It has come to my attention that several very high power DIY and commercial production Class-D amplifier projects are under way. These may be powerful enough to do damage to the human condition by inadvertantly radiating audio and radio noise into spectrums that violate not only the (USA) FCC rules, but can easily be harmful to your pets and yourself.
Considering that a switching power supplies and or switching amplifiers may be using frequencies well into the broadcast radio spectrum, the radiated harmonics can and will cause harm to those exposed. So some guidelines for construction may be in order:
* It is strongly recommended that metal cases be used for all of these projects where the power is intended to exceed 300 Watts per channel. This is prudent scientific engineering practice as the quality of the audio output will be improved as well as the shielding of spurious radiation in the radio frequency (RF) range. (300 Watts is somewhat arbitrary, but a 300 Watt switching power supply can easily have broadcast RF harminics of several Watts in the 300K Htz to 3000K Htz range or even beyond. The metal cases should be well grounded to the house power/ground circuit using very low impedence, low resistance methodologies. The metal cases should be closed whenever power is applied. *
No kidding, this stuff could fry your gonads.
Reference: http://www.diyaudio.com/forums/showthread.php?threadid=83424&goto=newpost
Considering that a switching power supplies and or switching amplifiers may be using frequencies well into the broadcast radio spectrum, the radiated harmonics can and will cause harm to those exposed. So some guidelines for construction may be in order:
* It is strongly recommended that metal cases be used for all of these projects where the power is intended to exceed 300 Watts per channel. This is prudent scientific engineering practice as the quality of the audio output will be improved as well as the shielding of spurious radiation in the radio frequency (RF) range. (300 Watts is somewhat arbitrary, but a 300 Watt switching power supply can easily have broadcast RF harminics of several Watts in the 300K Htz to 3000K Htz range or even beyond. The metal cases should be well grounded to the house power/ground circuit using very low impedence, low resistance methodologies. The metal cases should be closed whenever power is applied. *
No kidding, this stuff could fry your gonads.
Reference: http://www.diyaudio.com/forums/showthread.php?threadid=83424&goto=newpost
FastEddy
Thanks to your warning! You should not only specify risk in this warning, but more way to avoid it. Certification is about 300mW of RF power at 1 metter of appliance. That's easy to got more than this if you not follow basic rule as you said. Metal grounded cas is a first step to avoid radiation. Second step is to get the cleanest overshoot free HV PWM signal. Dont expect to push up the voltage if you can get clean square wave at low voltage. Limit the dV/dt will help a lot too to reduce EMI, but reduce effiency....But their another way to limit dV/dt without sacrifice in effiency (commond mode output filter for full bridge amplifier). Shielding of input modulator and analog processing (filter) will avoid noise pick-up that can be directly amplified and present at output terminal. Another point, very important, is to never have a chance to touch HV part of the amplifier. 200-360V DC can stop your heart!
A cheap way to check EMI...First look at your TV if you have diagonal line...That's means more work to do in your amplifier! Second, buy a cheap AM radio and remove antenna. If you ear what's palying in your class d, that's means more work again. That's normal to have radio emission, but you should not overpass your local radio station! AM radio without antena is a good (cheap) way to ensure that your under 300mW at 1 meter...Next part for you is about conducted emission.
Fredos
Thanks to your warning! You should not only specify risk in this warning, but more way to avoid it. Certification is about 300mW of RF power at 1 metter of appliance. That's easy to got more than this if you not follow basic rule as you said. Metal grounded cas is a first step to avoid radiation. Second step is to get the cleanest overshoot free HV PWM signal. Dont expect to push up the voltage if you can get clean square wave at low voltage. Limit the dV/dt will help a lot too to reduce EMI, but reduce effiency....But their another way to limit dV/dt without sacrifice in effiency (commond mode output filter for full bridge amplifier). Shielding of input modulator and analog processing (filter) will avoid noise pick-up that can be directly amplified and present at output terminal. Another point, very important, is to never have a chance to touch HV part of the amplifier. 200-360V DC can stop your heart!
A cheap way to check EMI...First look at your TV if you have diagonal line...That's means more work to do in your amplifier! Second, buy a cheap AM radio and remove antenna. If you ear what's palying in your class d, that's means more work again. That's normal to have radio emission, but you should not overpass your local radio station! AM radio without antena is a good (cheap) way to ensure that your under 300mW at 1 meter...Next part for you is about conducted emission.
Fredos
" ... Metal grounded case is a first step to avoid [external] radiation. Second step is to get the cleanest overshoot free HV PWM signal. ..."
... and this greatly improves signal to noise ratios = better and better audio ...
" .. Another good test is to place two amplifiers next to each other (unshielded) and see how much they interact. If they're synchronized and they make lots of noise then you have issues. If they aren't and there isn't any interaction then that's obviously good. ..."
This bares (bears?) repeating ... very good advise = a simpler test methodology than exotic radiation sensing or other, and it can be done with a 'scope and a signal source (signal generator or even a music source), even old fashioned, surplus 'scopes and generators work for this. Comparing your new amp design / prototype with an amp that has passed FCC muster is a good idea too (shielded, unshielded, in the cases or not).
This is called "cross talk" and is associated with common mode rejection (CMRR) = audio range signal source through one amp, then measuring the output of its neighbor, then reversing the amps for a comparison. Lots can be learned here about the quality of shielding, etc. Space the amps about 3 feet (1 meter) apart to get approximations comparable to what the FCC may be interested in. (Sometimes an FCC examiners will pass you when they see your test bench with this kind of set up.)
... and this greatly improves signal to noise ratios = better and better audio ...
" .. Another good test is to place two amplifiers next to each other (unshielded) and see how much they interact. If they're synchronized and they make lots of noise then you have issues. If they aren't and there isn't any interaction then that's obviously good. ..."
This bares (bears?) repeating ... very good advise = a simpler test methodology than exotic radiation sensing or other, and it can be done with a 'scope and a signal source (signal generator or even a music source), even old fashioned, surplus 'scopes and generators work for this. Comparing your new amp design / prototype with an amp that has passed FCC muster is a good idea too (shielded, unshielded, in the cases or not).
This is called "cross talk" and is associated with common mode rejection (CMRR) = audio range signal source through one amp, then measuring the output of its neighbor, then reversing the amps for a comparison. Lots can be learned here about the quality of shielding, etc. Space the amps about 3 feet (1 meter) apart to get approximations comparable to what the FCC may be interested in. (Sometimes an FCC examiners will pass you when they see your test bench with this kind of set up.)
An interesting way to generate emissions (so you can test the immunity of your amp) is to take the output filter off of a fixed frequency PWM class d amp and attach a junk speaker to the output. Use some speaker cables of decent length and you have a good RF generator.
Another cheap trick is to place a cell phone near the amp under test and have someone call that phone. You can hear lots of funky noises due to the phone emitting a lot of crap into the amp. Move the phone around and see where it leaks into the amp the most. This is also a decent way to see if your shielded wiring really works or not.
Another cheap trick is to place a cell phone near the amp under test and have someone call that phone. You can hear lots of funky noises due to the phone emitting a lot of crap into the amp. Move the phone around and see where it leaks into the amp the most. This is also a decent way to see if your shielded wiring really works or not.
BWRX: ... all good tricks ... Got more?
Appreciate it. We might come up with a "poor boy's" check list for DIY construction /Êsmall manufacturer prototyping of Class-D devices ...
" ... Use some speaker cables of decent length and you have a good RF generator. ..." This is exactly what the FCC "radio cops" would do if they ever come a callin' ... and if anyone builds really powerful Class-D (or other) amps that leak, especially in a big city, they will come = count on it!!
This is effectively "pirate radio", whether your project is broadcasting music, politics or just plain old noise & interference. (Reason: the RF noise could interfer with aircraft radio traffic = crash = FCC, FAA, FBI, CIA, NSA, Homeland, etc. ... like stink on poo poo.)
Appreciate it. We might come up with a "poor boy's" check list for DIY construction /Êsmall manufacturer prototyping of Class-D devices ...
" ... Use some speaker cables of decent length and you have a good RF generator. ..." This is exactly what the FCC "radio cops" would do if they ever come a callin' ... and if anyone builds really powerful Class-D (or other) amps that leak, especially in a big city, they will come = count on it!!
This is effectively "pirate radio", whether your project is broadcasting music, politics or just plain old noise & interference. (Reason: the RF noise could interfer with aircraft radio traffic = crash = FCC, FAA, FBI, CIA, NSA, Homeland, etc. ... like stink on poo poo.)
FastEddy don't be damn paranoid. The frequencies involved in SMPS and class D amplifiers, including parasitistics, are always below 100Mhz so their wavelenghts are way too large to interact with animal or human body cells. A typical 50Khz SMPS carrier is 6000 meters long, a 250Khz class D carrier is 1200 meters long, and even a piece of worst-case 100Mhz MOSFET gate ringing is 3 meters long.
What is well known to interact with living cells are microwaves. That's because the wavelenghts involved are much shorter (30cm for 1Ghz, 3cm for 10Ghz <-- a true killer). That includes MOBILE PHONES and any form of WIRELESS INTERNET. Furthermore, those devices must radiate 1 watt or more of harmful microwaves in order to reach the nearest repeater (which may be near your home and radiating way more power).
On the other hand, you will hardly see a SMPS or a class D amplifier radiating more than a fraction of a miliwatt (even without a metal case), mainly because huge antennas would be required in order to radiate the long wavelenghts involved efficiently.
Note that recent studies carried out in Europe showed that microwaves at "mobile phone" dosages are enough to induce chemical changes in living cells, including ADN modifications.
BTW: Do you have a mobile phone? or a WI-FI router? or maybe a PDA? If yes, you'd better take care of your gonads and stop blaming harmless stuff
What is well known to interact with living cells are microwaves. That's because the wavelenghts involved are much shorter (30cm for 1Ghz, 3cm for 10Ghz <-- a true killer). That includes MOBILE PHONES and any form of WIRELESS INTERNET. Furthermore, those devices must radiate 1 watt or more of harmful microwaves in order to reach the nearest repeater (which may be near your home and radiating way more power).
On the other hand, you will hardly see a SMPS or a class D amplifier radiating more than a fraction of a miliwatt (even without a metal case), mainly because huge antennas would be required in order to radiate the long wavelenghts involved efficiently.
Note that recent studies carried out in Europe showed that microwaves at "mobile phone" dosages are enough to induce chemical changes in living cells, including ADN modifications.
BTW: Do you have a mobile phone? or a WI-FI router? or maybe a PDA? If yes, you'd better take care of your gonads and stop blaming harmless stuff
Eva said:
they are also plenty of research refuting that.
the book isn't written on low wattage and long-term exposure to microwaves (in the wireless bands). with a maximum signal strength in the <1w range, and more often than not much lower than that.
cell phones have been in use for a good 20 some years (and more if you count the more powerful analog types). and we haven't seen noticeable increase in any radiation-induced diseases. not yet at least.
Eva...Just a quick question...You never feel the heat on a cool coil? If you play with fews Kw as you say, you have probably feel that starnge phenomene....Even with a 250-500Khz frequency...What I read in certification manual, is that maybe too the hight dV/dt of the squared wave that cause HF interference...
Fredos
Fredos
" ... The frequencies involved in SMPS and class D amplifiers, including parasitistics, are always below 100Mhz so their wavelenghts are way too large to interact with animal or human body cells. ..."
Fine, good even.
The point is not the human condition, but the other electronic systems in your area that may be effected ... Being more concerned about: "... broadcasting music, politics or just plain old noise & interference. (Reason: the RF noise could interfer with aircraft radio traffic = crash = FCC, FAA, FBI, CIA, NSA, Homeland, etc. ... like stink on poo poo.)"
(If I were really, really paranoid, I would have mentioned the military HF audio work using mixed signal audio noise to cause the Taliban to throw up on the battle field ... or the use of multiplexed ultra sound audio to kill cancer cells ... or the VLF radio harmonics that seem to interfer with the breeding cycle of whales ... or those new medical side scan digital ultra sound devices to make movies of your liver in action.)
Fine, good even.
The point is not the human condition, but the other electronic systems in your area that may be effected ... Being more concerned about: "... broadcasting music, politics or just plain old noise & interference. (Reason: the RF noise could interfer with aircraft radio traffic = crash = FCC, FAA, FBI, CIA, NSA, Homeland, etc. ... like stink on poo poo.)"
(If I were really, really paranoid, I would have mentioned the military HF audio work using mixed signal audio noise to cause the Taliban to throw up on the battle field ... or the use of multiplexed ultra sound audio to kill cancer cells ... or the VLF radio harmonics that seem to interfer with the breeding cycle of whales ... or those new medical side scan digital ultra sound devices to make movies of your liver in action.)
fredos said:
I have never experienced that up to the date. I use 50Khz and my coils are designed to leak as little magnetic flux as possible. I always use a loop antenna attached to an oscilloscope probe in order to check how much my circuits are leaking in each plane.
However, I have experienced other funny things like heavy gauge wires expanding and contracting or moving when 150A pulses are made to flow through them.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.