Hi Chris.
Sorry, mainly my way of directing thought towards this issue not trying to say others did not feel it also important.
I was just using those numbers as an example to focus our attention on the part that is important. I am new here and wanted to keep my posts simple so all could understand.
Everything in life can matter. But the question is how likely in the OP's current situation will he need twisted power lines in metal conduit?
Sometimes internal filtering is used to prevent that noise from getting out? I would think that is pretty much the function of a power supply other than to store energy.
Which is what I have been saying. This burst of energy the capacitors get every cycle has more high frequency content than what is coming down your 120V line. Those filter capacitors are there to deal with it.
I meant big capacitors relative to the rest of the piece of electronic equipment we are talking about. Generally they are the largest capacitors hanging about.
You only have one phase coming in your home. If you have your stereo on one 120V leg and a noisy device on the other 120V leg of your 240V service the two currents sum together and go in and out of your 240V lines to your house. Any difference goes through the ground from your panel.
We are talking about house wiring here were we not? I am not really familiar with the twisted stuff for houses.
The ground is precisely where you want the noise to go. Where would you put it?
I have read the previous posts. At least using the transformer got turfed. I was trying to explain why the conduit and using twisted cable for power wiring is unnecessary. I do not think it crusading, my saying that the new plan of attack by the OP is a waste of time.
I had spent some time learning about protecting military equipment from conducted and radiated EMI from the kHz range up into the GHz. I have worked in industrial and hospital environments. While I am no expert on all things audio or power related I have seen a few things. Just thought I would give my opinion. I'll try to keep my posts down as not to seem to be crusading.
Hi Chris. Can't recall, maybe Avemco (?), harvested from ITC cart machines.
Within the audio band I didn't see any real difference to choose from between them and Hammond filament transformers. I'ld expect that, it would be a very poor EI indeed that didn't make a solid effort to 20 kHz. Above that, you may have a point but that's awfully easy to filter in tube supplies. http://www.diyaudio.com/forums/tubes-valves/114026-dht-ac-filament-noise.html
Last edited:
I know an isolation transformer worked perfectly for a recording studio project and I would use one again. I spent two weeks with a scope and spectrum analyzer, impedance metering hammering ground rods before things got quiet, so this is my experience. Audiophiles have one high-gain (phono preamp) device whereas a recording studio runs a dozen microphones into a console with a lot more gain, which shows you line noise and RF like nothing you can imagine.
When I worked as a tech in medical research and they had a commercially made shielded room for the "experiments", involving brain cells (micro-volts, preamp gains of a few thousand). It had an isolation transformer and dedicated ground from the 6th floor down to the basement. Not my idea but it worked, the room was electrically quiet enough.
When I worked as a tech in medical research and they had a commercially made shielded room for the "experiments", involving brain cells (micro-volts, preamp gains of a few thousand). It had an isolation transformer and dedicated ground from the 6th floor down to the basement. Not my idea but it worked, the room was electrically quiet enough.
AC Wiring Losses
I did a quick calculation (not to be confused with speculation) on a 50' run of 14AWG NMD with a 120VAC source and 15A load, to compare use of larger wiring.
I could not find any EMI/RFI line filter specs with DC resistance, so who knows the voltage drop (at 60Hz). I thought it would be silly to save 0.05 ohms on your wiring and then add say 0.1 ohm with the filter.
P.S. - Jim, I hope your 12ga. stranded/litz wire meets electrical code, for insulation flammability, voltage etc. I'd hate for the inspector to have you pull it out.
I did a quick calculation (not to be confused with speculation) on a 50' run of 14AWG NMD with a 120VAC source and 15A load, to compare use of larger wiring.
I could not find any EMI/RFI line filter specs with DC resistance, so who knows the voltage drop (at 60Hz). I thought it would be silly to save 0.05 ohms on your wiring and then add say 0.1 ohm with the filter.
P.S. - Jim, I hope your 12ga. stranded/litz wire meets electrical code, for insulation flammability, voltage etc. I'd hate for the inspector to have you pull it out.
Attachments
Hi everyone.
I've been in Spain for the past week, so have not had a chance to look at this thread.
Still no conclusion yet. It sounds like the braided wire and steel pipe are a waste of time. It looks like (from the last post) that I need a heavier wire still to make the ~50' run across the house. Perhaps 8 ga to keep the drop less that 1%.
I have already moved the feed breaker to the top of the panel, and will check to make sure the light breaker for that room is on the other side of the panel (although it seems the 220 for the dryer and oven could cause cross-contamination if either appliance are running while the light is dimmed and the stereo is playing). I suppose I could keep the dryer and/or oven off while listening to the stereo critically...
PS I am enjoying the back and forth here on all the related subjects coming up due to my original question. I am learning a lot!
I've been in Spain for the past week, so have not had a chance to look at this thread.
Still no conclusion yet. It sounds like the braided wire and steel pipe are a waste of time. It looks like (from the last post) that I need a heavier wire still to make the ~50' run across the house. Perhaps 8 ga to keep the drop less that 1%.
I have already moved the feed breaker to the top of the panel, and will check to make sure the light breaker for that room is on the other side of the panel (although it seems the 220 for the dryer and oven could cause cross-contamination if either appliance are running while the light is dimmed and the stereo is playing). I suppose I could keep the dryer and/or oven off while listening to the stereo critically...
PS I am enjoying the back and forth here on all the related subjects coming up due to my original question. I am learning a lot!
If your service entrance panel is like mine, the order of the breakers inside will make no difference because that are all connected to the same think copper bus.
You best best, is to simply use a large enough gauge of wire and then very near the point of use have a good transformer/filter
If you have a refridgerator or dryer running _anyplace_ in the house and it's acoustic noise,that it sends throught ehdrywall and air will hugly overshadow any electrically noise.
Just this weekend I made some quick mesurements with a microphone and fond the noise from my sun's PC power supply fan was far larger then any electrical noise. And his PC was upstairs (one floor up and one closed door away.
A better use of your time and money is to worry about stuff that has a 10x or 100x larger effect on the sound. And this is acoustic noise. contruction technique can have a HUGE effect. For example have then place rock wool insulation in every space inside all interior walls. Use sound deadening board under the drywall. You can also attach the drywall with sound isolation clips, these are cheap and made a huge difference. Building codes allow this as these techniques are used for shared walls high end in multi-unit housing. The best is to use every technique, staggered studs, sound borad, isolation clipps and double layers of 5/8 drywall. All this is approved and to code. and not really expensive and will matter more than 100 times more then the kind of wire you use., maybe much more than 100 times depending oon the layout of the rooms Other small things mater a lot. For example interior doors, use the solid wood kind and not the hollow ones that, and weather stripping on interior doors makes them almost sound proof. In a few places in my house I use double french doors that can be closed to seal off the noise from one area to the others and the 6 foot wide glass still allows an open look. Buy double glasses dors and they are much more sound proof. All this adds up and is cheap and easy to do during construction.
The point here is to identify the important stuff and don't worry abot what does not matter (like the order of circuit breakers that share a power bus.)
Like I said, that fan in a PC in another room matters more. The sound of water in pipes matters too, You can buy platic isolators for water pipe. drill oversize holes and the isolator goes between the pipe and the hole. For waste lines, cast iron is much quiter than PVC but expensive you use it onlyon the vertical drops.
That was the plan (more or less) however the transformer (which I posted photos of) seems to have been shot down by the collective here. It has a good amount of capacitance in it, and may be the right thing (I always thought so) but then some think it may add more noise than it removes...
I am aware of the acoustic noise issue. All floors and ceilings are insulated. The floor of the 'media room' will be braced to prevent it resonating as well. Interior walls are not acoustically insulated unfortunately, and I am not pulling more Sheetrock down just to do that. Seriously considering a second 5/8 layer (over the 3/8 used originally) for the rest of the room in question. This is a bit outside the scope of this original question however...
Very interesting observation! Time to chuck the PC then I guess!
These things are being considered/addressed...
But at the same time, if there is a 'better' way to feed the power, why not do it that way? It costs the same- all I did was try to reduce the influence of other electrical equipment in the house from having an affect on the power feed to the stereo rig.
Waste lines are all copper. All pipes (that I have access to) are insulated with typical foam; penetrations with 'great stuff' fire block. Both bathrooms are to be demo'd/replaced and the pipes at these areas will be similarly treated. Again, a bit off topic from the original one, but welcome advice nonetheless!
If you're having major renovation done, maybe it makes sense to replace all easily accessible romex house wire with "BX" cables. AKA "AC" IIRC. This cable has the neutral and ground wires inside the flexible armor as a twisted pair, about one twist every 2 inches. This looks similar to balanced shielded microphone cable but with heavier wire. This will act as a reverse shield, to contain hum and powerline noise.
No ground rod or combination of will clear a fault by tripping a 20 amp breaker. I am a Senior Electrical Inspector and training officer for the city of Los Angeles (retired).
Best ground is isolated, However, if you are using a metal boc Two ground wires must be run, one isolated and the other a normal ground. Isolated grounds must come directly from the MAIN service panel and go directly to your Isolated ground receptacle. Best run in metal conduit for isolation from noise. Ground rods are only for lightning and High voltage faults to service entrance wires, most caused by auto vs utility pole. If you want I can give you a copy of Grounding essentials Powerpoint ny Tom Trainor, Previous Chief Electrical Inspector and President if IAEA.
Joseph Youhouse
Best ground is isolated, However, if you are using a metal boc Two ground wires must be run, one isolated and the other a normal ground. Isolated grounds must come directly from the MAIN service panel and go directly to your Isolated ground receptacle. Best run in metal conduit for isolation from noise. Ground rods are only for lightning and High voltage faults to service entrance wires, most caused by auto vs utility pole. If you want I can give you a copy of Grounding essentials Powerpoint ny Tom Trainor, Previous Chief Electrical Inspector and President if IAEA.
Joseph Youhouse
New computer and I have not installed Powerpoint viewer yet, sorry about that.
Joeyou3 sent me the PPT in question (also as PDF). Thanks! A good read. Nothing there about audio or noise on the lines, but lots about safety and proper grounding/bounding. The safety of return paths.
I'm not sure if I can post it here for copyright reasons. I will check.
If you care to research, much of it deals with the great improvements in the NEC of 1999. Article 250 of the NEC is where you want to read. The 2008 edition of the code should contain the good grounding practices.
I'm not sure if I can post it here for copyright reasons. I will check.
If you care to research, much of it deals with the great improvements in the NEC of 1999. Article 250 of the NEC is where you want to read. The 2008 edition of the code should contain the good grounding practices.
So by chance have you tested to see what voltage your utility is providing you? Are you going to design around peak usage of your neighbours or minimum...? My point is, for all you know the 1% voltage drop could be giving you 120v exactly at the equipment connection point. Or, you maybe even getting more than 120v.
Bashing your head against a wall over a 1% drop is a waste of time. You'll get greater variations during the day from the utility.
It's not about what the line voltage is steady state, but rather preventing it from being "mushy" under dynamic conditions. Example - movie with dinosaur walking stomping feet, you get a power draw of 1000W then 100W then 1000W then 100W... so the line voltage will go up and down. I'd have a lot of plaster on the floor... and the neighbors calling the cops if my system was cranked like that.
Hi Printer2,
Sorry, I should have commented sooner.
-Chris
Sorry, I should have commented sooner.
That's fair enough.
Well then, welcome to diyaudio. I hope you find enough to keep you interested. You will find members who have levels of understanding from less than what's normal to very advanced. I understand where you are coming from.
For what it's worth, I think twisted pairs in metal conduit is completely unnecessary. Having said that, most Romex is twisted once you get beyond one pair and a ground. Anything I've ever worked with anyway. So those people who feel this is a critical thing, just buy four conductor Romex and you're set.
Yes. Most computing equipment use these filters to prevent the electrical noise they generate from getting out into the wild. In commercial phone systems, they often also employ common mode chokes to keep each flavor of noise on it's own side of the fence.
If the designer has done their job properly, yes. However, since these filters are in fact used to pass EMC radiation standards, you can guess how well noise is normally controlled in equipment. That's actually a little unfair to designers. Noise can be generated by switching power supply operation - some operate in the MHz regions these days. Other things like impact printer heads and similar things may generate a fair amount of noise. Just imagine how much electrical noise was generated by old power drills, vacuum cleaners and mixers! The list goes on and on. I can say that from personal experience, most consumer devices have really poor power supply designs. Audio sound equipment tends to be really bad. The push for higher and higher supply capacitance ignores the fact that internal inductance is higher with increasing capacitor values. Same effect as the capacitors are miniaturized. Very little though seems to be directed towards actually delivering clean, low noise power to the circuitry. Some do, most do not.
The filter caps can't deal with these higher frequencies very effectively. There is often series inductance to deal with, and the caps are not sized appropriately to deal with those higher frequencies. The best way to deal with rectifier noise would be to extend the charging time per cycle (= smaller capacitors), or use another method that draws energy through more of the AC cycle. You want to reduce the generated noise and fast rectifiers coupled with larger filter capacitors only generate more noise. This noise also couples back through the transformer secondary windings to bath other circuits in the electronic sludge. Nice eh?
I'm sorry, you lost me there. My point is the basement "modder" that uses large filter capacitance as one of their money generating procedures. I feel actual service centers that do this should be judged more severely, since they are supposedly properly trained in electronics. Customers look to these people as experts in the field, and some work I have recently seen points to the work of morons.
They may send one phase to the local transformer, but we normally have two phases. There is 208 VAC (nominal) between them. That has sure to have been increased as the 117 VAC (nominal) now measures 125 VAC in many places. My house is one example of this. At least the noise on the line isn't bad.
Yes. I mentioned an example of twisted Romex earlier on. Most AC wiring is not twisted. I have seen twisted feeders entering the house from the pole. No idea in the below ground delivery.
Well, to the ground. However, the question is "when is a ground, not a ground"? One of my favorite questions to ask technicians. The answer? Depends on the system reference. If you look a the point of ground in a piece of equipment, you could look at the ground terminal as ... ground. But, if you consider the entire group of components that make up an audio system (for example), each ground wire is most definitely not a "ground". In truth, you do not want to run any current of any kind in the ground wire. It's there as a reference. RF current from radios will merrily travel down this connection, as well as the neutral connection. They are one and the same at the box grounding point. If current does travel down the "ground" lead, it encounters both resistance and inductance. This will create differentials across various "ground" connections depending on how much current flows down each, and what the impedances are in each leg. Notice the nasty word "impedance". AC noise filters that dump HF noise currents down the "ground" terminal will make this worse. Really, it should simply be made 100% common mode, radio RF should travel back along the neutral. Of course, it will appear everywhere due to capacitive coupling.
Okay, that certainly makes sense, and I agree.
Well then, your experience does entitle you to an opinion based on your experience. Not many people actually have first hand experience with this. The only thing to remember is that medical and military gear has far better power supply design than consumer stuff. So some of your assumptions in that direction should be flushed when considering the average bit of audio gear.
-Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.