We have previously published a Simple Speaker Protection (SSP) for amplifiers with balanced outputs :
https://www.diyaudio.com/community/...power-amplifier-with-balanced-outputs.244157/
But how about Class D amps with single supply and maybe also BTL outputs, or car amps such as TDA7850 ?
This is an attempt to find a single solution for all.
Not proven, but useful basis for discussions.
Hopefully someone has better ideas, or can make it even simpler.
Patrick
.
https://www.diyaudio.com/community/...power-amplifier-with-balanced-outputs.244157/
But how about Class D amps with single supply and maybe also BTL outputs, or car amps such as TDA7850 ?
This is an attempt to find a single solution for all.
Not proven, but useful basis for discussions.
Hopefully someone has better ideas, or can make it even simpler.
Patrick
.
Last edited:
Audio Express September 2003, Jene Keceli author
If step on AE toes, please remove attachment
If step on AE toes, please remove attachment
Good find. I have no access to AE, unfortunately.
Minor differences in the use of normal, lower cost optocouplers in post #1.
But very intelligent use of the NPN for slow start.
Cheers,
Patrick
Minor differences in the use of normal, lower cost optocouplers in post #1.
But very intelligent use of the NPN for slow start.
Cheers,
Patrick
Last edited:
Here Version 2. Perhaps not much simpler ?
Further suggestions for simplification most welcome.
Patrick
Further suggestions for simplification most welcome.
Patrick
In the current Mouser inventory, the lowest cost dual optocoupler with sufficient CTR that I can find is the LTV-825.
One for each channel.
Other better suggestions most welcome.
Patrick
One for each channel.
Other better suggestions most welcome.
Patrick
We have been asked about this.
Apparently there seems to be a demand for speaker protection modules.
Of course if you accept normal mechanical relays, you can just use those in post #2.
Those who have subscribed for the Le Monstre 2024 GB would have seen a MOSFET relay version of the same basic idea.
This uses the same circuit as our proven low distortion MOSFET relays.
www.diyaudio.com/community/threads/hiraga-le-monstre-2024.413301/post-7818683
www.diyaudio.com/community/threads/distortion-measurements-of-a-mosfet-relay-for-loudspeaker-protection.370678/
That was tailored for +/-12V rails but extendable.
I have even replaced the protection circuit on my own F5X in favour of this Le Monstre version.
But we are refining that further with a version which is as universal as possible.
As mentioned in the PDF, it should be able :
Able to cope with +/-50V rails, single ended or balanced output
Or +100V single rail with BTL output
Supplied directly by the same power supply to the amplifier
Latch in protection mode once triggered
Current consumption below 15mA
Left & Right version with the same PCB for direct mounting next to the amplifier
All components easily available, no integrated circuits except for opto MOSFET drivers
The only exceptions are single rail amplifiers with capacitor or transformer coupled outputs.
But then they won't need any additional DC protection. 😊
PCBs have been ordered to verify the said refinements.
As soon as this is completed, we shall make it available, independent on amplifier projects, as has been the case till now.
Patrick
PS
Apart from all the advantages of MOSFET relays, it must be noted that they fail by short circuit.
It is more unusual for mechanical relays to fail when breaking open.
Last edited:
How expensive, I was asked.
Because of the simplicity, I counted current price of the components (with no quantity discount) at DK to be less than 20 USD.
And simplicity also means less component to fail, i.e. higher reliability.
Patrick
Because of the simplicity, I counted current price of the components (with no quantity discount) at DK to be less than 20 USD.
And simplicity also means less component to fail, i.e. higher reliability.
Patrick
The relay needs a flyback diode to protect Q1. This will provide a turn-on delay but no immediate turn-off. It is unlikely that isolation is needed so it would be simpler to use transistors instead of optos. Simplicity may appeal to some, but I think cost and coverage of other problems like overcurrent are more important.
Over current, over temperature, etc. are amplifier design specific and can either be part of the amplifier circuit, or in the power supply, or both.
Of course we have working prototypes for those.
But they are not universally applicable.
Happy New Year,
Patrick
Of course we have working prototypes for those.
But they are not universally applicable.
Happy New Year,
Patrick
+/-85V rails means that all the devices used have to be rated at 250V or above, to be absolutely on the save side.
In addition, 15mA means 2.5W dissipation.
Can it be done technically ?
Certainly, but it will not be such a compact unit.
And the choices of MOSFETs are quite limited.
Patrick
In addition, 15mA means 2.5W dissipation.
Can it be done technically ?
Certainly, but it will not be such a compact unit.
And the choices of MOSFETs are quite limited.
Patrick
Will the opto transistor adequately short the base of the switching transistor to ground?
Some optos have a pretty high saturation voltage. In the past I've added a couple of series diodes
between the emitter and ground so there is a more definite turn-off when the base is shorted to ground.
Some optos have a pretty high saturation voltage. In the past I've added a couple of series diodes
between the emitter and ground so there is a more definite turn-off when the base is shorted to ground.
Of course.
Prototypes soldered, ready for testing this weekend after business trip.
Patrick
Yes, because the switching is done at a MOSFET gate which is on at ~2V.
The optocoupler used has a saturation voltage of maximum 0.4V.
Patrick
