Well, everybody's opinion has been interesting.
The first reason for NOT having an LED was that it added NO real usefulness, and it would be difficult to mount. We would have to drill an extra hole in the front panel, etc..
The second reason is that it would be difficult to POWER the LED, because we are using something like +/- 30V supplies, and current balance between the voltages is critical. The LED could actually offset the voltages, if put on one side only to 'ground'. The best solution would be to add two power resistors, one on each side of the LED, and use it as a floating load across the two supplies of one channel. However, what about the second channel? It won't have the same load. Is that important?
Third, my old friend Enid Lumley could hear LED's used as indicators. Was she crazy? Well, yes, but she could hear better than anyone I have ever known. Why cross Enid? Others have independently found the same thing.
Maybe we could suppress any 'RFI' or whatever else that might be generated in the LED by adding a sizable capacitor across the LED. Might help, who knows?
I hope that 'reasonable people' can see that it just was NOT worth it to add an LED, since there was no 'on/off' switch in any case, and just touching the top of either the power supply or control chassis would prove whether it is plugged in.
Now you know! '-)
The first reason for NOT having an LED was that it added NO real usefulness, and it would be difficult to mount. We would have to drill an extra hole in the front panel, etc..
The second reason is that it would be difficult to POWER the LED, because we are using something like +/- 30V supplies, and current balance between the voltages is critical. The LED could actually offset the voltages, if put on one side only to 'ground'. The best solution would be to add two power resistors, one on each side of the LED, and use it as a floating load across the two supplies of one channel. However, what about the second channel? It won't have the same load. Is that important?
Third, my old friend Enid Lumley could hear LED's used as indicators. Was she crazy? Well, yes, but she could hear better than anyone I have ever known. Why cross Enid? Others have independently found the same thing.
Maybe we could suppress any 'RFI' or whatever else that might be generated in the LED by adding a sizable capacitor across the LED. Might help, who knows?
I hope that 'reasonable people' can see that it just was NOT worth it to add an LED, since there was no 'on/off' switch in any case, and just touching the top of either the power supply or control chassis would prove whether it is plugged in.
Now you know! '-)
OK, let me get this straight.
- The supplies in the BT are so poor that a 5mA current will "unbalance" them.
- Series connections between rails require resistors on each side, ignoring Kirchoff's law
- If a known crazy person says that she can "hear" a 5mA device with insanely low noise connected nowhere near a signal path, that should be the primary engineering consideration
- RFI (or some other emanation) that has never been detected in the physical universe should be addressed by bypass caps
And you wonder why I'm cynical about fashion "high end audio"?
- The supplies in the BT are so poor that a 5mA current will "unbalance" them.
- Series connections between rails require resistors on each side, ignoring Kirchoff's law
- If a known crazy person says that she can "hear" a 5mA device with insanely low noise connected nowhere near a signal path, that should be the primary engineering consideration
- RFI (or some other emanation) that has never been detected in the physical universe should be addressed by bypass caps
And you wonder why I'm cynical about fashion "high end audio"?
I thought it was well known that an LED which is more than a few volts away from ground suffers from electron tunnelling, so it turns into a microwave oscillator. That is why two resistors would be needed per LED, pus a cap across it.
Also, to get good channel balance the left and right power supplies have to be built at exactly the same time (and almost the same place) by identical twins running their soldering irons off the same mains spur. Any subsequent current draw imbalance will upset the built-in symmetry.
Ordinary mortals would of course be unaware of the special requirements of high-end award-winning audio.
Also, to get good channel balance the left and right power supplies have to be built at exactly the same time (and almost the same place) by identical twins running their soldering irons off the same mains spur. Any subsequent current draw imbalance will upset the built-in symmetry.
Ordinary mortals would of course be unaware of the special requirements of high-end award-winning audio.
A center tap is not required for Kirchoff's Law to apply. At least in this universe.
I'm a symmetrical kind of guy. I don't like adding extras on one power supply only. However, my explanation of why I didn't add an LED from 'practical hassle' reasons should have been enough.
Actually, I would have liked to have an LED built inside a 2A circuit breaker, added to the AC input of each power supply module. It would look really 'cool', business like, and only add, maybe $500 to the cost.
It would seem that one cannot win. IF you avoid something, because it would just add hassle in construction, certainly do NOTHING for the sound quality, AND maybe cause a problem, I get criticized.
IF I do something that does cause time and effort to be expended as well as added cost, I get criticized. I can't win.
Speaking of the 'devil', I spoke to Jack Bybee this morning. He has a newer, even MORE expensive product, and he can't make them fast enough. He is on a mission to help audiophiles, and he can't break away to have lunch. That's the way it goes.
Actually, I would have liked to have an LED built inside a 2A circuit breaker, added to the AC input of each power supply module. It would look really 'cool', business like, and only add, maybe $500 to the cost.
It would seem that one cannot win. IF you avoid something, because it would just add hassle in construction, certainly do NOTHING for the sound quality, AND maybe cause a problem, I get criticized.
IF I do something that does cause time and effort to be expended as well as added cost, I get criticized. I can't win.
Speaking of the 'devil', I spoke to Jack Bybee this morning. He has a newer, even MORE expensive product, and he can't make them fast enough. He is on a mission to help audiophiles, and he can't break away to have lunch. That's the way it goes.
Because then it would be good engineering and not performance art. And lots of people can do good engineering, but not many can do performance art.
Morons demanded more.
Oh, wait - here you are, demanding even more, yet again.
Now that I have stated why I don't have a LED in the CTC Blowtorch, what about 'why' I don't use a center tap on any of the 4 power transformers in the CTC Blowtorch?
To save time, I will first say, to remove ANY connection of the power supply transformers from the system 'ground' or 'common' (a more realistic term in most of my designs)
To save time, I will first say, to remove ANY connection of the power supply transformers from the system 'ground' or 'common' (a more realistic term in most of my designs)
Last edited:
I might also say that I heard from our old friend and colleague, Charles Hansen of Ayre.
Charles and I have NO problem discussing the same topics, except for the fact that we are 'competitors' in the audio marketplace, and he is one of my more formidable competitors.
Still, we have lots of 'common ground' to relate to, including getting older.
Charles and I have NO problem discussing the same topics, except for the fact that we are 'competitors' in the audio marketplace, and he is one of my more formidable competitors.
Still, we have lots of 'common ground' to relate to, including getting older.
For more than a year I had the intention to inform the pro audio people (at prodigypro) about one of my findings but being too busy with work I had no time to do it, but after reading all this LED discussion I thought I will share one of my disturbing experiences.
I'm not really an audiophile (although I'm very interested in great sound) I'm a sound engineer and mastering engineer who also designs and builds many of his tools ( various signal processors, equalisers, mastering consoles, tape machine electronics etc). I also have and use some very refined devices from different pro audio manufacturers.
This whole introduction is just to understand the context where I'm coming from.
Not so long ago I moved to a different place in a building where my listening room was about 25-30meters above earth level. In this new location I had some serious RF problems. At 1-1.5 kilometers away from the window you can see the TV FM & AM radio antennas of most of the station here in the capital city.
I use for monitoring some active monitors made by one of the most important speaker makers from western Europe (I will not name the manufacturer). They are not some very expensive ones but they cost some grands so we expect them to produce some very good sound. And yes, these speakers can sound very nice but in this new location the sound was mostly horrible (not acoustic problems, I use acoustic absorbers/diffusers). A relatively constant metallic sound and depending on the hours of the day you could hear if your ears were close to the speakers higher than normal level of noise and/or demodulated audio from the radio transmissions. At some very few hours almost every day my playback system could sound very good but most of the time it sounded horrible. So I decided to find the source of the problem. One by one I isolated every device till I discovered that even if I don't connect any cable to the balanced input of the active speaker the RF noise and/or the demodulated audio was still hearable. Wow!
Next step was to use a male XLR plug and short the + and - inputs to ground and check what is going on. Nothing, no improvement. So I started to disassemble the speakers and try different tricks to eliminate the RF problem. There was no pin1 problem with the XLR connector. I twisted every cable inside the speaker, used ferrites, I used techniques to short RF at the loudspeaker driver , tried to use ceramic capacitors at various points to create low impedances at RF. After all these tricks there was a very small improvement but the noise and the demodulated audio was still there. And the sound was still horrible.
After some days of thinking I decided to do something radical and to shield with copper the whole interior of the active monitor. I did this mod and the shielding was done with copper sheets glued to the interior walls and soldered them to produce a continuous box. I had big hopes that this will solve all the problems. It didn't. Again there was a small improvement but the problem was still there.
I was disturbed by this experience and I could not imagine what I omitted in the whole process. I left the speakers for weeks and decided to listen to them only in those few hours when the RF noise was non audible. It was extremely frustrating.
One day completely by accident I had a crazy idea. There are two LEDs on the front panel of each speaker one for signaling that the speakers are powered and the other to signal overload. I thought that the only exposed parts of the speaker to the outer world are now only these two LEDs. The LEDs being diodes it is not totally unacceptable to think that they could detect/demodulate RF. At the very first moment it sounded absurd and almost dismissed the idea. I was thinking, there is no reason, they are utilitarian LEDs with no real connection to the audio electronics. I had a look at the schematic and saw that the audio electronics (the discrete power amps plus crossover plus balanced receiver circuit) shared only the GND at the power supply point with the two LEDs. I was still thinking it would be a crazy idea to disable the LEDs, but I tried everything and had no idea what else I could do. So decide to open again the speakers and completely disconnect the LEDs from the circuit.
Like a miracle all the RF noise and demodulated AF vanished. Also the whole metallic sound was gone. I was not believing this was possible. So to check what was going on I connected a signal generator (multi sine wave) and a spectrum analyzer to the amps and compared the spectras between the two variants, speakers with LEDs connected and speakers with LEDs removed. The difference in noise and distortion components was something like 25-30dB.
All this trouble was generated by some LEDs.
chrissugar
P.S. what I learned from this is that one thing is how it performs a device on the test bench and a different thing is how they perform in real world situations connected to unknown devices, in RF hostile environments and with bad power. (all those PPM distortion specs will be something different outside the lab). There is a positive side of this story. For some years I test all the gear I design and build in this RF hostile place, and if here the device performs flawless, it will perform probably anywhere.
I'm not really an audiophile (although I'm very interested in great sound) I'm a sound engineer and mastering engineer who also designs and builds many of his tools ( various signal processors, equalisers, mastering consoles, tape machine electronics etc). I also have and use some very refined devices from different pro audio manufacturers.
This whole introduction is just to understand the context where I'm coming from.
Not so long ago I moved to a different place in a building where my listening room was about 25-30meters above earth level. In this new location I had some serious RF problems. At 1-1.5 kilometers away from the window you can see the TV FM & AM radio antennas of most of the station here in the capital city.
I use for monitoring some active monitors made by one of the most important speaker makers from western Europe (I will not name the manufacturer). They are not some very expensive ones but they cost some grands so we expect them to produce some very good sound. And yes, these speakers can sound very nice but in this new location the sound was mostly horrible (not acoustic problems, I use acoustic absorbers/diffusers). A relatively constant metallic sound and depending on the hours of the day you could hear if your ears were close to the speakers higher than normal level of noise and/or demodulated audio from the radio transmissions. At some very few hours almost every day my playback system could sound very good but most of the time it sounded horrible. So I decided to find the source of the problem. One by one I isolated every device till I discovered that even if I don't connect any cable to the balanced input of the active speaker the RF noise and/or the demodulated audio was still hearable. Wow!
Next step was to use a male XLR plug and short the + and - inputs to ground and check what is going on. Nothing, no improvement. So I started to disassemble the speakers and try different tricks to eliminate the RF problem. There was no pin1 problem with the XLR connector. I twisted every cable inside the speaker, used ferrites, I used techniques to short RF at the loudspeaker driver , tried to use ceramic capacitors at various points to create low impedances at RF. After all these tricks there was a very small improvement but the noise and the demodulated audio was still there. And the sound was still horrible.
After some days of thinking I decided to do something radical and to shield with copper the whole interior of the active monitor. I did this mod and the shielding was done with copper sheets glued to the interior walls and soldered them to produce a continuous box. I had big hopes that this will solve all the problems. It didn't. Again there was a small improvement but the problem was still there.
I was disturbed by this experience and I could not imagine what I omitted in the whole process. I left the speakers for weeks and decided to listen to them only in those few hours when the RF noise was non audible. It was extremely frustrating.
One day completely by accident I had a crazy idea. There are two LEDs on the front panel of each speaker one for signaling that the speakers are powered and the other to signal overload. I thought that the only exposed parts of the speaker to the outer world are now only these two LEDs. The LEDs being diodes it is not totally unacceptable to think that they could detect/demodulate RF. At the very first moment it sounded absurd and almost dismissed the idea. I was thinking, there is no reason, they are utilitarian LEDs with no real connection to the audio electronics. I had a look at the schematic and saw that the audio electronics (the discrete power amps plus crossover plus balanced receiver circuit) shared only the GND at the power supply point with the two LEDs. I was still thinking it would be a crazy idea to disable the LEDs, but I tried everything and had no idea what else I could do. So decide to open again the speakers and completely disconnect the LEDs from the circuit.
Like a miracle all the RF noise and demodulated AF vanished. Also the whole metallic sound was gone. I was not believing this was possible. So to check what was going on I connected a signal generator (multi sine wave) and a spectrum analyzer to the amps and compared the spectras between the two variants, speakers with LEDs connected and speakers with LEDs removed. The difference in noise and distortion components was something like 25-30dB.
All this trouble was generated by some LEDs.
chrissugar
P.S. what I learned from this is that one thing is how it performs a device on the test bench and a different thing is how they perform in real world situations connected to unknown devices, in RF hostile environments and with bad power. (all those PPM distortion specs will be something different outside the lab). There is a positive side of this story. For some years I test all the gear I design and build in this RF hostile place, and if here the device performs flawless, it will perform probably anywhere.
All this trouble was generated by some LEDs.
More likely connecting wires. Presumably the LEDs are forward biased?
- Status
- Not open for further replies.