Kindhornman, correct me if I'm wrong, but paralleling 4 100V caps will still liit you to 100V, but you will have 4 times their capacitance (assuming they are the same value).
Putting two say 100uF/63V caps is series will give a voltage limit of (2*63) 126V, but half their capacitance (100:2) 50 uF.
Correct?
Putting two say 100uF/63V caps is series will give a voltage limit of (2*63) 126V, but half their capacitance (100:2) 50 uF.
Correct?
dvv,
I would only assume that with the much larger surface area of the multiple capacitors you would have a much harder time having a failure due to a failure of the film between layers. What is the normal failure mode of a film capacitor, what is happening? Someone smarter than me can answer that. I didn't say you increased the voltage rating but I would equate it though not really to paralleling two 1/4 watt resistors and having a 1/2 watt capability between two of them. Someone is welcome to correct me or shoot holes in that thinking.
I would only assume that with the much larger surface area of the multiple capacitors you would have a much harder time having a failure due to a failure of the film between layers. What is the normal failure mode of a film capacitor, what is happening? Someone smarter than me can answer that. I didn't say you increased the voltage rating but I would equate it though not really to paralleling two 1/4 watt resistors and having a 1/2 watt capability between two of them. Someone is welcome to correct me or shoot holes in that thinking.
Have you measured or is this Wikipedia....
Again you say loud , i said domestic , have you ever measured how much power is required to raise VC TEMP to 150, i have measured while testing and have never seen such temps unless in sound reinforcement where there is no xover. In domestic use its pretty rare to be using more than 10-15 watts RMS EVEN at elevàted volume and yes , I'm aware of this strawman argument from those pushing the virtues of Horns ...
I have seen caps fail in xovers big time and i never said it had anything to do with ESR , i mentioned this could be more of a concern than heating VC's in domestic use, you should measure...
Sorry, no use for me discussing this with somebody who doesn't want learn.
One quick reference about voice coil heating and temperature rise for those who care to read.
http://www.onesystems.com/pdf/education/Inside_Only_Voice_Coils.pdf
http://www.aes.org/e-lhttp://www.ci...nder - JBL) (1984).pdfib/browse.cfm?elib=5251
http://www.onesystems.com/pdf/education/Inside_Only_Voice_Coils.pdf
http://www.aes.org/e-lhttp://www.ci...nder - JBL) (1984).pdfib/browse.cfm?elib=5251
Failures in hi powered Xovers are usually in the bass or Midbass regions (80-100V) where electrolytic types are use and not 400v poly. As to cap failure it is not common unless used for PA , it was mentioned as an aside to VC induced distortion due to temps , in the past i had done alot of testing on this and could not determine this was an issue for domestic used speakers.
addressing Strawman VC argument :
Power testing was done at 1W/2W/4W/8W/16W Continuous sweeps from 10-40K, there was no variation in
* Fr
* Impedance mag/Phase
* THD
Additionally VC temps were measures with infra-red therm and increase in temps over ambient were minimal (1-8 deg) and cool down was almost instantaneous even with woofers, from memory i believe we even tested at 32W but I'm not absolutely sure , but 16W was the standard testing procedure when matching Prs.
The best thermal transfers were from ribbon drive units, thoughts at the time were the alloy diaphragm being directly cooled by the ambient air passing over the diaphragm, being fully exposed at all times.
My conclusion at the time ( 18yrs ago) and still such today is VC induced distortion in modern day loudspeakers for domestic listening environment is not an issue, as most quality drive units do not suffer from high VC temps to make a difference..
A few of the Drivers tested with xover:
Morel MDT30
Dynaudio Esotar tweeter
Seas 6.5 inch poly
Seas Alloy cone 8 inch woofer
Vifa 12 inch woofers
addressing Strawman VC argument :
Power testing was done at 1W/2W/4W/8W/16W Continuous sweeps from 10-40K, there was no variation in
* Fr
* Impedance mag/Phase
* THD
Additionally VC temps were measures with infra-red therm and increase in temps over ambient were minimal (1-8 deg) and cool down was almost instantaneous even with woofers, from memory i believe we even tested at 32W but I'm not absolutely sure , but 16W was the standard testing procedure when matching Prs.
The best thermal transfers were from ribbon drive units, thoughts at the time were the alloy diaphragm being directly cooled by the ambient air passing over the diaphragm, being fully exposed at all times.
My conclusion at the time ( 18yrs ago) and still such today is VC induced distortion in modern day loudspeakers for domestic listening environment is not an issue, as most quality drive units do not suffer from high VC temps to make a difference..
A few of the Drivers tested with xover:
Morel MDT30
Dynaudio Esotar tweeter
Seas 6.5 inch poly
Seas Alloy cone 8 inch woofer
Vifa 12 inch woofers
Dont be so hard on yourself, feel free to show how you arrived at 150C, i know it's easier for you to deflect..
...
Last edited:
I agree workload sharing will be default offload individual capacitors and will reduce their output impedance.
However, their voltage rating will be limited by the smallest rating in any group. Thus, paralleling two 1/4 W resistors is not the same thing as paralleling two caps. Much less so that paralleling two transistors. Each item has its own rules.
HornMan ,
More strawman arguments from those peddling horns for non domestic use .....
Last edited:
Hi Steven, thanks for the links, but the 2nd one does not seem to work. Is it behind a pay wall?
Wayne, doesn't the whole argument also hinge on speaker sensitivity, or the voltage and current required to drive them?
In my last little adventure, I pushed my own speakers, in my room, to 96 dB SPL, but at 3 rather than the usual 1 metre distance. Aside from the fact that altogether it was far too loud for listening longer than say 20-30 minutes, depending on the endurance of your ears, the peak actually hit 70W into nominally 8 Ohms, minimum 6.5 Ohms.
Nothing happened except a lot of music too loud for comfort, under normal circumstances I drive them with well below 1 W AVERAGE power, obviously more on transients, but all that is well below anything remotely looking like critical.
I take it that's what you meant with normal use, under home conditions. PA is something else again, there you are using literally tens of kilowatts for a stadion. Having worked with and for The Greateful Dead, John is better qualified to comment on such power levels.
Anyway, the point is that if one's speakers are less efficient, one will need more driving power for the same effect, and thus be more exposed to possible overdrive problems. I think Kindhorman mentuioned typical bookshelf speakers, with an efficiency of say 86 dB/2.83V/1m, but he didn't mention that such small bookshelf speakers usually have less tolerance for large power inputs, and are there again more in danger of being overdriven.
In my last little adventure, I pushed my own speakers, in my room, to 96 dB SPL, but at 3 rather than the usual 1 metre distance. Aside from the fact that altogether it was far too loud for listening longer than say 20-30 minutes, depending on the endurance of your ears, the peak actually hit 70W into nominally 8 Ohms, minimum 6.5 Ohms.
Nothing happened except a lot of music too loud for comfort, under normal circumstances I drive them with well below 1 W AVERAGE power, obviously more on transients, but all that is well below anything remotely looking like critical.
I take it that's what you meant with normal use, under home conditions. PA is something else again, there you are using literally tens of kilowatts for a stadion. Having worked with and for The Greateful Dead, John is better qualified to comment on such power levels.
Anyway, the point is that if one's speakers are less efficient, one will need more driving power for the same effect, and thus be more exposed to possible overdrive problems. I think Kindhorman mentuioned typical bookshelf speakers, with an efficiency of say 86 dB/2.83V/1m, but he didn't mention that such small bookshelf speakers usually have less tolerance for large power inputs, and are there again more in danger of being overdriven.
Life gets interesting if the amp clips with an asymmetric signal - you get a DC component at the output. IMHO, that's something worth thinking about, especially when designing active speakers. If the design is along the lines of filters=>power-amps=>drivers, then there's nothing protecting your tweeters from the power amp's self-inflicted DC offset. One good way round that is to make the power amp part of the filter. e.g. design a vcvs filter and make the power amp itself the active element in the filter circuit.
Yes and No D,
16W and 32W is irrelevant of sensitivity or listening distance , this does not affect the test results, the sensitivity and listening distance affects how much power you need to achieve your listening SPL goal, If at 100w RMS I'm sure temp rise would be different , experience had shown 100w RMS sustained to be very unlikely in a domestic environment , even on my low sensitivity speakers i rarely exceed 60 watts continuous and the heat from amp and xover is substantially , ribbon temps are barely 2 -8 deg above ambient, cool down is instantaneous.
These guys need to actually measure instead of this strawman argument, we have only seen VC temp issues in PA sound reinforcement setups, at obscene RMS levels, not domestic...
16W and 32W is irrelevant of sensitivity or listening distance , this does not affect the test results, the sensitivity and listening distance affects how much power you need to achieve your listening SPL goal, If at 100w RMS I'm sure temp rise would be different , experience had shown 100w RMS sustained to be very unlikely in a domestic environment , even on my low sensitivity speakers i rarely exceed 60 watts continuous and the heat from amp and xover is substantially , ribbon temps are barely 2 -8 deg above ambient, cool down is instantaneous.
These guys need to actually measure instead of this strawman argument, we have only seen VC temp issues in PA sound reinforcement setups, at obscene RMS levels, not domestic...
Last edited:
addressing Strawman VC argument :
Power testing was done at 1W/2W/4W/8W/16W Continuous sweeps from 10-40K, there was no variation in
* Fr
* Impedance mag/Phase
* THD
Additionally VC temps were measures with infra-red therm and increase in temps over ambient were minimal (1-8 deg) and cool down was almost instantaneous even with woofers, from memory i believe we even tested at 32W but I'm not absolutely sure , but 16W was the standard testing procedure when matching Prs.
???
No change in variation in THD from 1--16W?
Measuring VC temperature with an IR thermometer?
There are better ways to measure the temperature of a conducting piece of wire with a known temperature coefficient of resistance.
Last edited:
A.wayne, how would you measure the temperature of a conducting piece of wire with a known temperature coefficient of resistance? That is really a tough one.
Thermal management is one of the hardest areas of driver design. No way has this anything to do with marketing BS. Thermal mismanagement is a distortion mechanism, it leads to compression, and it is one of the most populair failure modes. I have never seen a failed capacitor in a xover, other than elco's from old age, but the smell of burnt voice coil former and lacquer is something I know all too well.
Thermal management is one of the hardest areas of driver design. No way has this anything to do with marketing BS. Thermal mismanagement is a distortion mechanism, it leads to compression, and it is one of the most populair failure modes. I have never seen a failed capacitor in a xover, other than elco's from old age, but the smell of burnt voice coil former and lacquer is something I know all too well.
Last edited:
You are discussing cause, i never disputed that , so please pay attn ..
I SAID IN DOMESTIC APPLICATION we could not measure or detect VC temp to be an issue.
You are shouting semantics ...
There is no burning of VC at typical domestic listening levels , never , sold hundreds of prs of speakers for domestic use never had one fail in the field. Only PA club systems exhibit what you claim.
So ,
Please list your testing procedure and personal results when you have tested , i voiced mine, semantics not getting us anywhere, we tested as high as 32watt RMS, how many exceeding 32w RMS in a domestic environment and in practical use torture testing we have used in an attempt to see if we had issues , pushed amps to clipping when testing on music, we had no such issues or damage..
PS: 32 watt rms translate to 384 watts peak on dynamic music , easily , you have a 400 watt rms/ch amplifier , think about it
I SAID IN DOMESTIC APPLICATION we could not measure or detect VC temp to be an issue.
You are shouting semantics ...
There is no burning of VC at typical domestic listening levels , never , sold hundreds of prs of speakers for domestic use never had one fail in the field. Only PA club systems exhibit what you claim.
So ,
Please list your testing procedure and personal results when you have tested , i voiced mine, semantics not getting us anywhere, we tested as high as 32watt RMS, how many exceeding 32w RMS in a domestic environment and in practical use torture testing we have used in an attempt to see if we had issues , pushed amps to clipping when testing on music, we had no such issues or damage..
PS: 32 watt rms translate to 384 watts peak on dynamic music , easily , you have a 400 watt rms/ch amplifier , think about it
Last edited:
- Status
- Not open for further replies.