Series current limiter

Hi,

I am often repairing solid-state audio amplifiers, and would like to build some kind of current limiting circuit.
Ideal would be some "black box" similar to limiting current with a light bulb, because it would be great if it didn't need a ground connection.
Is that possible?

If not, could someone help me design a current limiter, which won't be too sensitive to input voltage variation?

Another thing that concerns me is, that if I use some current limiting devices on both power rails, is it possible that the two limiters would fight against each other or oscillate?

Many thanks

Unless you have two power supplies, you just need a single current limiter. You can put a 1 ohm resistor or less and a transistor in series with the whole circuit, monitor the voltage across the resistor (which equals I*R so you monitor the current through it because R is fixed) and use a comparator to sense when your current reaches the limit. When the comparator will sense the current limit, it will activate an op amp which will decrease the voltage across the base-emitter junction of the transistor (the transistor has to be fully turned on if the current is below the limit) and thus increase its resistance and keep the current fixed at the desired value.

I need two of them, because there can be faults like transistor's case shroted to ground, which affects only one of the rails.

I am aware of the op amp current limiter, but I was hoping that there exists some device circuit that doesn't need external power supply at all (which I know is highly unlikely), or something that could operate from up to +/-100V rails, the maximal current can be cca 1-2A for short time (10-15 sec).
100V is quite impractical to regulate, but it can be done. The problem is that I wish to make some universal kind of limiter, so the voltage can range from say +/-25V to +/-100V.

Any suggestions?

One handy item for decreasing the voltage on amplifier supply rails is the variac. Lower AC in means lower DC out.

beenthere said:

One handy item for decreasing the voltage on amplifier supply rails is the variac. Lower AC in means lower DC out.

Click to expand...

You mean I should attach servo and operate it based on secondary current?

Seriously, I don't want to lower the rails because that would change the quiescent current, and also could create DC shift of output, which really won't help too much with finding possible faults.
I would like to have a simple once-for-all solution, in the way that I connect it to he main rails of the amp, and the amp either triggers one protection, both, or stays with some reasonable current.

My post #17 in this thread would be a good start.

http://forum.allaboutcircuits.com/showthread.php?t=13594

studiot said:

My post #17 in this thread would be a good start.

http://forum.allaboutcircuits.com/showthread.php?t=13594

Click to expand...

Do you know where I can find that book in electronic form? That would be really great to read whole.

Its too old, but see your PM.

This document: http://www.national.com/appinfo/power/files/f4.pdf has a section (beginning on p12) outlining current limiters.

Well, I suppose you could use a TL783 high voltage linear regulator with a resistor, similar to how you'd use an LM317 in a lower voltage situation. However, the dropout voltage is pretty large at higher current levels (10v to 18v depending upon temp and current).

But, if you're trying to allow even a moderate current flow with a high voltage differential across it, you'll have a bit of excitement with the noise and smoke generated by it's sudden demise. 250mA with a 100v differential is 25W. Even a hefty copper heatsink with a fan wouldn't likely be enough.

thingmaker3 said:

This document: http://www.national.com/appinfo/power/files/f4.pdf has a section (beginning on p12) outlining current limiters.

Click to expand...

Thanks for that, I managed to make a working limiting circuit, at least in simulator.

And one more question, does anybody know what is a typical inductance of wirewound ceramci 5W resistor, in my case 0.1ohm? Thanks

Ohmite makes non-inductive wirewound resistors. Digikey and Mouser are two distributors that carry Ohmite products.

Thanks, but I don´t need non-inductive resistor, I would just like to know what is the typical inductance.

It will vary widely between manufacturers, resistance, and wattage rating.
Different gauge wire, different diameters of winding, different spacing of the coils, and number of turns will all make a big difference in the inductance.

Wheeler's Formula will get you within about 3%, if you know all of the above parameters.

Ok, that sounds reasonable, I will have to try it with my meter, but I suppose it is too small to measure.

You also forgot the permeability of the surrounding material

kubeek said:

You also forgot the permeability of the surrounding material

Click to expand...

I would think that sand/ceramic would have a pretty low permeability

Heres a simple solution, however, it is a little old school. The attached circuit will limit at a value of .7 volts across either emitter resistor. Select transistor with appropriate Vce max and power dissapation for your application.

Raychem poly fuses might also solve your problem