I found a cool thread in the archives where someone
didn't like to use speaker relays (amp out --> relay contacts -->
speaker) due to the arcing problems across the relay contacts
when there was DC offset present on the amplifier output,
as the inductance from the speaker caused back emf -> arc,
most notably, the woofers were the worse culprit that
caused this.
Rob Elliot's sound pages indicated that there is a fix
for this.
relay COM --> speaker
relay N.O. --> amplifier
relay N.C. --> GND
As the lever goes to the NC position, the arc is
shunted to gnd protecting the arc from passing
to the speakers.
True/false ?
So..
Do you still experience this arcing problem in a non-DC offset
scenario ? Perhaps the amplifier got too hot and your
protection circuit turned off the relay to disengage the speaker from the amplifier or some other protection circuit
enabled......
Will the voltage present at the speaker (the music)
kick back to the relay contacts causing arcing ?
if yes,
...then
......Can the protection circuit mute the audio input signal first,
......then a split second later, disable the output relay
......so there is no voltage across the speaker to cause
......the arcing ?
....else
.......cool
end.
If relay contacts arc, will the two contacts weld together
and not "open up" to disengage the speaker from output?
if no
...then
......If my relay is destroyed due to arcing, would not
......this be acceptable as the relay did the job of protecting
......the speaker?
....else
.......not cool
end.
didn't like to use speaker relays (amp out --> relay contacts -->
speaker) due to the arcing problems across the relay contacts
when there was DC offset present on the amplifier output,
as the inductance from the speaker caused back emf -> arc,
most notably, the woofers were the worse culprit that
caused this.
Rob Elliot's sound pages indicated that there is a fix
for this.
relay COM --> speaker
relay N.O. --> amplifier
relay N.C. --> GND
As the lever goes to the NC position, the arc is
shunted to gnd protecting the arc from passing
to the speakers.
True/false ?
So..
Do you still experience this arcing problem in a non-DC offset
scenario ? Perhaps the amplifier got too hot and your
protection circuit turned off the relay to disengage the speaker from the amplifier or some other protection circuit
enabled......
Will the voltage present at the speaker (the music)
kick back to the relay contacts causing arcing ?
if yes,
...then
......Can the protection circuit mute the audio input signal first,
......then a split second later, disable the output relay
......so there is no voltage across the speaker to cause
......the arcing ?
....else
.......cool
end.
If relay contacts arc, will the two contacts weld together
and not "open up" to disengage the speaker from output?
if no
...then
......If my relay is destroyed due to arcing, would not
......this be acceptable as the relay did the job of protecting
......the speaker?
....else
.......not cool
end.
If there is enough DC present to arc across a contact, the presence of a relay is the very least of your problems.
Regarding the remander of your post, what is this obsession with arcing? It takes mucho voltage to arc, and at normal listening levels it's not going to happen. Additionally, relays are designed to deal with a small bit of arcing...probably more than most amps are capable of generating.
Relay outputs have been used in very high-quality gear for many many years. There must be a reason why...
If you want a relay, use one. If you don't like them, then don't. I use them all the time with no ill effects (and I'm pretty picky...)
Regarding the remander of your post, what is this obsession with arcing? It takes mucho voltage to arc, and at normal listening levels it's not going to happen. Additionally, relays are designed to deal with a small bit of arcing...probably more than most amps are capable of generating.
Relay outputs have been used in very high-quality gear for many many years. There must be a reason why...
If you want a relay, use one. If you don't like them, then don't. I use them all the time with no ill effects (and I'm pretty picky...)
I use them all the time with no ill effects (and I'm pretty picky...)
Cool, I want to use a relay. Do you know with model #
or type you use?
I posted this topic after reading this 2 year old
thread.
http://www.diyaudio.com/forums/showthread.php?threadid=1549&highlight=relay
Silverpike said;
djk, you say in your reply that relays will arc and weld when tripped during overload. Have you acutally seen this happen?
I am curious to know, because in Slone's amp design book, he reccomends relays over crowbar circuits, which is in direct contrast to your advice.
It is hard for me to imagine a relay arcing. I say this because most amps I see have rails of 80V or less. I don't think 80V is enough to cause an arc. Wouldn't the short of the output cause the rails to drop potential because of the limited current supply also? Also, would the relay arc before the protection fuses blow?
In addition, the particular design I am building is based on lateral MOSFETs. Slone indicates in his book that the vast majority of lateral MOSFETs fail "soft", which is to say that when they go, they will short the gate to source instead of emitter to collector like BJTs. Does this failure mode have any relevance to relay arcing?
I would like to know more about your experiences.
djk said;
"djk, you say in your reply that relays will arc and weld when tripped during overload. Have you acutally seen this happen?" Are you kidding? More times than I can count.
"I am curious to know, because in Slone's amp design book, he reccomends relays over crowbar circuits, which is in direct contrast to your advice." So what.Do you want protection or not? "It is hard for me to imagine a relay arcing." In Electronics 101 they teach that an inductor will try to maintain the current through itself.If you try and interupt the current, ie open the relay while driving DC into an inductive load, the back EMF will try to rise to infinity.Is that high enough to arc? "Also, would the relay arc before the protection fuses blow? " A fast blow, AGC type fuse, will pass 200% of rated current for one hour.If correctly sized they will prevent fire after amplifier failure.I used to fix a lot of Crown DC300A, Peavey CS800, etc.When the customer brought them in the first thing I checked was the fuse.If it had the ABC ceramic fuse I told them it wasn't bad.If they had put in an AGC glass fuse I told them it would be a bunch of money.The fuse element melts and tries to open up under overload.Then because of the inductive load the voltage rises and an arc is struck through the metal vapor of the fuse.ABC fuses have powder in them to quench this arc.The extra energy carried by the arc wipes out emitter resistors, bridge rectifiers, and crowbar triacs that normally would not have failed. "Slone indicates in his book that the vast majority of lateral MOSFETs fail 'soft', which is to say that when they go, they will short the gate to source instead of emitter to collector like BJTs. " This is true but the lateral MOSFET is very expensive on a safe area watt per dollar bassis and I don't care for the sound of the bass of amps using these kind of outputs.
Silverpike said;
quote:
"I am curious to know, because in Slone's amp design book, he reccomends relays over crowbar circuits, which is in direct contrast to your advice." So what.Do you want protection or not?
Protection I would like, but I could sure do without the rudeness.
If my post came across as accusatory, then I apologise, I meant no offense. I am only trying to reconcile the two opinions.
Since you obviously have much experience with design reliability, you are one of the best resources I have seen on this forum. Your posts are informative, and include examples. However, the reason I am mentioning Slone is that he is no idiot. I am not going to dismiss an author with 30+ years of design experience in home and commercial audio lightly.
Does the inductance of the load affect the magnitude of the back EMF? Is this a problem with any speaker, or is there a greater risk with only certain kinds? Most crossover networks are not purely inductive. Does the capacitance have any additional effect?
djk said;
A relay keeps the amp quiet during turn on and turn off. A crowbar will protect the load from DC. You can build an amp with either or both. The amps that I have seen that had both seem to run forever because the designer didn't cut corners in other areas either. The only work of Sloan that I have seen is the 200W MOSFET job from Radio Electronics about three years ago. There was plenty of room for improvement in that design. The woofer is the big culprit for inductance.
***************** END ******************************
I'm seeking other opinions, just like a cancer patient
seeks another doctor. heh "yes, said the second doctor,
you are gonna die".
<-- /joke
The relay I found was a Magnecraft PCB mount relay.
DPDT with 20A contacts. Not miniature type, it's decent
size, about 2"L x 1.4"W x 1"H
I've seen the "other" smaller relays used in
some Levinson gear and Lamm amplifiers,
that particular relay is my second choice, not
as BeeFy as my Magnecraft...... hehe
I want to parallel the DPDT contacts for 40A rating
and perhaps double redudency .... just in case ...
Cool, I want to use a relay. Do you know with model #
or type you use?
I posted this topic after reading this 2 year old
thread.
http://www.diyaudio.com/forums/showthread.php?threadid=1549&highlight=relay
Silverpike said;
djk, you say in your reply that relays will arc and weld when tripped during overload. Have you acutally seen this happen?
I am curious to know, because in Slone's amp design book, he reccomends relays over crowbar circuits, which is in direct contrast to your advice.
It is hard for me to imagine a relay arcing. I say this because most amps I see have rails of 80V or less. I don't think 80V is enough to cause an arc. Wouldn't the short of the output cause the rails to drop potential because of the limited current supply also? Also, would the relay arc before the protection fuses blow?
In addition, the particular design I am building is based on lateral MOSFETs. Slone indicates in his book that the vast majority of lateral MOSFETs fail "soft", which is to say that when they go, they will short the gate to source instead of emitter to collector like BJTs. Does this failure mode have any relevance to relay arcing?
I would like to know more about your experiences.
djk said;
"djk, you say in your reply that relays will arc and weld when tripped during overload. Have you acutally seen this happen?" Are you kidding? More times than I can count.
"I am curious to know, because in Slone's amp design book, he reccomends relays over crowbar circuits, which is in direct contrast to your advice." So what.Do you want protection or not? "It is hard for me to imagine a relay arcing." In Electronics 101 they teach that an inductor will try to maintain the current through itself.If you try and interupt the current, ie open the relay while driving DC into an inductive load, the back EMF will try to rise to infinity.Is that high enough to arc? "Also, would the relay arc before the protection fuses blow? " A fast blow, AGC type fuse, will pass 200% of rated current for one hour.If correctly sized they will prevent fire after amplifier failure.I used to fix a lot of Crown DC300A, Peavey CS800, etc.When the customer brought them in the first thing I checked was the fuse.If it had the ABC ceramic fuse I told them it wasn't bad.If they had put in an AGC glass fuse I told them it would be a bunch of money.The fuse element melts and tries to open up under overload.Then because of the inductive load the voltage rises and an arc is struck through the metal vapor of the fuse.ABC fuses have powder in them to quench this arc.The extra energy carried by the arc wipes out emitter resistors, bridge rectifiers, and crowbar triacs that normally would not have failed. "Slone indicates in his book that the vast majority of lateral MOSFETs fail 'soft', which is to say that when they go, they will short the gate to source instead of emitter to collector like BJTs. " This is true but the lateral MOSFET is very expensive on a safe area watt per dollar bassis and I don't care for the sound of the bass of amps using these kind of outputs.
Silverpike said;
quote:
"I am curious to know, because in Slone's amp design book, he reccomends relays over crowbar circuits, which is in direct contrast to your advice." So what.Do you want protection or not?
Protection I would like, but I could sure do without the rudeness.
If my post came across as accusatory, then I apologise, I meant no offense. I am only trying to reconcile the two opinions.
Since you obviously have much experience with design reliability, you are one of the best resources I have seen on this forum. Your posts are informative, and include examples. However, the reason I am mentioning Slone is that he is no idiot. I am not going to dismiss an author with 30+ years of design experience in home and commercial audio lightly.
Does the inductance of the load affect the magnitude of the back EMF? Is this a problem with any speaker, or is there a greater risk with only certain kinds? Most crossover networks are not purely inductive. Does the capacitance have any additional effect?
djk said;
A relay keeps the amp quiet during turn on and turn off. A crowbar will protect the load from DC. You can build an amp with either or both. The amps that I have seen that had both seem to run forever because the designer didn't cut corners in other areas either. The only work of Sloan that I have seen is the 200W MOSFET job from Radio Electronics about three years ago. There was plenty of room for improvement in that design. The woofer is the big culprit for inductance.
***************** END ******************************
I'm seeking other opinions, just like a cancer patient
seeks another doctor. heh "yes, said the second doctor,
you are gonna die".
<-- /jokeThe relay I found was a Magnecraft PCB mount relay.
DPDT with 20A contacts. Not miniature type, it's decent
size, about 2"L x 1.4"W x 1"H
I've seen the "other" smaller relays used in
some Levinson gear and Lamm amplifiers,
that particular relay is my second choice, not
as BeeFy as my Magnecraft...... hehe
I want to parallel the DPDT contacts for 40A rating
and perhaps double redudency .... just in case ...
The standard relay used in stereo gear is this one, 24VDC, 4PDT, each contact rated at 5A. Up to about 100W output or so, one relay is used, with one channel using two contacts. Over 100W or so two relays are used (usually with the coil windings in series) and each relay has all four contacts used for one channel.
I have a 330WPC amp here (for instance) that uses two of these. Works beautiful.
The posts you referred to are interesting, but I've seen a lot of high power amps fry (used to be a roadie for a band, was a daily occurance to smoke an amp) and never, never seen a relay 'fuse' itself. What they posted was pure speculation and conjecture.
I recommended the relay above for two reasons:
1) It's what everyone else uses, and is a fairly common configuration. I'd shy away from using a relay that is going to be hard to find once you need a replacement.
2) It's cheap!! Buy a dozen of the things and you'l be set for life!
I have a 330WPC amp here (for instance) that uses two of these. Works beautiful.
The posts you referred to are interesting, but I've seen a lot of high power amps fry (used to be a roadie for a band, was a daily occurance to smoke an amp) and never, never seen a relay 'fuse' itself. What they posted was pure speculation and conjecture.
I recommended the relay above for two reasons:
1) It's what everyone else uses, and is a fairly common configuration. I'd shy away from using a relay that is going to be hard to find once you need a replacement.
2) It's cheap!! Buy a dozen of the things and you'l be set for life!
Thanks for the info. I'm gonna dremel
open my sealed relay to examine it
and check it on on the bench connected
to live speakers - hehe Torture
tests are fun.
open my sealed relay to examine it
and check it on on the bench connected
to live speakers - hehe Torture
tests are fun.
Relay
Another tip:
Check out the LRZ1 from www.amplimo.nl, this is a relay specifically developed for audio outputs. It has a main wolfram (tungsten?) 100A contact bridged with a gold contact for low-level fidelity. Both contacts open and make in reverse order. The best there is for this, and it will set you back all of 6.5 euros.
Jan Didden
Another tip:
Check out the LRZ1 from www.amplimo.nl, this is a relay specifically developed for audio outputs. It has a main wolfram (tungsten?) 100A contact bridged with a gold contact for low-level fidelity. Both contacts open and make in reverse order. The best there is for this, and it will set you back all of 6.5 euros.
Jan Didden
- Status
- Not open for further replies.