Failure with easy spk protection circuit

[bigpanda]

Hi guys,

I tried to build the speaker protection unit at

http://www.astro.uu.se/~marcus/private/m250.html

It's easy and simple, but I just can't get it work. I didn't lay a pcb for this. After I wire all the components together, I apply a 12V DC to it and all I can see is the LED on my relay turns on and heard that the contacts close up without any delay at all. I tried to apply a DC of 5V to the speaker input pts but nothing appened. It's a shame to screw up such a simple circuit, but can someone please kindly shed some light on this? All compontents are new and checked before soldered together and wiring had been checked and rechecked. I thought I might had a 'no contact' if I made a mistake but I got an 'always contact' instead. I had a 200W amplifier almost finish and I need this one to protect my speakers for testing (although they are not expensive at all). can someone please kindly give some advice?

Tks

[Netlist]

What happens when you apply a negative voltage? Say -1 to -5V

/Hugo

[ekaerin]

Hi there,

I'm sitting at the comp. drink'n beer. I'm on my way out, but:

333uF and 22k, did you check at what time you reach 0.6V at the
relay trans. base.

Did you ground your 5V with the 12V.

____Then : Check if all transistors are OK. By theory the guy should work.

/ Mattias

[janneman]

The circuit of the two transistors on the top that receive the speaker signals is ill defined as to gain etc. It is possible that it conducts without speaker input signal. Also, the switching point of the relay transistor is ill defined, and depends a lot on the particular relay and transistor gain. A schmitt-trigger-type circuit is normally preferred for a fast pull-in of the relay to avoid arcing etc. See D Self's articles.

So, when it works, the delay will vary between different transistors and relays. Probably no problem for you for a one-off implementation, you can adjust it to taste.

Remove the wire from the base of the relay switching transistor to the above two transistors, and try again. You have a delay now?
If you try repeatedly, make sure you discharge the delay cap between tries.

Jan Didden

[bigpanda]

Tks guys,

janneman :
Your are trying to seperate the two f'n circuits, right? That's exactly what I have done already. I don't get any delay either. Maybe I have to pulled everything out and check the components individually again? I hate doing this because I did not use a specified pcb.
Do I adjust the 1K to vary the delay (if there is any at all) What you are implying is that even if I had this working and somehow if I change the relay in the furture, I might had to tune this again?

ekaerin :
If I apply a 5V DC to the amp. pts., I would have to apply 5V DC with reference to the ground, aka the +ve of my 5V dry cell go to amp. pts nad -ve go to the ground. Is that what it is suppose to be?

[hughmon]

The turn on time constant's so short that you may just not be noticing it with a transistor with good beta (assuming the circuits built right with good parts). Try upping the value of the delay cap to 3,300uF and see if the delay's noticeable. If you do that, put some diode like a 1N4004 around the 47K (cathode to +12V) for a quicker discharge of the delay cap at turn off. As for no response to a 5V offset in the DC detectors, it could be a construction issue, or it could be insufficient beta in the detecting transistors. Not to be knocking anything, but this is a relatively poorly designed circuit, and I personally would look elsewhere for something better.

Nitpicking 101: Not that this is something that's ever likely to happen, but with the output of the two amps summed in the offset detector, if one amp should fail positive while the other fails negative (vaguely possible in a bridge configuration) the two opposite volatges will cancel resulting in no shutdown signal from the offset detector. It's better, especially if bridging, to use a separate detector for each channel.

[janneman]

Checking each component isn't that usefull in these cases. Better to build it up step by step, first the relay switching transistor with the RC delay network. Put a DC meter on the C and see what happens when you switch on (use a DC meter with high input impedance min 100k). You've got to get this working first. Double check all your polarities and values and pin-outs. Changing an e for a c is easily done and can keep you busy for days. Ask a friend to check the wiring etc.

Jan Didden

Edit: A rough calculation tells me the voltage across the C rises by almost 1V per second. Depending on the transistor gain, that means the relay pulls in anywhere between a half and 3/4 of a sec. There is a delay, but you miss it probably because it is quite short. Change the 47 k series R to 200 k and double the cap. If you can, lower the 12V to 5V. You should get a clear delay of up to 10 seconds.

[bigpanda]

Ok, I got the delay part working at last.

After I tried with different values of cap. and res., I got a 3-4sec delay with 330u and 91k (actually there are some other combinations but I stick with the original 330u and use this combination).

I connect the DC detection part, the circuit doesn't work again but this time there is 'no connection'. What am I going to do now?

[hughmon]

Assuming no wiring mistakes, one or both of the offset detecting transistors could be either leaking, or coming on due to sporadic noise. Try connecting something like a 270K resistor between the base and emitter on each of the two transistors and see if that helps. A .01uF cap paralleling the 270K resistors may also help with noise.

[Gofer]

It is easy to screw up simple circuits, I now everything about it.
Therefore I might have some usefull (hopefully) comments.
I have noticed that the dc detection part is copied from the elektor magazine long time ago.
Please check the polarity of the diodes and the orientation of the transistors. The base is in the middel.
Make one of the dc detection transistors to conduct by placing a 1.5 V battery (or whatever) with a resistor of 1K in series among the base and emittor. Does the contactor open?
Also a mistake I would make is to wire the contactor relay the wrong way. Make shure that when there is no power to it the speakers are not connected to the amplifier.

I hope this helps and if not I will make one myself and let you know. (I am going to build latmos beasts so I need it too).

By the way does anybody has a copy of the dc protection circuit of the good old 1982 Crescendo?

Good luck, Govert-Jan