Speaker Protection
Hello Dave
I would not assume the speaker protection works, the IC might fuction correctly but the relay used to disconnect the speaker in the event of a DC fault condition does look like it can break a 60V DC without the contacts welding short . In other words your speaks are not really protected.
Regards
Arthur Rappos
whams said:Hey guys...
I used the White Noise Audio loudspeaker protection unit for my monoblocks. It works a treat 😉
The protection unit can accept 12 VAC to power it so when I got my transformers made, I asked for a 12V secondary also so I power the unit straight from my transformer.
These amps sound fantastic....I am so happy with them. A big thanks and thumbs up to Bruno and Jan-Peter.
Cheers,
Hello Dave
I would not assume the speaker protection works, the IC might fuction correctly but the relay used to disconnect the speaker in the event of a DC fault condition does look like it can break a 60V DC without the contacts welding short . In other words your speaks are not really protected.
Regards
Arthur Rappos
Re: Speaker Protection
Hi Arthur,
I decided to take your comments to the main man of White Noise Audio (David White) to get his take on it. He is a great guy to deal with and has assured me that his speaker protection units do the job that they are asked to do.
Here is what he said :
"Hi Dave,
Nice looking power amps, congrats. Hmmm. Well the LPM uses only transistors, not an IC in sight. As regards the relay its rated at 16A and is certainly good for 75V dc - the voltage rails used in my highest power amps. If you made a habit of always turning your amp on under full drive, or hot plugging the input with the volume turned to maximum on a regular basis then you would eventually ruin the relay contacts ( I suspect you'd ruin a lot more before the relay ). Even then it'd take several years. I suspect that the life of the relay is governed by fatigue in the plastic ( deliberately chosen so as not to introduce distortion ) actuating mechanism rather than ruination of the contacts. Don't forget too that the relay isn't ever going to operate under worst case conditions - i.e breaking a heavy fault current at 60V dc - because the LPM actuates the relay at +/-1.5V of dc fault voltage. There will be precious little, if any, arcing at these modest voltage and current levels so the relay contacts will take this minimal punishment essentially for ever. Anyway how often do you expect a dc fault on your amps output. Rest assured LPMs are worth every penny, I make sure that all of my own amps have them
Regards, David"
I'm sticking with his protection module, I think his products are great. 😉
Cheers,
Dave
PHEONIX said:
Hi Arthur,
I decided to take your comments to the main man of White Noise Audio (David White) to get his take on it. He is a great guy to deal with and has assured me that his speaker protection units do the job that they are asked to do.
Here is what he said :
"Hi Dave,
Nice looking power amps, congrats. Hmmm. Well the LPM uses only transistors, not an IC in sight. As regards the relay its rated at 16A and is certainly good for 75V dc - the voltage rails used in my highest power amps. If you made a habit of always turning your amp on under full drive, or hot plugging the input with the volume turned to maximum on a regular basis then you would eventually ruin the relay contacts ( I suspect you'd ruin a lot more before the relay ). Even then it'd take several years. I suspect that the life of the relay is governed by fatigue in the plastic ( deliberately chosen so as not to introduce distortion ) actuating mechanism rather than ruination of the contacts. Don't forget too that the relay isn't ever going to operate under worst case conditions - i.e breaking a heavy fault current at 60V dc - because the LPM actuates the relay at +/-1.5V of dc fault voltage. There will be precious little, if any, arcing at these modest voltage and current levels so the relay contacts will take this minimal punishment essentially for ever. Anyway how often do you expect a dc fault on your amps output. Rest assured LPMs are worth every penny, I make sure that all of my own amps have them
Regards, David"
I'm sticking with his protection module, I think his products are great. 😉
Cheers,
Dave
Hi,
as I see it, this is more wishful thinking than reality. When Class D fails catastrophicaly, you will have one mosfet shortcircuited and output will rise to the rail in less than 100us. Even if your protection circuit racts instantly, there is still relay drop off time which is typically 10-20ms. In this time fault current might rise quite high. Bacause speaker load is inductive, you will definitely have arc across contacts. The only question is if the relay is capable of breaking such arc or not.
Best regards,
Jaka Racman
as I see it, this is more wishful thinking than reality. When Class D fails catastrophicaly, you will have one mosfet shortcircuited and output will rise to the rail in less than 100us. Even if your protection circuit racts instantly, there is still relay drop off time which is typically 10-20ms. In this time fault current might rise quite high. Bacause speaker load is inductive, you will definitely have arc across contacts. The only question is if the relay is capable of breaking such arc or not.
Best regards,
Jaka Racman
I'm getting worried here. I'm using off the shelf Velleman K4701's, same principle but only a cheapish 10A relay.... 
I agree with Jaka totally, the relay won't be fast enough to switch before the voltage increases to the rails, on any kind of power amp not just class-d.
However, I use a similar 16A relay to White Noise for DC protection myself and as David says, they don't need break such massive energy regularly and they are up to the job. They are rated to something like 2-3kW inductive load breaking capacity.
However, I use a similar 16A relay to White Noise for DC protection myself and as David says, they don't need break such massive energy regularly and they are up to the job. They are rated to something like 2-3kW inductive load breaking capacity.
what about using solid state relays ? I know they are pretty expensive but if one want to be safe....
Giorgio
Giorgio
There are two main kinds of solid state relays. The ones that use triacs at the output are obviously unsuitable...! "photovoltaic" types (which use a string of photodiodes to bias the gates of two source-to-source MOSFETs) could work, although the on resistance of standard AC types would wreck the output impedance (and add some distortion).patriz said:
I once made such a circuit using two banks of 200V parallel fets in such a configuration. It worked, and the whole thing had only about 20mohms resistance, but I could still hear it, in a not too audiophile system at that. The difference was not big but the effect was something like the archetypical "transistor amp" sound (choked mids).
Others report that mechanical relays in the output are also not free from colourations. While I'm a bit skeptical about the "sound" of a good relay, I prefer to err on the safe side and put the relay in the PSU instead (lots of discussions about this elsewhere in this and in the ucd180 thread).
Thanks Bruno, but if you put the relays in the psu once it has been disconnected by relays because of dc your dc sensor would'nt sense anymore dc switching the relays so you have psu again on the circuit but then the sensor will see again dc and so on in the threads you mentioned I didn't find how you solved this. Maybe I just missed it .
Giorgio
Giorgio
It is described in there. Add a circuit to latch the relay on forever once the DC protection is triggered, so that you need to unplug the amplifier to reset it.patriz said:
I use the normally-off contact of the relay itself to power its own coil after it has been momentarily actuated by the DC protection.
Simple eh? No extra parts needed for the latch function, just use the relay itself to hold itself on.
Growing concerned!!!
Bruno, I'm now becoming VERY concerned!!! By reading the numerous posts about DC protection it looks like the presence of destructive DC voltages at the output terminals is a fairly common event.
In the 6-channel amp I'm building I've got no space at all to put in such a circuit: even if I could accomodate the detection unit, I can't accomodate 12 relays.
How likely is it, in your experience, that after the initial testing and experimenting, when infant mortality is likely to occur, a MOSFET fails letting the full DC potential through?
In other words, how much is a DC protection circuit really needed?
What is your experience/track record with the UCD400 modules?
Bruno, I'm now becoming VERY concerned!!! By reading the numerous posts about DC protection it looks like the presence of destructive DC voltages at the output terminals is a fairly common event.
In the 6-channel amp I'm building I've got no space at all to put in such a circuit: even if I could accomodate the detection unit, I can't accomodate 12 relays.
How likely is it, in your experience, that after the initial testing and experimenting, when infant mortality is likely to occur, a MOSFET fails letting the full DC potential through?
In other words, how much is a DC protection circuit really needed?
What is your experience/track record with the UCD400 modules?
Re: Growing concerned!!!
Hi,
Don't worry. Such catastrophic faults are rare events. All modules that leave the factory have been fully tested. This catches "infant mortality" cases.
The large amount of posts does in no way reflect the likelyhood of a catastrophic failure but the potential seriousness of the result if you do have a failure without a DC protection installed. Applying 60V DC to a loudspeaker can lead to fire. To make an analogy, if a car manufacturer talks a lot about the airbags in their cars, you wouldn't think that their cars crash more often that other manufacturers'. You'd understand they want to keep their customers alive, in good shape and happy by making safer vehicles.
Likewise, DC protection is standard practice in amplifiers. For in case an audio manufacturer were to think this is not so important, there is also a legal requirement for all equipment sold to fail in a safe manner when it fails. To this end, mandatory tests are conducted on equipment where the failure of every single part is simulated, once by shorting and once by opening. During such a test, no fire may occur, no solder joints are allowed to come loose and any smoke produced may not be combustible.
This is the law. The law is not concerned with the likelyhood of a failure, but with the results of one, if and when it happens.
As part of their designed-in safety, every audio amplifier sold to the public as a finished product is equipped with a measure to prevent large DC voltages from appearing on the speakers. The design and implementation of such protections is a matter of course for every commercial amplifier project.
Since we do not sell standalone amplifiers, we don't have DC protection on every single module. This would be uneconomical. Instead, we equip our power supply modules with a relay that disconnects the supply. In this way, you need only one (albeit larger) relay for all channels, instead of one relay per channel. If you are building your own supply, I suggest you do the same because I heartily sympathise with you when you don't want to have 12 relays in your system!
Since the modules don't have their own DC protection, we find it very important that users understand what a DC protection is for, and hopefully insure that they have one on their project.
I find the fact that there is so much talk about proper protection here a good sign.
Typical field failure rates (per annum) across the consumer electronics industry vary roughly between 0.1% and 2%. Our modules are on the low end of that spectrum. This is necessary in order to insure that our customers can guarantee a good low failure rate on their product when they have large numbers of channels installed.
I know that other vendors fail to inform their users about DC protection, for fear of alarming their potential customers. Do not be mislead by this. Their field failure rate is not lower than ours. Even if by some magic (mil-spec incoming and outgoing inspection) they did manage to get a 0.01% failure rate, it is simply irresponsible to sell modules without instructions for safe use. Even if only one can go wrong on a user and burn down his house, this is one too many.
Of course they are not legally liable, because they didn't sell the finished product (the customer assembled that himself), but the attitude is extremely irresponsible. We prefer not to take this attitude. If our discussions alarm some people, so be it. This is better than having an uninformed user ring us up and tell us he's had a serious problem.
I hope this answer puts these things in perspective, and leaves you reassured that our preoccupation about product safety has nothing to say about the reliability (which happens to be another preoccupation).
Cheers,
Bruno
m.parigi said:
Hi,
Don't worry. Such catastrophic faults are rare events. All modules that leave the factory have been fully tested. This catches "infant mortality" cases.
The large amount of posts does in no way reflect the likelyhood of a catastrophic failure but the potential seriousness of the result if you do have a failure without a DC protection installed. Applying 60V DC to a loudspeaker can lead to fire. To make an analogy, if a car manufacturer talks a lot about the airbags in their cars, you wouldn't think that their cars crash more often that other manufacturers'. You'd understand they want to keep their customers alive, in good shape and happy by making safer vehicles.
Likewise, DC protection is standard practice in amplifiers. For in case an audio manufacturer were to think this is not so important, there is also a legal requirement for all equipment sold to fail in a safe manner when it fails. To this end, mandatory tests are conducted on equipment where the failure of every single part is simulated, once by shorting and once by opening. During such a test, no fire may occur, no solder joints are allowed to come loose and any smoke produced may not be combustible.
This is the law. The law is not concerned with the likelyhood of a failure, but with the results of one, if and when it happens.
As part of their designed-in safety, every audio amplifier sold to the public as a finished product is equipped with a measure to prevent large DC voltages from appearing on the speakers. The design and implementation of such protections is a matter of course for every commercial amplifier project.
Since we do not sell standalone amplifiers, we don't have DC protection on every single module. This would be uneconomical. Instead, we equip our power supply modules with a relay that disconnects the supply. In this way, you need only one (albeit larger) relay for all channels, instead of one relay per channel. If you are building your own supply, I suggest you do the same because I heartily sympathise with you when you don't want to have 12 relays in your system!
Since the modules don't have their own DC protection, we find it very important that users understand what a DC protection is for, and hopefully insure that they have one on their project.
I find the fact that there is so much talk about proper protection here a good sign.
Typical field failure rates (per annum) across the consumer electronics industry vary roughly between 0.1% and 2%. Our modules are on the low end of that spectrum. This is necessary in order to insure that our customers can guarantee a good low failure rate on their product when they have large numbers of channels installed.
I know that other vendors fail to inform their users about DC protection, for fear of alarming their potential customers. Do not be mislead by this. Their field failure rate is not lower than ours. Even if by some magic (mil-spec incoming and outgoing inspection) they did manage to get a 0.01% failure rate, it is simply irresponsible to sell modules without instructions for safe use. Even if only one can go wrong on a user and burn down his house, this is one too many.
Of course they are not legally liable, because they didn't sell the finished product (the customer assembled that himself), but the attitude is extremely irresponsible. We prefer not to take this attitude. If our discussions alarm some people, so be it. This is better than having an uninformed user ring us up and tell us he's had a serious problem.
I hope this answer puts these things in perspective, and leaves you reassured that our preoccupation about product safety has nothing to say about the reliability (which happens to be another preoccupation).
Cheers,
Bruno
comments on sound quality
I just got my ucd400's working this past weekend.
Much has been said about the sound quality of the ucd400, but I thought another person's experience might be helpful (though my amplifier points of reference may be somewhat outdated).
These comments are for the ucd400 with ad8620, electrolytic coupling caps shorted, polypropylene 4.7 uF caps on the inputs for DC offset protection. No DC protection on outputs (yet).
Configured as monoblocks, power supply used is nothing special, 280 VA power transformer, 10,000 uF filter capacitor per rail, 61 volts per rail.
Speaker used is a ribbon/dynamic array of my own design, all dipole. The amplifier powers the dipole panel that covers 110hz on up. Preamp is a Sony TA-P9000es. Source material is SACD with a decent player, and some 24/96 material.
Generally: Very impressive! Better than Parasound hca1500a (much better), better than Audio Research D115kmII tube amp (hoped for but not really expected). This is my first experience with a switching amplifier in my own system. The ucd400 is better than D115mkII for massed vocals, cymbals/percussion, and bowed string instruments, which is very impressive for any solid state amp. This amp does all the right things with my speakers. The images break free from the speakers, the speakers sound as flat as they measure, etc.
Sound quality of the setup is very similar to a Martin Logan CLS with an all-tube amplifying chain, e.g., an Audio Research SP10mkII and D115mkII. Very transparent and "live sounding" over a broad range of music types. I get a very good "relaxation response" from this setup. These are first impressions. Long term experience may differ. This is the best sounding solid state amplifier in my experience, though I have no experience with the best solid state amplifiers of the past 10 years (e.g., Pass Labs x-series).
Really good loudspeakers benefit a great deal from using the highest quality amplification, and this amp seems to be worthy of the best (IMHO). I look forward to the ucd700, since I have some power transformers that are too high a voltage for the ucd400's (is 77 VDC per rail enough for the ucd700?).
In a few weeks I will try this on an electrostatic setup, replacing an Audio Research M100 monoblock pair. I also hope to get my Martin Logan CLS speakers restored for a real test of the ucd400's
Future tests will include comparing the ucd700 to the ucd400. I may end up converting all my amplifiers with these if "long term enjoyment" is anything like my initial impressions. It would be nice to get away from occasional re-tubing and the power consumption, heat, and electric bills associated with the tube equipment.
I just got my ucd400's working this past weekend.
Much has been said about the sound quality of the ucd400, but I thought another person's experience might be helpful (though my amplifier points of reference may be somewhat outdated).
These comments are for the ucd400 with ad8620, electrolytic coupling caps shorted, polypropylene 4.7 uF caps on the inputs for DC offset protection. No DC protection on outputs (yet).
Configured as monoblocks, power supply used is nothing special, 280 VA power transformer, 10,000 uF filter capacitor per rail, 61 volts per rail.
Speaker used is a ribbon/dynamic array of my own design, all dipole. The amplifier powers the dipole panel that covers 110hz on up. Preamp is a Sony TA-P9000es. Source material is SACD with a decent player, and some 24/96 material.
Generally: Very impressive! Better than Parasound hca1500a (much better), better than Audio Research D115kmII tube amp (hoped for but not really expected). This is my first experience with a switching amplifier in my own system. The ucd400 is better than D115mkII for massed vocals, cymbals/percussion, and bowed string instruments, which is very impressive for any solid state amp. This amp does all the right things with my speakers. The images break free from the speakers, the speakers sound as flat as they measure, etc.
Sound quality of the setup is very similar to a Martin Logan CLS with an all-tube amplifying chain, e.g., an Audio Research SP10mkII and D115mkII. Very transparent and "live sounding" over a broad range of music types. I get a very good "relaxation response" from this setup. These are first impressions. Long term experience may differ. This is the best sounding solid state amplifier in my experience, though I have no experience with the best solid state amplifiers of the past 10 years (e.g., Pass Labs x-series).
Really good loudspeakers benefit a great deal from using the highest quality amplification, and this amp seems to be worthy of the best (IMHO). I look forward to the ucd700, since I have some power transformers that are too high a voltage for the ucd400's (is 77 VDC per rail enough for the ucd700?).
In a few weeks I will try this on an electrostatic setup, replacing an Audio Research M100 monoblock pair. I also hope to get my Martin Logan CLS speakers restored for a real test of the ucd400's
Future tests will include comparing the ucd700 to the ucd400. I may end up converting all my amplifiers with these if "long term enjoyment" is anything like my initial impressions. It would be nice to get away from occasional re-tubing and the power consumption, heat, and electric bills associated with the tube equipment.
@goudey,
Welcome to the club! May the UcD be with you! 😀
I'm going to put the Velleman kits on my amp, I just hope they're good enough when the going gets tough...
@Bruno,
If you have the time, can you look at these?
Velleman 4701
Is this any good, or am I ruining my sound quality with a device that won't work when failure occurs anyway?
I'm tempted to buy the HQ Hypex power supplie(s) but I have my BCC caps already 🙄
Welcome to the club! May the UcD be with you! 😀
I'm going to put the Velleman kits on my amp, I just hope they're good enough when the going gets tough...
@Bruno,
If you have the time, can you look at these?
Velleman 4701
Is this any good, or am I ruining my sound quality with a device that won't work when failure occurs anyway?
I'm tempted to buy the HQ Hypex power supplie(s) but I have my BCC caps already 🙄
Hi J-P and Bruno, any ideas when the new PS's will be available? they are still showing as "soon available" ?
Re: comments on sound quality
Congratulations, you are now officially infected with the Class D virus. Yes, these amps are killers.
Have fun
Gertjan
goudey said:
Congratulations, you are now officially infected with the Class D virus. Yes, these amps are killers.
Have fun
Gertjan
Quite succinctly put. Unless the relay is specified to break large currents in an inductive circuit, there's at least a significant chance of contact welding when push comes to shove. The "unpowered" connection scheme looks suspicious from a sonic perspective. It means the circuit uses the speaker signal...Yves Smolders said:
Re: comments on sound quality
It's designed for 90V (@ nominal mains voltage with no signal). It'll work fine at 77V but that means the supply voltage will be lower when you're driving it hard and you won't get the full 700W out of it. Maybe only 600 😀goudey said:
Hey Bruno,
How much longer will we have to wait for the long anticipated UCD700.
I am waiting in agony.
Thanks
Lawrence
How much longer will we have to wait for the long anticipated UCD700.
I am waiting in agony.
Thanks
Lawrence
The proto pcb's should be arriving anytime soon. These need to be stuffed and thoroughly debugged. Then a second PCB round is made to mop up oversights in the design and BOM and the inevitable errors committed while correcting those of the first round. The third round can then be produced in a larger series. The production 700 is still a couple of months away. I'm not sure how long a person can wait in agony without suffering nervous breakdown, so you might try to relax during that time. 😀lawbadman said:
Cheers.
Bruno Putzeys said:
Those modules are OK as long as the relay is upgraded to a better quality component. The relay is cheap and only rated to 10A. It being powered off the speaker is not an issue IMO as any extra or non-linear loading it may present whilst sat there in normal use is practically non existent compared to the impedance of the speaker.