Ideas on overload shutdown

[DIYdo]

Can anyone direct me to any known circuits that will shut an amp down in an overload condition. I would rather this method than VI limiting.

I have tried some ideas of my own using op-amps. Can't get the things to do what I want though.

Any ideas will be greatly appreciated.

[AndrewT]

Hi,
1. Temperature switch on the heatsink.
2. Thyristor detecting excess current in the output stage or supply rail.
3. Thyristor detecting high frequency in the VAS stage.

Temperature switch could be either reset on cooling or latching.

[ilimzn]

2. Thyristor detecting excess current in the output stage or supply rail.
3. Thyristor detecting high frequency in the VAS stage.

How would that work, a thyristor being a latching element? Crowbar (blows fuse)?

[DIYdo]

I was thinking more along the lines of putting all output power through a low value resistor, then by monitering the voltage across it operating a relay to remove operational power.

I think a thyristor latching design would be more appropriate to shunt off power in case of a blown transistor. Still all ideas are worth looking into.

[AndrewT]

Hi,
your almost there.
Thyristor monitors the voltage across the resistor and latches.
The latch could be to a mains relay or CCS feeding an LTP or crowbar on PSU or ....... just keep adding to the list.

Look up the Cyrus mk1 amplifier. It has a thyristor across the output device Re and latches until the signal falls to zero current then opens and waits for the next overload (next cycle). The latch (from memory) pulls the LTP or VAS to zero output signal.

My Crimson latches on detecting high frequency on the VAS and pulls both the LTP and VAS to zero output signal. It needs mains shut off to unlatch. Only ever happened once when testing with a 50kHz test signal.

[DIYdo]

I shall look into that. Many thanks.

[forr]

I immediatly thought to Crimson. I may have the schematics somwhere if you are interested. I establish it by reverse engineering. Very simple. The funny story is that I bought the Crimson boards to know how the latch was done. It was in 1981 and the amp is still living.

~~~~~ Forr

§§§

[AndrewT]

Hi Forr,
and it's still fitted to the new mkVIIIs.

Did I get it right? detects hi F and latches the LTP down to zero output?

[forr]

HI AndrewT,

I'll try to scan the Crimson schematics tomorow.

When an overlaod occurs, the Constant Current Sources of both the Long tail Pair and the Voltage Amplifying Stage are inhibited.

The LTP has a standard two bipolar NPN transistors CCS.
The collector voltage of the sensing transistor is the voltage reference for the CCS of both the LTP and the VAS.
At overlaod, a latching circuit sends a few mA in the sensed resistor, the sensing transistor saturates, its collector very low voltage turns the current off of both the LTP and the VAS.

How an overload is detected ? The PNP Common Emitter of the VAS is buffered by an NPN Emitter Follower having a 100µ capacitor and a 1 Kohm resistor across its base and its emitter.
When an overload occurs, the protection transistors of the output stage shortens this emitter (through a 2.2 kOhm) to the amp output.
So the current of the Emitter Follower increases : in its collector, it triggers a two transistor circuit working like a thyristor whose current output increases the voltage across the sensed resistor above.

There is a little delay to avoid the circuit to latch on short output overcurrents. Instead of switching off/on the power supply, a push button can be installed to unlatch the thyristor and restart the whole amp.

~~~~~~~ Forr

§§§

[davidsrsb]

Seconded here that the original Crimson circuit works very well, its never false triggered and has survived some abuse over he years.