Since almost all power amp work as a giant OP-AMP if you feed only AC signal to the speaker you should not have a DC signal in the power supply, but you will get DC signal in the ceter tap of the power transformer if the amp feed DC signal to the speaker. Detecting DC signal in the center tap and using this to cut power supply output should be possible.
I have started a desing ( uncomplete ).
T1, T2 & R1 can detect DC signal in groud path.
A pair of low saturation voltage FET can be used as " missing device "
There is still some desing to be made ........................
With this configuration you no component ( relay ) in signal path.
I have started a desing ( uncomplete ).
T1, T2 & R1 can detect DC signal in groud path.
A pair of low saturation voltage FET can be used as " missing device "
There is still some desing to be made ........................
With this configuration you no component ( relay ) in signal path.
Attachments
Hi,
interesting idea, but it is certainly not universally applicable. Any single ended stage (common source) between ground and + rail will send current into the center tap.
While I know many don't like a protection in their signal path, cutting the psu this way will allow the secondary filter caps of one polarity (to the right) to discharge into your speaker. Better to cut the line to the speaker first IMHO.
Hannes
interesting idea, but it is certainly not universally applicable. Any single ended stage (common source) between ground and + rail will send current into the center tap.
While I know many don't like a protection in their signal path, cutting the psu this way will allow the secondary filter caps of one polarity (to the right) to discharge into your speaker. Better to cut the line to the speaker first IMHO.
Hannes
I had some spare time so I made a quick test today
I simply add a 100 ohms resistor in the center tap of the transformer in a power amp.
The amplifier is a simple killer and I use it in a resitive load of 7.5 ohms and drive it up to 25 or 30 watts peak with music ( FM tuner )
Using an analog voltmeter acros the resistor readings did not go over 10 Mv (DC) using a digital voltmeter reading remain same average value.
Using an oscilloscope we can see a pulsating signal up to 0.6 volt , synchronised with rectifiers current pulse.
Across the resistor there was also a lot of noise ( i think it was rectifier noise )
I simply add a 100 ohms resistor in the center tap of the transformer in a power amp.
The amplifier is a simple killer and I use it in a resitive load of 7.5 ohms and drive it up to 25 or 30 watts peak with music ( FM tuner )
Using an analog voltmeter acros the resistor readings did not go over 10 Mv (DC) using a digital voltmeter reading remain same average value.
Using an oscilloscope we can see a pulsating signal up to 0.6 volt , synchronised with rectifiers current pulse.
Across the resistor there was also a lot of noise ( i think it was rectifier noise )
Yes he had a mysterious (small) black box that lived on the centre tap for DC protection. Never tested its behaviour.
Anyone tested this?
What happen when the current to the speakers go through here?
Well, the one transistor will trig when the positive pulses is present, and the other transistor will trig when the negative pulses is present.
That is, when the pulses are so big that they produce enough current to make 0,7V over the resistor. I rather use a device that meassures directly to the speaker output, filtered to disregard AC. Easier, and totally without influence to the sonics.
How to protect the speaker? Relay in series with the speaker lead is one, Relay in series with the fuses in to the amp, or as B&O used to do it: Trig a thyristor in the PSU to shortcirquit the PSU, and thereby stop any DC from getting to the speaker.
What happen when the current to the speakers go through here?
Well, the one transistor will trig when the positive pulses is present, and the other transistor will trig when the negative pulses is present.
That is, when the pulses are so big that they produce enough current to make 0,7V over the resistor. I rather use a device that meassures directly to the speaker output, filtered to disregard AC. Easier, and totally without influence to the sonics.
How to protect the speaker? Relay in series with the speaker lead is one, Relay in series with the fuses in to the amp, or as B&O used to do it: Trig a thyristor in the PSU to shortcirquit the PSU, and thereby stop any DC from getting to the speaker.
The concept is sound but practically there are several issues.
As mentioned before, the AC currents will trigger the detection transistors. The problem is how to distinguish a (small) DC current from a quite large AC current.
What you could do is to low-pass filter the voltage across the detection resistor before you 'look' at it. (Which of course is exactly what you would do if you wanted to detect DC on the speaker).
Basically you are mixing up two things, detection and protection. You can much easier detect DC at the speaker output, so you don't need the resistor in the supply line, and then you can still do the protection in the supply lines if you want that.
But, there is a reason why nobody does that this way...
jan
Member
Joined 2009
Paid Member
Jan, the rail caps will act as filters - you need only look at the signal across the resistor in the ground return and size that resistor appropriately. Post 3 shows the results of doing the experiment. Even large AC signal is a non issue because there is still some rail cap to take care of that before the gnd current sensing resistor.
SAGE
I can Exclusively reveal for the 1st time, that Les Sage of Sage Audio, both told me & showed me, what was in the little potted black boxes of his for DC protection. It was a fast acting Polyswitch !
Some years later, one of Maplins magazines ampifier designs also included a Polyswitch for the same reason.
I can Exclusively reveal for the 1st time, that Les Sage of Sage Audio, both told me & showed me, what was in the little potted black boxes of his for DC protection. It was a fast acting Polyswitch !
Some years later, one of Maplins magazines ampifier designs also included a Polyswitch for the same reason.
Member
Joined 2009
Paid Member
You have to make sure you rail caps are rated for rail-to-rail voltages not just rail-to-ground because when that poly opens the shorted rail collapses and the full rectifier voltage appears across the remaining rails caps. Other than that, I like this approach.
Sounds right. They were rated for the expected max power 250w / 500w etc. and required power to be removed to reset.
- Status
- Not open for further replies.