real appreciation for the skill and knowledge it takes to fabricate something. Most folks who enjoy all the technologic marvels around us today have no clue about that it takes to produce those things - they live in alternative reality . . .
----
Has a lot to do with the fact that universities are operating as businesses nowadays and not as institutions of learning. Some here will no doubt reject that characterisation. But, consider that many of the major universities in the US and in the UK are fabulously wealthy. Put making money ahead of educating youngsters, and pay university deans six figure salaries based upon how much money the institution makes and you have a recipe for disaster.
Back in my day, BTW, we had to spend 6 weeks in 'filing school': You were given a lump of metal (literally) from which you had to file various complex shapes. 'Hard labour' and frustration that resulted in cut knuckles, slashed thumbs but at the end of it, a real appreciation for the skill and knowledge it takes to fabricate something. Most folks who enjoy all the technologic marvels around us today have no clue about that it takes to produce those things - they live in alternative reality . . .
More to the point though 😉....that doesn't explain the FUD about wire directionality, JC is old enough to know better.
Here are two other books that I have found valuable for 'deep' understanding of what is happening in wire and contacts. They both can be very useful on occasion.
JN has already 'trashed' one of them, years ago. I hope people realize this is just a 'debating' technique normally used by JN over the decades, and not to be taken seriously. Of course, I have many more books related to physics, as well as engineering. It would be interesting if JN would tell us what references he uses, don't you think?
JN has already 'trashed' one of them, years ago. I hope people realize this is just a 'debating' technique normally used by JN over the decades, and not to be taken seriously. Of course, I have many more books related to physics, as well as engineering. It would be interesting if JN would tell us what references he uses, don't you think?
Attachments
No, it wouldn't!
your juvenile willy-waving attempts are cringe-making embarrassing!
wHy do you need to do this, JC, with your pedigree?
your juvenile willy-waving attempts are cringe-making embarrassing!
wHy do you need to do this, JC, with your pedigree?
I think that it is time for JN to give us a list of references, so that we can all be on the same page.
Primary engineering interests in 'long' coax cable runs is the signal timing... thus, velocity of prop. needs to be known.
Also the signal dispersion.... waveform change. Especially if peak level needs to be known.... The impulse signal is reduced in amplitude and spread out in time. ........have to deconvolve the waveform knowing the dispersion properties;
http://lss.fnal.gov/archive/other/lbl-cc-2-1b.pdf
THx-RNMarsh
Also the signal dispersion.... waveform change. Especially if peak level needs to be known.... The impulse signal is reduced in amplitude and spread out in time. ........have to deconvolve the waveform knowing the dispersion properties;
http://lss.fnal.gov/archive/other/lbl-cc-2-1b.pdf
THx-RNMarsh
Last edited:
Bumping https://www.diyaudio.com/forums/the...wtorch-preamplifier-iii-1410.html#post5707239 because of little response so for (thanks to those who did).
On a tangent for those interested in low-level high resolution measurements, over at ASR I just updated a thread about measuring linearity of the lowest 4 bits of a DAC directly which is normally impossible. The level is so low that it is almost is completely covered in analog noise, reducing effective resolution to some 20 bits or so, best case.
However, with heavy time-domain averaging I was able to reduce uncorrelated noise enough to clearly see the LSB dancing in high resolution.
By chance, it seems I have found a processing glitch (all positive sample values are off by 1 LSB) in my DAC during these tests, producing a deterministic very low level error (down below analog noise floor) with according distortion, additionally verified by modelling of the error replacing the real device: RME Adi-2 Pro FS (AK4490 DAC): positive sample values are always offset by -1 ! | Audio Science Review (ASR) Forum
On a tangent for those interested in low-level high resolution measurements, over at ASR I just updated a thread about measuring linearity of the lowest 4 bits of a DAC directly which is normally impossible. The level is so low that it is almost is completely covered in analog noise, reducing effective resolution to some 20 bits or so, best case.
However, with heavy time-domain averaging I was able to reduce uncorrelated noise enough to clearly see the LSB dancing in high resolution.
By chance, it seems I have found a processing glitch (all positive sample values are off by 1 LSB) in my DAC during these tests, producing a deterministic very low level error (down below analog noise floor) with according distortion, additionally verified by modelling of the error replacing the real device: RME Adi-2 Pro FS (AK4490 DAC): positive sample values are always offset by -1 ! | Audio Science Review (ASR) Forum
I still have somewhere my official diploma in manual metal turning and milling (CNC was unheard at that time). And yes, I have a micro lathe and a desktop CNC mill and can turn manually and CNC mill whatever non standard parts I need, to about 10-20um precision. Also CNC mill single side PCBs that are not worth sending to a board house (like large power supply boards, with 12mm copper traces).
I had a similar problem generating 24 bit files from 64 bit floats. It was a long time ago but the solution for that problem was using ceil() rather than floor() for rounding. One left distortion after a 1LSB TPDF dither and the other had a perfect noise floor. There are also several discussions on the web about the issue that for signed integers the max positive and min negative values are not the same magnitude, I don't really follow the issue in the detail some programmers do.
Last edited:
Scott -- pretty simple, "0" (zero) is needs to be defined either as a positive or negative number; there isn't -0 and +0 in 2's complement. By convention zero is +0, but I'm sure someone will come along with a case where the opposite is true. 🙂
E.g. a 8-bit 2's complement integer is [-2^7, 2^7-1] or [-128,127]
E.g. a 8-bit 2's complement integer is [-2^7, 2^7-1] or [-128,127]
You already know I've never trashed any books. I have pointed out when you take things out of context in an attempt to bury us with bs, you are of course referring to our discussion long ago where you were trying to justify near superconducting.
For my references regarding superconductors, I do not need any books, I just walk down the hall and ask the best sources on this planet.
For EM theory, I walk to a different door and again ask the best in the world.
For liquid helium based 3 and 4 inch diameter diodes, i do not have to leave my office.
To adapt NEC to a scientific machine, same. For winding high accuracy magnets...perhaps you get the picture?
I am trying my best to not remain the "reference" for high accuracy motion control, there are two individuals I am teaching who I believe can surpass me.
Eventually what I do will get into the textbooks. But more importantly, I hope the advances I am achieving will be used in the beamline by medical researchers to develop tools for cancer that work. Half of the synchrotron beamlines in the world are used for biology.
Jn
Last edited:
- Status
- Not open for further replies.