Current limiting behavior??

[stoc005]

Just for fun I inputted the entire Leach amp into LTSpice and wanted to see how the current limiter worked. With a 1 ohm resistive load, the current limiter does work but it is oscillating. When it goes into overcurrent, the drive is "removed" which makes the current in the outputs decrease. Once the current decreases, the drive is allowed to increase the output current, which makes the current limiter reduce the drive again. So, I get around a 10KHz oscillation when the amp is driven into current limiting by a 1KHz signal. Anybody seen this before? Does the real amp do this? I'd hope not. But it makes sense since there is no hysteresis on the current limiter transistors. To me, this is just not a well behaved current limiter and lots of amps use this technique. Comments??

[richie00boy]

That's normal behaviour.

[darkfenriz]

Why worry about short circuit stability

[stoc005]

Because I do not want a large amount of high frequency "hash" going into my tweeters.

[AndrewT]

but you have just shorted your amp with a total of 1r0.
What voltage is left to drive your speakers?

[JPV]

Originally posted by AndrewT
but you have just shorted your amp with a total of 1r0.
What voltage is left to drive your speakers?

It is still a problem because it is not necessary a continuous short that is happening but it can be some kind of overload that will make the amplifier oscillating in a strange way.
This is inherent in a limiting technique that feeds back the control to the input.
It should be better that the circuit should create a led warning if the condition is transient and if the condition remains switch off ones and for all the amplifier.

JPV

[wg_ski]

The magnitude of the oscillation is generally a LOT less than full signal voltage. A few volts p-p won't hurt anything. If you're worried about it, hang a capacitor from base to emitter and set the time constant with the resistor that feeds it to 100 milliseconds or so. That will slow the feedback down to the point of being stable.

[AndrewT]

and that capacitor is needed to allow valid transient signals to pass even though the resistor setting for the trigger points is based on DC limitations.
I would expect that a carefully set up limiter would pass approximately twice the transient current that triggers in the DC condition.

[Hurtig]

Short circuit protection sdhould NOT be made as a current limiter... You should use a circuit that shouts down the amp when current is to high, and then wait ex 3 seconds before output is enabled again.

[stoc005]

If you change the caps to around 10uf instead of .1 uf, I see the oscillation go away and the behavior to be a bit more "civilized". The output just flat tops and won't let the current/voltage output get any higher. The amp behaves normally if the current stays below 9.5 amps per output device. They are rated at 15 amps. If it were me, I'd prefer this over the original way it worked.