Hi jacco
If the op amp are correctly used,they are very high reliable.
Never i view a op amp failure in normal work conditions.
In more years of designing,two times an op amp of my project had fails:during a thunderstorm(1 op amp over 5 in a preamp first time and 3 over 5 during the second thunderstorm).
Do not preoccupe for the reliability of this circuit:the first I built is in 1984 and it work today:never i have made any substitution
If the op amp are correctly used,they are very high reliable.
Never i view a op amp failure in normal work conditions.
In more years of designing,two times an op amp of my project had fails:during a thunderstorm(1 op amp over 5 in a preamp first time and 3 over 5 during the second thunderstorm).
Do not preoccupe for the reliability of this circuit:the first I built is in 1984 and it work today:never i have made any substitution
I'm building a mosfet-circuit that brakes DC-supply voltage from the poweramp when protectioncircuit sends a protect signal (V+)
Schematic:
The good thing is that mosfets are faster and more reliable than relays because there is no mechanical function. Also Relays for big DC-currents are expensive and takes a lot of space..
http://files.upl.silentwhisper.net/upload6/Amp_PSU.pdf
I'm gonna etch the PCB for this circuit today and someday this week the components will come so I can't tell yet if it works good or not though.
Schematic:
The good thing is that mosfets are faster and more reliable than relays because there is no mechanical function. Also Relays for big DC-currents are expensive and takes a lot of space..
http://files.upl.silentwhisper.net/upload6/Amp_PSU.pdf
I'm gonna etch the PCB for this circuit today and someday this week the components will come so I can't tell yet if it works good or not though.
Hi Anatech!
Some large caps after mosfets would slow down things over mosfets somewhat and keep impedance low but also carry a hefty charge in case of failure.
If I have to choose myself I want as low rail impedance as possible with good quality caps very close to last stage.
I tend to agree with you, but as you say, give it a try.
Some large caps after mosfets would slow down things over mosfets somewhat and keep impedance low but also carry a hefty charge in case of failure.
If I have to choose myself I want as low rail impedance as possible with good quality caps very close to last stage.
I tend to agree with you, but as you say, give it a try.
Hi,
is a FET in line to the amp worse than a fuse? Both will show an effective resistance with the attendant volts drop. Both will modulate the voltage arriving at the amp in time with the current pulses.
What about the alternative if semiconductors in line are so abhorent.
Use the in line fuses followed by a pair of FET crowbars.
How easy to design this circuit?
Back to that PDF.
is K1 an opto isolated thyristor?
What limits the voltage on the 470uF delay capacitor?
is a FET in line to the amp worse than a fuse? Both will show an effective resistance with the attendant volts drop. Both will modulate the voltage arriving at the amp in time with the current pulses.
What about the alternative if semiconductors in line are so abhorent.
Use the in line fuses followed by a pair of FET crowbars.
How easy to design this circuit?
Back to that PDF.
is K1 an opto isolated thyristor?
What limits the voltage on the 470uF delay capacitor?
I dont like the power fet protection at all
We are after spk protection if a fet fails for whatever we have DC
No good !!!
I prefer the crowbar but I have found that end users will try to defeat these (bigger fuses)
I have used the crowbar a few times
To me the best method is to bite the bullet and use a high power Relay/Contactor.
I normally use a circuit the gives me a time delay before connecting the spk (no switch on thump).
I then use a circuit to monitor the dc offset so that above a few volts and a time of no more than 0,2 sec the contactor is forced off.
The advantage here is that if there is a dc offset(thump) on switch on the spk is not connected !!
So virtually any device can fail and the spk is safe! I know this may seem primitive but to be honest I have seen many devices fail in the feild despite sophisticated protection ideas
As for the Mosfet ideas there,s nearly enough parts to build a spare amp.
best regards Trev
We are after spk protection if a fet fails for whatever we have DC
No good !!!
I prefer the crowbar but I have found that end users will try to defeat these (bigger fuses)
I have used the crowbar a few times
To me the best method is to bite the bullet and use a high power Relay/Contactor.
I normally use a circuit the gives me a time delay before connecting the spk (no switch on thump).
I then use a circuit to monitor the dc offset so that above a few volts and a time of no more than 0,2 sec the contactor is forced off.
The advantage here is that if there is a dc offset(thump) on switch on the spk is not connected !!
So virtually any device can fail and the spk is safe! I know this may seem primitive but to be honest I have seen many devices fail in the feild despite sophisticated protection ideas
As for the Mosfet ideas there,s nearly enough parts to build a spare amp.
best regards Trev
Wow, thanks for all your answers!
K1 is a relay. Multisim 7 has an odd symbol for it..
I understand this because at first I did like the Relay/contactor idea more. The good thing with relay is that if the whole amp and psu blows completely the speaker is still safe. But still, the mosfet/psu version is alot better than not no have any protection at all which i don't have now. But who knows, I might rebuilt this one in a relay version. But as said, im gonna give the mosfet version a try
I Etched the PCB this evening, It wen't quite good. The only bad thing is that the printing was a little faded on the black so I had to reinforcement some large copper areas (for the Dc currents between mosfets and connector) . But it's ok. I'll show it when it's finished. Hopefully components will show up on wednesday.
K1 is a relay. Multisim 7 has an odd symbol for it..
I understand this because at first I did like the Relay/contactor idea more. The good thing with relay is that if the whole amp and psu blows completely the speaker is still safe. But still, the mosfet/psu version is alot better than not no have any protection at all which i don't have now. But who knows, I might rebuilt this one in a relay version. But as said, im gonna give the mosfet version a try
I Etched the PCB this evening, It wen't quite good. The only bad thing is that the printing was a little faded on the black so I had to reinforcement some large copper areas (for the Dc currents between mosfets and connector) . But it's ok. I'll show it when it's finished. Hopefully components will show up on wednesday.
Hi,
relays cutting a DC offset from output to speaker need a very large current rating for this duty.
The contacts can arc or even worse, weld closed on high current.
Then your assumed safe protection has gone.
Close rated fuses in the supply rails before the amp output stage and let the NFB correct for modulation of the supply rails.
relays cutting a DC offset from output to speaker need a very large current rating for this duty.
The contacts can arc or even worse, weld closed on high current.
Then your assumed safe protection has gone.
Close rated fuses in the supply rails before the amp output stage and let the NFB correct for modulation of the supply rails.
Hi to all
the reasons for adding a protection circuits is not only to protect from amplifier fail but also to eliminate the ON transient when power on.
The mosfets protection are very good to protect from a DC offset in the amp output's but it does not work for bump on power on.
A problem is inside a Mos used like switch:your internal on resistance is not linear with frequency,so it is not a good solution for high end amp.
Also to use the capacitors after this Mos is not a good thing because when the protection is on and the power supply is disconnected,the charge of this cap flows on the speakers.
Another problem:
it is difficult to find a low on resistance for P channels mosfets,and the resistance are different for N and P.
For the P polarity in case of high power amp their dissipation cannot be neglected.
the reasons for adding a protection circuits is not only to protect from amplifier fail but also to eliminate the ON transient when power on.
The mosfets protection are very good to protect from a DC offset in the amp output's but it does not work for bump on power on.
A problem is inside a Mos used like switch:your internal on resistance is not linear with frequency,so it is not a good solution for high end amp.
Also to use the capacitors after this Mos is not a good thing because when the protection is on and the power supply is disconnected,the charge of this cap flows on the speakers.
Another problem:
it is difficult to find a low on resistance for P channels mosfets,and the resistance are different for N and P.
For the P polarity in case of high power amp their dissipation cannot be neglected.
parsecaudio said:
Hi Parsec
The intent of my comment for the use a a few discrete transistors compared to a single quad op-amp is not for cost or number of separate components (this is DIY not commercial stuff here) but to have a very high reliability due to the limited number of compoenents to get a high Mean Time Between Failure number.
A quad op-amp like the TL084 may have something like maybe 40 transistors... (I have not checked the exact number here...)
jacco vermeulen said:
Hi Jacco
yes 4 times (provided that all individual op-amp of the Quad are used)!
Also. I like the circuit of post #23. I do not remember what is the current high limit of the IRF540 (is it high enough for high current demands?).
Hi Andrew,
I've seen a few welded relays over the years, but they are in the minority. I've also seen TO-3 transistors with flame cuts in the casing and others where the arc folled the emitter lead (gone) into the socket.
I'll take relay protection any day. Mute the input at the same time and you will reduce pitting on the output relay with every turn off.
Most amplifiers use fuses before the rectifiers where they may help with lowering peak charging currents. A good thing in my book. I doubt they are audible in a properly designed amp. In a poorly designed amp, everything is audible.
Crowbars can be very effective, but they almost guarantie a service job, and not a cheap one either. Still, cheaper than some speakers.
-Chris
I've seen a few welded relays over the years, but they are in the minority. I've also seen TO-3 transistors with flame cuts in the casing and others where the arc folled the emitter lead (gone) into the socket.
I'll take relay protection any day. Mute the input at the same time and you will reduce pitting on the output relay with every turn off.
Most amplifiers use fuses before the rectifiers where they may help with lowering peak charging currents. A good thing in my book. I doubt they are audible in a properly designed amp. In a poorly designed amp, everything is audible.
Crowbars can be very effective, but they almost guarantie a service job, and not a cheap one either. Still, cheaper than some speakers.
-Chris
Hi,
If the amp had gone wrong then the only extra cost is two fuses.
If the amp had spuriously triggered, then the total cost is two fuses and some time to replace them.
With all these topological alternatives, the prime purpose is to limit damage extending beyond the first failure. That limitation includes saving speakers and transformers. Most schemes will achieve at least some of these benefits. Is there a simple (cheap) scheme that achieves all the benefits?
I forgot about that one, thanks for the reminder.
if the crowbar was needed, it is time for service anyway, even if only to adjust the sensitivity of the trigger level.
If the amp had gone wrong then the only extra cost is two fuses.
If the amp had spuriously triggered, then the total cost is two fuses and some time to replace them.
With all these topological alternatives, the prime purpose is to limit damage extending beyond the first failure. That limitation includes saving speakers and transformers. Most schemes will achieve at least some of these benefits. Is there a simple (cheap) scheme that achieves all the benefits?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.