Hi
RF hash can certainly be a tricky problem to deal with. It can be difficult to shield the circuit from external RF sources. When I was in school, one of the buildings was built with an RF shielding of grounded mesh within all the exterior walls as it was designed to be used by engineering students developing radio circuits. You cannot pick up a radio signal inside, but have to drape your antenna out the window.
Unfortunately, now I live in an area with much more RF noise, and do have a problem with RF hashing up my output, although I can get it down to a relatively small level. Careful grounding techniques are a helpful first step. Having the input filter as close to the input devices as possible helps too...IMO. Bonsai has a good point about the RF entering the input stage via the feedback compensation. Yet another reason to take the compensation from VAS instead of the output in SE GNFB circuits. Many amps require a series output coil to block this RF from the speaker wires, AKA antennae. I find the EMI radiating from florescence lighting ballasts can be maddening as well, so you’re not alone here. I prefer good ol' incandescent.
RF hash can certainly be a tricky problem to deal with. It can be difficult to shield the circuit from external RF sources. When I was in school, one of the buildings was built with an RF shielding of grounded mesh within all the exterior walls as it was designed to be used by engineering students developing radio circuits. You cannot pick up a radio signal inside, but have to drape your antenna out the window.
Unfortunately, now I live in an area with much more RF noise, and do have a problem with RF hashing up my output, although I can get it down to a relatively small level. Careful grounding techniques are a helpful first step. Having the input filter as close to the input devices as possible helps too...IMO. Bonsai has a good point about the RF entering the input stage via the feedback compensation. Yet another reason to take the compensation from VAS instead of the output in SE GNFB circuits. Many amps require a series output coil to block this RF from the speaker wires, AKA antennae. I find the EMI radiating from florescence lighting ballasts can be maddening as well, so you’re not alone here. I prefer good ol' incandescent.Hi,
A couple of months a go a visiting speaker manufacturer and I were conducting some tests on a system which included a new design of his.
Part way through these measurements I put on a familair CD as a reality check on the overall balance of sound, and the result was atrocious. It was so bad that I couldn't bear to listen to it, and we set to in the hope of finding something broken/failed somewhere in the system.
We found nothing which made any improvement, until after about 30 minutes when the S/M charger for the lap-top which we were using in connection with some frequency measurements was switched off, as the battery was by then fully charged.
The difference between the charger being switched 'on' in the same room, compared with it being 'off', was like night and day to me, as it adversely affected the sound by so much. The mains wiring in my audio test area is a very carefully installed (and costly!) dedicated entirely separate system from the remaining property circuits, and which the offending charger was plugged into, so it is most unlikely that this was other than air-borne hash causing this problem.
Incidentally, like CBS240, a long while ago I worked within a Faraday cage environment when at Joseph Lucas, who were responsible for (almost) all of the UK's motor vehicle electrics in those days, and we were developing interference shielding for the then new plastic-bodied (glass-fibre) cars.
Regards,
A couple of months a go a visiting speaker manufacturer and I were conducting some tests on a system which included a new design of his.
Part way through these measurements I put on a familair CD as a reality check on the overall balance of sound, and the result was atrocious. It was so bad that I couldn't bear to listen to it, and we set to in the hope of finding something broken/failed somewhere in the system.
We found nothing which made any improvement, until after about 30 minutes when the S/M charger for the lap-top which we were using in connection with some frequency measurements was switched off, as the battery was by then fully charged.
The difference between the charger being switched 'on' in the same room, compared with it being 'off', was like night and day to me, as it adversely affected the sound by so much. The mains wiring in my audio test area is a very carefully installed (and costly!) dedicated entirely separate system from the remaining property circuits, and which the offending charger was plugged into, so it is most unlikely that this was other than air-borne hash causing this problem.
Incidentally, like CBS240, a long while ago I worked within a Faraday cage environment when at Joseph Lucas, who were responsible for (almost) all of the UK's motor vehicle electrics in those days, and we were developing interference shielding for the then new plastic-bodied (glass-fibre) cars.
Regards,
Bobken said:
I found the same problem at my colleague, who owned Denon system and whom I visited with my notebook and external soundcard. It was impossible to listen when powered from SMPS. Only battery power was usable. At home, in my system which I designed to be EMI proof, I have no problem with SMPS supplied notebook. And there are so many nuances between unlistenable (clicks, buzz etc.) and perfect.
Hi Pavel,
A lot of people have ridiculed my 'listening' observations, especially when there is no apparent theoretical reason for the phenomena which I have observed and commented upon.
However, the incident I referred to was just another occasion when although nothing untoward had been seen during conventional measurements, I immediately 'heard' this unwanted problem as soon as I listened to some familair sounds through this system.
Of course, we were not testing specifically for any RF hash, but one sometimes needs to know exactly what can potentially have some adverse effects on the overall sonics of systems, before knowing what and how to measure for them, and there are still some areas in this respect which are not very well understood, in my experience.
Regards,
A lot of people have ridiculed my 'listening' observations, especially when there is no apparent theoretical reason for the phenomena which I have observed and commented upon.
However, the incident I referred to was just another occasion when although nothing untoward had been seen during conventional measurements, I immediately 'heard' this unwanted problem as soon as I listened to some familair sounds through this system.
Of course, we were not testing specifically for any RF hash, but one sometimes needs to know exactly what can potentially have some adverse effects on the overall sonics of systems, before knowing what and how to measure for them, and there are still some areas in this respect which are not very well understood, in my experience.
Regards,
lumanauw said:
Hi Lumanauw,
My background has been in analogue audio for around 40 yrs, and although I have used some digital equipment, I have done very little development work on them until recently, preferring to stay with what I am more familair with. Around 18 mths ago I was asked to do some work on a DAC for another UK manufacturer, and I was extremely surprised to hear exactly the same 'sonic' characteristics when trying different components in digital circuits, as I am familair with in analogue audio.
I addition to the resistors mentioned in your reference here (which hasn't copied over with this quote, I see), I found just the same with capacitors.
This defies logic and any conventional training/wisdom etc., even more than with analogue, in my view, but nevertheless this is what I hear.
Regards,
For those with no scope or method of checking RF trash , get a copy of Spectralab Trial ( valid 30 days ) and look at the spectrum . Then turn on the RF gadget or even the TV and you will see some new spurs popping up .
In some amps I couldn't hear any degradation in the audio but the spectrum surely is different. The spurs will be there even without a test signal. Maybe a bit different (?) . I didn't look at them too closely. It just occured to me to note down their amplitudes and frequencies. Will do that later.No time now.
In some amps I couldn't hear any degradation in the audio but the spectrum surely is different. The spurs will be there even without a test signal. Maybe a bit different (?) . I didn't look at them too closely. It just occured to me to note down their amplitudes and frequencies. Will do that later.No time now.
Hi, Bobken,
The most interesting part is that Bruno Putzey is very familiar with digital (look at the list of his papers, mostly about digital), yet he authored a paper with that subject. Interesting POV from a digital background man.
I don't understand that paper wholy, don't understand the mechanisme yet, but reading the title of Bruno's AES paper above (and considering his technical capability), I think I won't be surprised if you find out like your quote above, or if you can hear sonic difference between digital coaxial cables (where it should not make a difference).
The most interesting part is that Bruno Putzey is very familiar with digital (look at the list of his papers, mostly about digital), yet he authored a paper with that subject. Interesting POV from a digital background man.
I don't understand that paper wholy, don't understand the mechanisme yet, but reading the title of Bruno's AES paper above (and considering his technical capability), I think I won't be surprised if you find out like your quote above, or if you can hear sonic difference between digital coaxial cables (where it should not make a difference).
lumanauw said:
Hi Lumanauw,
I am familair with Bruno's work, but (although perhaps this has more-recently changed?) we depart on some obvious matters like the effect of using Black Gate caps in circuits. Personally, I avoid electrolytics at all costs unless there is no alternative, but although Bruno apparently measured some and concluded that they were nothing special, I found that BG 'N' non-polarised caps are overall the best-performing electrolytic caps available, 'sonically'. Maybe he didn't test the 'N' types, though, but as BGs have ceased in production, anyway, it is all academic now.
You are quite right about wires, and I have something up my sleeve here for if I ever fall on hard times financially.
By chance, I discovered an easy way to substantially improve the sonics of a piece of commercial digital gear which I had been asked to do some work on, to see if I could improve it. One short piece of badly-chosen, lousy wire installed internally and which carries only digital signals, really brought down the subjective performance of this piece of equipment, to an extent which shocked me.
I'm keeping this to myself for now, as I am sure that there is some commercial mileage in revealing this, if the need ever arises!
Regards,
No question RF is an issue. It's huge for some, less so for others. I have a CCFL over my bench, but I keep a regular bulb handy for use during any low level measurements. My ceiling lights are also on a dimmer, and often have to be turned off during tests. Even the computer upstairs where I'm typing right now can have an obvious effect. I guess these things don't bother me so much because I know about them, and it's routine to run around turning things off, looking for a change. I also keep a multi-band radio handy on the bench (the Sony 2010 in the pictures of dirty benches). It's an old trick to probe circuits with a small radio, looking for oscillations and emissions. It's also useful for probing household devices- my humidifier was nasty until I installed an LC filter on the speed control. I'm also suspicious of the power pole outside. I can hear, and sometimes see, the arcing on the insulators if the weather conditions are right, mostly high humidity. That must be producing some dreadful crud. My present opinion, open to change, is that many/most RF suppression techniques, applied to audio devices, degrade the sound. It's always better to kill it at the source, but obviously that's not always possible. A scope is handy, but isn't going to help a bit with cell frequencies and much other rubbish.
lumanauw said:
Hi,
I haven't (yet!) experienced that phenomenon, but tending to stick to similar brands (and very few electrolytics anyway) I usually don't have this choice.
At a pure guess, it could be because the 'path' through them is shorter, or something similar. 40 yrs ago I would have agonised over such a 'discovery', but nowadays I merely accept these things at face-value, and take good advantage of any such potential benefits.
Regards,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.