Certain posts containing personal attacks and their collateral have been removed from this thread. Gentlemen, kindly keep the discussion civil.
So, they make in all cases wideband (>300kHz, 50MHz) noise floor measurements, the devices are "doing something" - yes decreasing HF transfer like a passive components would - the decrease in HF results of course in reduction of the wideband (HF) noise floor and then they say it would be audible, though no measurement of noise in audioband (BW = 22kHz), except one strange plot with 1dB improvement that might have resulted from repeating error, has been provided. Reduction in HF will be also achieved with any LP filter. In my opinion this deduction from HF noise floor to audibility is a pure pseudo-science and the audio magazine just needed "something" measured to support the claim on sound improvement. I know this situation personally, I was asked several times by local audio producers and sellers to measure "something" on link cables and power cables.
Last edited:
Yes - this 1 db difference in THD+N is really interesting. My inner technician doubts it legitimacy. Yes there where fancy boxes used to do the investigation - but what about the operator? Not 100% convinced. I would have liked to see an independent organisation doing the measurements. A high-end magazine is not. So, not ready to invest in these things and put into my gear. It will take more rigor and stringens to do that.
//
//
If we want to keep these people doing what they are good at, they will have to be paid better, but I don't know that it can ever match what the finance industry can do. The only people that will stay in research are the ones that are not motivated by money and there are only a few of those. Kudos to DPH, but when he sees his buddy driving down the street in the latest Porsche, you can't blame him for wanting change career path.
Power centres running stuff under Win 3.0. Ouch! I'd be afraid of getting hacked
I can see that it is everywhere the same with salaries in science. I had worked in two scientific institutes, those were the most interesting years of my career, however paid badly. Finally I started to co-operate, on long-term consulting basis, with a US company on quite unique industrial technology and this made my income very reasonable.
Many years ago I worked in the power industry, then for the power industry as a consultant. Back then we used PDP11 running mainly home-grown software and home-grown networking. Not a TCP/IP stack in sight. No external connections.
Then they wanted to get data from the plant to the PC on their desk. The PC on their desk was on the office LAN which was connected to the internet. I worked out a way of doing this fairly safely, but warned them that it reduced security.
Then they noticed that PCs were much cheaper than PDP11 or PLC, so PCs started appearing at the plant end. Then the IT people issued an edict that all PCs came under their control, and must be on the LAN so they could be 'managed' efficiently. I warned them that this would be throwing away security, and anyway the IT people would not understand the special requirements of a PC doing plant control (e.g. don't try to 'upgrade' it overnight to a new system version which was not supported by the SCADA software). About this time I left the industry and went back to unversity.
I now hear dire warnings about plant computers being hacked, or facing threats of hacking. Why do people not know that you cannot have something which is both connected (so you can see the data) and isolated (so it is safe from external interference)? It is like a newbie designing a PSU with the wrong transformer, so he wants to run two separate PSUs from one CT secondary.
Then they wanted to get data from the plant to the PC on their desk. The PC on their desk was on the office LAN which was connected to the internet. I worked out a way of doing this fairly safely, but warned them that it reduced security.
Then they noticed that PCs were much cheaper than PDP11 or PLC, so PCs started appearing at the plant end. Then the IT people issued an edict that all PCs came under their control, and must be on the LAN so they could be 'managed' efficiently. I warned them that this would be throwing away security, and anyway the IT people would not understand the special requirements of a PC doing plant control (e.g. don't try to 'upgrade' it overnight to a new system version which was not supported by the SCADA software). About this time I left the industry and went back to unversity.
I now hear dire warnings about plant computers being hacked, or facing threats of hacking. Why do people not know that you cannot have something which is both connected (so you can see the data) and isolated (so it is safe from external interference)? It is like a newbie designing a PSU with the wrong transformer, so he wants to run two separate PSUs from one CT secondary.
So noise/distortion etc. gives rise to different electron energy states in a conductor from music? I'm sure a proven Maxwell demon would easily score a Nobel prize.john curl said:
Generally yes, but in this case the failure to look at it was spectacular.
I have a 1988 wireless world that talks about home automation being the next big thing!
The facebook angle is interesting as many companies have forgotten that you should deliver solutions your customers want rather than what you have to sell. Square pegs being hammered into round holes everywhere.
I have a 1988 wireless world that talks about home automation being the next big thing!
The facebook angle is interesting as many companies have forgotten that you should deliver solutions your customers want rather than what you have to sell. Square pegs being hammered into round holes everywhere.
Agreed.
No baseline, no error bars, no repeat measurement. Once they found a result they liked, it was off to the printer.
1 dB at -97dB? Suspicious.
Does anybody really believe that a high end audio mag would print a neutral or negative article? Or even a test showing that a ferrite bead (as they mentioned) could do the same thing.
Imagine the threat of lawsuit, or even worse, the withholding of articles for evaluation. The nag would go under.
Jn
Yes - and then you find ludicrously complex security on things that really don't need it, and systems that should be ludicrously hard to get into with wide open back doors.
A company I deal with, their IT has locked down the LAN and email so much that you can't essentially send any files. So all the developers send stuff by personal email to their phones, then download to the workstations. By making the security silly, they force people to bypass it completely in order to do their job... Aaargh...
The model in many places. By the end (my end) they froze everything to a 5 year out of date Redhat and I could no longer build upgrades to anything because they had had to keep everything pointed to old libraries for the in house tools.
We had a very senior product line director that, on his own, issued a memo that all free or "beta" software should be removed immediately. I reminded him that would include most of the Linux distribution everyone used.
Our adversaries (whoever they may be) must think it strange that we make it so easy to access our critical infrastructure. Someone needs to tell IT people that computers are now ubiquitous, so they can't claim all computers as IT; imagine what it would have been like if 50 years ago the typing pool claimed control over all keyboards!
One power station did at least make an attempt to be safe. It had a 'diode' in a communications link between two computers, so messages could only go one way. One computer than became known as POTOS: 'processor on the other side' of the diode.
One power station did at least make an attempt to be safe. It had a 'diode' in a communications link between two computers, so messages could only go one way. One computer than became known as POTOS: 'processor on the other side' of the diode.
- Status
- Not open for further replies.