@richie00boy,
What relay would you put in there? It's easy to solder on these boards so I might give it a try.
If it doesn't "feel right", i'm going with the high-quality power supplies from Hypex - maybe Jan-Peter could ship me a version without caps but with a couple of short cables so I can use my existing caps.
What relay would you put in there? It's easy to solder on these boards so I might give it a try.
If it doesn't "feel right", i'm going with the high-quality power supplies from Hypex - maybe Jan-Peter could ship me a version without caps but with a couple of short cables so I can use my existing caps.
Yves, you need to brush up your memory 😉
http://www.diyaudio.com/forums/showthread.php?postid=581423#post581423
http://www.diyaudio.com/forums/showthread.php?postid=581423#post581423
Schrack relay RS Part No:198-6977
Hello Richie00boy
Can I ask you what qualifies the Schrack relay RS Part No:198-6977 as being suitable for speaker protection on an amplifier with a DC power supply of +/- 60V DC. Have you tested it with fault condition with 60V DC into 4 ohm load, and tried to break the DC condition from the load.
Regards
Arthur
richie00boy said:
Hello Richie00boy
Can I ask you what qualifies the Schrack relay RS Part No:198-6977 as being suitable for speaker protection on an amplifier with a DC power supply of +/- 60V DC. Have you tested it with fault condition with 60V DC into 4 ohm load, and tried to break the DC condition from the load.
Regards
Arthur
So I'm thinking of going class D. These modules sound like they are very nice. I've got an old Adcom GFA7000 5 channel amp with a couple of dead channels. The transformer has 5 sets of outputs for the original 5 channel. 125watts x 5 bipolar design. My only problem is I don't know what the wattage of the transformer is. Rail voltage is +/- 65 volts dc. Anybody know anything about this amp? Can I use the bones from this amp for a 5 channel class D using these modules.
Thanks, Dave
Thanks, Dave
Re: Schrack relay RS Part No:198-6977
The part is rated at 4kVA switching capacity. Of course this is for a resistive load, but there should be adequate capacity to survive at least one highly inductive & high energy switch in an audio amp situation. The relay would then be replaced along with the rail fuses, to ensure maximum reliability. A £5 relay is cheaper to replace than treasured speakers, and it would be foolish to skimp on replacing it.
I must confess to not having fully tested the part up to the rails you mention with a highly inductive load, although it is on my list of things to do 🙂
PHEONIX said:
The part is rated at 4kVA switching capacity. Of course this is for a resistive load, but there should be adequate capacity to survive at least one highly inductive & high energy switch in an audio amp situation. The relay would then be replaced along with the rail fuses, to ensure maximum reliability. A £5 relay is cheaper to replace than treasured speakers, and it would be foolish to skimp on replacing it.
I must confess to not having fully tested the part up to the rails you mention with a highly inductive load, although it is on my list of things to do 🙂
Speaker DC protection
Hi guys,
I was thinking about building a dual mono amp based on the UCD400. I read lots of posts and concerns about speaker protection from DC if the MOSFET(s) blow. Unlike input singals where read relays can be used with little degradation to sound quality, I am not keen using big power relays in series with the output of the power amp. This relay will surely degrade the sound, it may be OK when high current is passing through but during low level listening or quite parts of the music will effect the sound quality. I was therefore thinking of putting the relay between the output of the amp and ground. If the output MOSFET blows up the relay will be energised shorting the output to ground and protecting the speaker(?). A second relay could simultaneously disconnect the mains and shut down the unit. What do you think?
Hi guys,
I was thinking about building a dual mono amp based on the UCD400. I read lots of posts and concerns about speaker protection from DC if the MOSFET(s) blow. Unlike input singals where read relays can be used with little degradation to sound quality, I am not keen using big power relays in series with the output of the power amp. This relay will surely degrade the sound, it may be OK when high current is passing through but during low level listening or quite parts of the music will effect the sound quality. I was therefore thinking of putting the relay between the output of the amp and ground. If the output MOSFET blows up the relay will be energised shorting the output to ground and protecting the speaker(?). A second relay could simultaneously disconnect the mains and shut down the unit. What do you think?
Yes, that is one way of doing it. I'm just curious why you think a powerful relay will a) degrade sound and b) pass larger signals different to smaller ones. If anything a powerful relay will have much lower contact resistance and nice conductivity, making it better than a small relay for sound quality.
There are some cheap relays out there that use part of the magnetic system as a conductor which can be non-linear for sound, but you can get good relays quite easily.
There are some cheap relays out there that use part of the magnetic system as a conductor which can be non-linear for sound, but you can get good relays quite easily.
Re: Speaker DC protection
I expect firework, shorting this to ground will give huge currents with the power supply caps discharging via the amp and relay. Probably the relay will be damaged (contacts welded together??) by this act. Then I would prefer to put the relay in the DC rails and leave it with that, so no relays in between amp and speaker. Note that when you put the relay in the DC rail, you probably need to put a cap (value??) over the relay contact to avoid arcing when you try to brake the contact while a large current is flowing.
Maybe Bruno knows what value should be used.
Best regards
Gertjan
greg filip said:
I expect firework, shorting this to ground will give huge currents with the power supply caps discharging via the amp and relay. Probably the relay will be damaged (contacts welded together??) by this act. Then I would prefer to put the relay in the DC rails and leave it with that, so no relays in between amp and speaker. Note that when you put the relay in the DC rail, you probably need to put a cap (value??) over the relay contact to avoid arcing when you try to brake the contact while a large current is flowing.
Maybe Bruno knows what value should be used.
Best regards
Gertjan
Small signal relays use reed elements, you can not get reed relays to pass large currents. To pass large currents you will have to use automotive type or similar relays which do not perform well with small signals (quite parts of the music).
Gertjan,
I agree that you are going to have a large current flowing thourgh the output stage and the relay but we are talking about a catastophic failure. The relay would propably survive assuming it is large enough.
I agree that you are going to have a large current flowing thourgh the output stage and the relay but we are talking about a catastophic failure. The relay would propably survive assuming it is large enough.
Peavey had a pretty cool idea for their CS-800 amplifiers. They mounted a TO-220 triac between speaker hot and ground, where the gate was fed by the speaker output going through a lowpass RC network with a corner frequency somewhere around 5 Hz, if memory serves. Since the amps had a highpass corner frequency of 30 Hz (proaudio, you know the drill), the triac wouldn't trigger with normal signals. On DC present on the output, however, the DC would pass through the LPF at the gate in less than a half-second, turn on the triac, which would short the speaker output to ground, then eat a lot of current and self-destruct shorted to ground, thus protecting the speaker.
I thought it was weird to use an active device as an anti-fuse to protect a speaker, but the circuit worked marvelously well. The triacs were dirt cheap to replace, and also a damn sight simpler than reconing a speaker, I can tell you.
Likely the time constants would need to be rejiggered for home audio use - it'd be a pain to replace the triac every five minutes of "Terminator 2" - but the principle is likely still usable.
Francois.
I thought it was weird to use an active device as an anti-fuse to protect a speaker, but the circuit worked marvelously well. The triacs were dirt cheap to replace, and also a damn sight simpler than reconing a speaker, I can tell you.
Likely the time constants would need to be rejiggered for home audio use - it'd be a pain to replace the triac every five minutes of "Terminator 2" - but the principle is likely still usable.
Francois.
DSP_Geek said:
Why use these destructive methods if you can just put the relay in the DC rails??? There is some decoupling on the UcD modules, so I would expect no sonic degradation from the relay. A relay is probably more expensive than a triac. On the other hand, with the triac approach you would also blow fuses that also cost money. And yes, triggering on home-theater like signals would be an issue. This is probably an issue for any DC protection. The perfect (DC) protection would be a protection that either measures the cone position of the woofers (too expensive) or tries to guestimate it based on the signal going in the woofers and cut the signal when the maximum cone excursion is reached. With a closed system, that would probably be pretty easy to do, may think about it. Hereby documented and not patentable anymore (probably somebody in pro-audio already thought of it anyway).
Best regards
Gertjan
ghemink said:
Note the relay would have to break a lot of current - easily 10 or 15 amps for a 400W/side amplifier driving 4 ohm speakers. There must be no possibility of contacts welding together under load or current interruption, and you don't know whether the relay itself will work until it's needed.
You said it: a good relay is more expensive than a couple of small triacs, especially since the rest of the circuit would have to be duplicated anyway. The cost does add up over Peavey's production volumes. As for fuses, you'd need them with a relay anyway, so that's a moot point.
In any event, this circuit is insurance - you hope it's never needed, but it's nice to have anyway. Even though it might be destructive, the default mode of an amplifier module is continued operation, so the vast majority of these failsafes are never invoked.
Francois.
Hi,
in class D amplifiers DC failure sensing is very easy to implement. All you have to watch is switching action of output stage. If you don't have change in polarity for some miliseconds (to count in for some heavy clipping), then you can safely trigger DC protection.
Best regards,
Jaka Racman
in class D amplifiers DC failure sensing is very easy to implement. All you have to watch is switching action of output stage. If you don't have change in polarity for some miliseconds (to count in for some heavy clipping), then you can safely trigger DC protection.
Best regards,
Jaka Racman
What if...
About DC protection:
Suppose that the DC protection circuit is powered from the positive rail voltage. Now suppose that the positive side of the PSU fails, for example the rectifier and/or a fuse blows. The positive rail voltage will be gone and the UCD module will be powered with only 0V and -60VDC. Will there be a negative DC-voltage on the speaker output in a situation like this?
Should I make an independent PSU for the DC protection circuit to be on the safe side? Or is the module configured to close down in a situation where one of the voltages is missing?
About DC protection:
Suppose that the DC protection circuit is powered from the positive rail voltage. Now suppose that the positive side of the PSU fails, for example the rectifier and/or a fuse blows. The positive rail voltage will be gone and the UCD module will be powered with only 0V and -60VDC. Will there be a negative DC-voltage on the speaker output in a situation like this?
Should I make an independent PSU for the DC protection circuit to be on the safe side? Or is the module configured to close down in a situation where one of the voltages is missing?
Re: What if...
Both the UcD180 and UcD400 shut down if one of the voltages is missing (no DC output), and will work afterwards without any hiccups. I actually tested that by mistake when the one of the secondaries on my trafo broke loose from the bridge rectifier.
thomsva said:About DC protection:
Suppose that the DC protection circuit is powered from the positive rail voltage. Now suppose that the positive side of the PSU fails, for example the rectifier and/or a fuse blows. The positive rail voltage will be gone and the UCD module will be powered with only 0V and -60VDC. Will there be a negative DC-voltage on the speaker output in a situation like this?
Should I make an independent PSU for the DC protection circuit to be on the safe side? Or is the module configured to close down in a situation where one of the voltages is missing?
Both the UcD180 and UcD400 shut down if one of the voltages is missing (no DC output), and will work afterwards without any hiccups. I actually tested that by mistake when the one of the secondaries on my trafo broke loose from the bridge rectifier.
Jaka Racman said:
But that only works when the output is clipping, if for some reason you have like 40V DC at the output, for example due to DC at the input and having removed the DC blocking cap, then 10A DC will flow in the case of 4Ohm speakers which will probably fry the woofers. On the other hand, it is an interesting method to detect clipping.
Best regards
Gertjan
Hi,
picture shows what I had in mind. By using slow optocouplers with 15us turn off time, both optocouplers are conducting while amplifier switches. When one Mosfet is on for more than 15us (due to the failure or clipping), one optocoupler turns off and logic signal DC_Protection goes high. But you are right about coupling caps. If you remove them, think about the consequences. I think they are originally designed in for a reason.
Best regards,
Jaka Racman
picture shows what I had in mind. By using slow optocouplers with 15us turn off time, both optocouplers are conducting while amplifier switches. When one Mosfet is on for more than 15us (due to the failure or clipping), one optocoupler turns off and logic signal DC_Protection goes high. But you are right about coupling caps. If you remove them, think about the consequences. I think they are originally designed in for a reason.
Best regards,
Jaka Racman
