Christer said:
Thyristor short is a good method. Must detect overload with a method that does not affect sound quality. Switching current transformer possibly.
Goal is to protect speaker, right?
Solid State outputs hard to protect and have best sound. These parts are cheap. If one of these fail, you still need to protect the speaker.
mwh-eng said:
Or a voltage divider on the output with a high Zin buffer (op amp,
JFET,...) followed by pretty much anything you need to detect DC
and trigger the thyristors.
Slone has an argument against crowbars, but that seems only
relevant for commercial amps, since it is no problem if you are
aware of it and do your own repair when broken. The point was
that often the thyristor blows too and needs to be replaced,
or it won't protect you a second time.
Agree too both points. Even an expensive lateral MOSFET or two
is coffee money compared to a speaker.
Another idle thought for output protection. Instead of having a relay in series with the speaker lead is it possible to have an open relay from the output to ground. In the event of a DC fault the relay could be closed shorting the output and protecting the speakers. This would seem to protect the integrity of both the power supply leads and the speaker leads (ie. no fuses or relays)and allow the current to blow an appropriate fuse on the primary side of the power supply. Jonathan Bright
" Instead of having a relay in series with the speaker lead is it possible to have an open relay from the output to ground. "
This sounds like it would protect the speakers but you would want to be sure you also have a good VI protection scheme that you really trust, lest you save the speakers at the cost of the output transistors. I think it is important that both the amplifier and the speakers get some kind of protection and the means are most likely different for each.
I don't know how difficult it would be to implement but being a bit conservative on these maters I would be interested in something that also disconects the mains if any other fault persists more than a set time. There must be a latching relay (or the equivalent) somewhere so that the mains wold stay disconected until you do a manual reset. This keeps you from ignoring the problem.
This sounds like it would protect the speakers but you would want to be sure you also have a good VI protection scheme that you really trust, lest you save the speakers at the cost of the output transistors. I think it is important that both the amplifier and the speakers get some kind of protection and the means are most likely different for each.
I don't know how difficult it would be to implement but being a bit conservative on these maters I would be interested in something that also disconects the mains if any other fault persists more than a set time. There must be a latching relay (or the equivalent) somewhere so that the mains wold stay disconected until you do a manual reset. This keeps you from ignoring the problem.
It should be possible to connect two mosfets back to back to form a solid state bidirectional switch. Then it could be controlled with an optocoupler. A small transformer would have to power the 12V floating supply to the mosfets. I remember making one of those bidirectional switches to act as both a rectifier and switching regulator, both mosfet gates of which were to be controlled by a single feedback circuit. It does seem too complicated for speaker protection for diyers though.
How about the idea of using a standard solid state relay to interrupt the connection to the speaker in case of overload. Either of these options may introduce too much series resistance anyway.
How about the idea of using a standard solid state relay to interrupt the connection to the speaker in case of overload. Either of these options may introduce too much series resistance anyway.
subwo1 said:
I like the back to back mosfet idea. Could use 2 each in parallel to lower the resistance or use an amp with negative output impedance to negate this resistance. How will you sense an overcurrent condition? Also, how fast of a response is needed to protect a small base driver voice coil? In the case the system is tri-amped with active crossovers, the tweeter could see a DC current overload.
Hi sam9,
Series inductors delay NFB node sensing of loudspeaker back emfs. Would need to be no more than 250nH to maintain 0.01% distortion specifications.
Hi jonathan,
Like your idea, any driving circuit ?
Once an amplifier is faulty it needs attention anyway. Your relay would be external, and gets us away from contact resistance waveform distortion.
Hi subwo1,
Lookin good ...... but ...... is there a switch-on surge ?
Can output devices not run at the +55V anyway, to simplify psu requirements ?
Series inductors delay NFB node sensing of loudspeaker back emfs. Would need to be no more than 250nH to maintain 0.01% distortion specifications.
Hi jonathan,
Like your idea, any driving circuit ?
Once an amplifier is faulty it needs attention anyway. Your relay would be external, and gets us away from contact resistance waveform distortion.
Hi subwo1,
Lookin good ...... but ...... is there a switch-on surge ?
Can output devices not run at the +55V anyway, to simplify psu requirements ?
Graham M. I can't help with a specific circuit as I don't understand enough of the basics of electronics but I can recall building a kit from a local mag' 20 yrs ago that was fairly simple and I think it did two things; (i) it delayed connecting the speakers until the dc conditions had stabilised and then (ii) disconnected them if dc subsequently appeared at the output. There was a suitably long time constant to avoid tripping it prematuely on suatained bass notes. I just thought that the latter function (ii) could be used to close the relay rather than open it. If you want to pursue things at your end of the world I imagine you have access to "E&WW". JL-H has published two relevant circuits. One is the components surrounding the relay on his 80 watt Mosfet design published around '83-'84. The other is the twin rail regulated PSU in the May 1989 edition. It fully protects its associated amp' by shutting down the PSU when dc appears at the output. Jonathan Bright
Hi Graham, You were right about the turn on surge. I guess I was wishfully thinking that the positive and negative charging currents would balance out the circuit then. I have moved the bypass capacitors to ground and added the resistor at the junction point to lower the ac gain. Thanks for your accolades.
Attachments
Hi Graham, I forgot to say that those mosfets are rated for only 100v. If 200v ones are used at higher voltage, one has to watch out for power dissipation and current rating. If mosfets are paralleled, then the gate resistors should be increased to a couple hundred ohms to prevent parasitic interaction between them. But then the problem is that the frequency response and sound quality decrease.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.