A simple speaker protection circuit

[AndrewT]

Thanks Mh,
I didn't see the constant current through r3!

[mhennessy]

Originally posted by AndrewT
Thanks Mh,
I didn't see the constant current through r3!

No worries

BTW, you're right about the zener diodes - as shown they only work as normal diodes anyway. I'd omit them and just connect the base of T1 to the emitter of T2. No worries about the 6-7V reverse-VBE damage as the other transistor clamps this to 0.6V.

Another BTW - much as the power-saving relay scheme is a good idea, it might not function in this circuit because C2 causes T3 to turn on slowly. He could fix this by adding another transistor to make a schmitt trigger... I've used current sources with relays, but this has been more about varying supply voltages than saving power... (a good idea if using the half-wave/small C scheme you rightly suggested)

Douglas Self did an article about relay release speed, conclusions were pretty much common sense IIRC. Don't forget the delay caused by the input RC filter (I guess about 65ms currently)

Cheers,

Mark

[Russ White]

Excellent feedback Mark and Andrew!

I very much appreciate your thoughtful advice.

Here is what I have now based on your inputs.

Any further feedback is greatly appreciated.

R4 is there to adjust the sensitivity of the DC protection.


:EDIT: schematic was screwed up. Fixed.

[mhennessy]

Hi Russ,

This looks better than before, and will probably work ok. Comments?

R4 probably won't affect the sensitivity in any desireable way (because it depends on the HFE of the transistors, and will affect the turn-over frequency of the input filter). It's better to have lower values here because the input transistors will need some current - 100K is as high as I'd go.

R3 probably isn't needed.

R7 definitely not needed.

The circuit doesn't self-reset very quickly because there's nothing to discharge C2 (apart from the intrinsic diodes in the op-amp - not good!). Add an 1n4148 in parallel with R2, pointing "up"...

C3 would probably be better if it was 100uF, perhaps more. Testing required...

The circuit still suffers from having to discharge the power-on delay cap via the input transistors (and R1). This morning I quickly sketched up a circuit that uses a PNP emitter follower after the input pair to get around this problem. It also uses a current source to drive the relay. I've hacked around your latest diagram to show you the idea...

There's an optional DC-fault LED indication, and longer power-on delays are possible because you can use large values of C with no issues. Whether it's better for your application is obviously for you to decide...

Hope this helps,

Mark

[Russ White]

Wow thanks Mark, those are some good pointers.

What would you think about something like this?

[soundtech]

Not wishing to undermine all your hard work, but Velleman have a kit for this.
Part # K4700, the assembly manual with schematic are available for download on their website.
www.velleman-kit.com

Paul

[mhennessy]

Originally posted by Russ White
Wow thanks Mark, those are some good pointers.

What would you think about something like this?

Hi Russ,

That looks pretty good. IC1 will ensure the relay is snapped on quickly, allowing the power-saving scheme to work. I might have added another transistor to my last circuit to achieve that, but it's personal preference thing. I like transistors!

Now, C2/R4 can be lower/higher respectively. I chose those values in my circuit to ensure sufficient base current in the relay drive transistor. Certainly 22u and 100K would be where I'd start now...

Regarding IC1: Watch the total supply voltage (assuming it's fed from a half-wave rectified version of the main transformer). Obvious, I know...

Be sure your chosen device is happy to have its inputs close to the supply rails (I've been caught out in the past with the TL08x devices). It might be worth trying a resistor (1-10K, say) at the non-inverting input of the op-amp so that the input doesn't see such a low impedance close to the rail (some devices can be funny about this)

It might help to add a feedback resistor to IC1 to give it schmitt action. Say 100K-1M between out and +in, (note you'll also need to add the input resistor mentioned above). This will ensure the o/p changes state cleanly with no spurious transitions (perhaps caused by the unregulated supply). Any funny business happening at this time might affect the operation of the power-save relay drive circuit.

Time to start bread-boarding

Mark

[mhennessy]

Originally posted by soundtech
Not wishing to undermine all your hard work, but Velleman have a kit for this.
Part # K4700, the assembly manual with schematic are available for download on their website.
www.velleman-kit.com

Paul

Yes, I know this kit well. While this might solve the problem, no-one here would learn anything from just buying a kit.

Also, it's somewhat complicated, don't you think? Just look at all those electrolytic capacitors waiting to dry out!

I don't know if Russ is planning to add this to his product range, but it would certainly be a compact and economical alternative to the Vellemen offering

Mark

[Russ White]

Originally posted by soundtech
Not wishing to undermine all your hard work, but Velleman have a kit for this.
Part # K4700, the assembly manual with schematic are available for download on their website.
www.velleman-kit.com

Paul

This is not about "obtaining" a solution, but rather designing one. There are more than a couple designes I could directly copy, but I wouldn't learn anything that way. This is the joy for me, the journey, not just the destination.

t has been very fun so far.

[Russ White]

Originally posted by mhennessy


Hi Russ,


Now, C2/R4 can be lower/higher respectively. I chose those values in my circuit to ensure sufficient base current in the relay drive transistor. Certainly 22u and 100K would be where I'd start now...

Regarding IC1: Watch the total supply voltage (assuming it's fed from a half-wave rectified version of the main transformer). Obvious, I know...

Be sure your chosen device is happy to have its inputs close to the supply rails (I've been caught out in the past with the TL08x devices). It might be worth trying a resistor (1-10K, say) at the non-inverting input of the op-amp so that the input doesn't see such a low impedance close to the rail (some devices can be funny about this)

Time to start bread-boarding

Mark

Hi Mark,

Once again very good input! Thanks!

I was thinking I might add an additional zener on the non-inverting opamp input to limit the voltage there to below the rail voltage, maybe something like a 18V zener.

I will add the resistors as you say, I was actually thinking the same thing, just had not fleshed it out that far yet.

I will probably use something like a LF411 for the circuit (since I have a bunch of them and they are relatively cheap).

Cheers!
Russ