Hi, Series Caps on the amp output for speaker " protection " , yes or no ?. ( Ideea being - amp goes bad DC at output, cap = no dc , speakers are " happy " ) .
I know they are used in single supply rail amps with half of the rail voltage on them. that's why we have dual rail power on amps to have 0 ( nearly zero ) dc offset, so no cap needed = better sound?. In 2-way , 2-way speakers there's always a cap for the midrange, tweeter, unless bi-amped ( even in bi-amped cap on tweeter is recomended as I read ).
Let's say in a PA amp, 200W per channel, a 4700uF cap rated ( 80v or so ) for 8ohms it will cut off at around 4 Hz ?, To much current trough the cap?, or the only concern is " worse sound quality" ?.
- Bruno.
I know they are used in single supply rail amps with half of the rail voltage on them. that's why we have dual rail power on amps to have 0 ( nearly zero ) dc offset, so no cap needed = better sound?. In 2-way , 2-way speakers there's always a cap for the midrange, tweeter, unless bi-amped ( even in bi-amped cap on tweeter is recomended as I read ).
Let's say in a PA amp, 200W per channel, a 4700uF cap rated ( 80v or so ) for 8ohms it will cut off at around 4 Hz ?, To much current trough the cap?, or the only concern is " worse sound quality" ?.
- Bruno.
My impression is sound quality could be altered using the cap due to any stray inductance in the cap's foil winding structure, or and ESR the cap may have if it's high enough. Best protection I saw was a fuse that was included in the amplifier's feedback loop to minimize the effect of the fuse's internal resistance. A high value resistor was in parallel with the fuse to keep the amplifier's feedback loop from going crazy should the fuse open. I think a Hafler model did that.
Majority of commercial amplifier manufacturers have used a speaker protection relay circuit instead of capacitors in speaker output lines.
A single supply amp has the REQUIRED DC bias for a ~4700uF Electrolytic capacitor. A dual rail amp would require an AC/bipolar rated capacitor and there is no reasonable way to do that at for more than about 68uF. A DC detection relay circuit is a much better idea.
It may depend on what your goal is. Are you building your own design amplifier or repairing/improving some existing equipment?
Easy solutions would be to use a standard schematic from any commercial amplifier or buy a speaker protection module from eBay.
Another solution worth to consider might be a mosfet based circuit - like one discussed in this thread:
https://www.diyaudio.com/community/threads/speaker-protection-board.377896/
xXBrunoXx - You may want to consider fully complementary topologies because faults tend to cause the output voltage to go to zero rather than to a rail.
Ed
Ed
As long as the relay is rated for at least 10A, sticking contacts is unlikely. But the real threat is contact oxidization that fails to connect as the relay ages.
1. Mechanical relay contacts do not last forever.
2. FET solid state relays may cause enough distortion to offend purists.
3. TRIAC crowbar protection may cause further damage to the amp and there may be high frequency issues.
It's possible to shut down the power supply but no one does that.
1. Mechanical relay contacts do not last forever.
2. FET solid state relays may cause enough distortion to offend purists.
3. TRIAC crowbar protection may cause further damage to the amp and there may be high frequency issues.
It's possible to shut down the power supply but no one does that.
In most amplifiers from 70-ies and 80-ies that I have dealt with these contacts have not caused any hearable problems.
But passing 1 kHz through the relay sometimes makes it "sing" i.e. you can hear the tone coming from it.
I will most probably make some measurements in near future (mosfets just arrived) and post them in the thread linked in the previous post.
There may be a difference, there may not be a difference - time will tell...
That’s only true in simulation. It will still go to a rail with only a couple microvolts of real world offset voltage and loss of feedback.
Mechanical relays have as much contact resistance and DISTORTION due to the work functions of the metals involved as the FET ones do. They only reason to use them is/was because they are cheap. But manufacturers cheap out and use ones that are TOO small, resulting in welded contacts during a fault. The big ice cube ones didn’t do that. BUT what normally happens is an amp faults and blows the fuse. That results in circuit interruption before the relay is even triggered. Then there is DC at the speaker the NEXT time it’s powered up (after replacing the fuse) and the relay simply fails to close. Any cheapie will do that and be 100% effective. You do see all these threads around here “My amp is in protect mode, there’s no sound, what do I do?”.
wg_ski - My statement comes from mental simulation. Try it yourself - mentally disconnect a random node in the front-end. A differential pair with a VAS loaded by a current source will have a high probability of sending the output to a rail. A fully-complementary front-end can tolerate quite a few faults with zero output voltage provided that the VAS is loaded with a resistor to ground.
Ed
Ed