Cooked the 4 ohm Resistor

[guitarkeys.com]

Ok. This is out of an SLA that has been run HARD the last two years. I mean without regard for safely max volume hard. The wire finally got so hot it came out of the end which alerted me to the issue. It's a rental.

I believe it was a 10W resistor so I'm thinking of going with a 20W - No other components show signs of excessive heat. But I did see some aluminum clad 50 Watt resistors that would certainly mean I wouldn't have to mess with them again. They are about 10 bucks each at mcm, but if they will work for the piezos they might be worth it in this case.

Any advise is appreciated. Well... almost any advise.

[Bill Fitzmaurice]

Any advise is appreciated. Well... almost any advise.

That's a sign of abuse. I realize that part and parcel of a rental agreement is that the renter must flunk an IQ test, but still...

[escapemcp]

If a 10W resistor has lasted this long, then a 20W one will be plenty, surely??

[LelandCrooks]

I provide 25w ceramics on the 4 ohm side. Just for precaution. I had one do that years ago, a piezo array was grounding to the frame of the driver on a old Dr250 that I'd reworked to hold the then new melded array. Look for a short in the cabinet.

[guitarkeys.com]

Thanks Leland. I double check that just to be sure but I think the only short is between the renters shirt collar and hat.

He's a kucklehead and I can't fix that, but he pays and has no issue with covering any repair costs (parts and labor).

[byacey]

Don't forget that the power rating on the resistor is derated when it isn't allowed free air movement for convection cooling. Burying it within insulation batting or under foam can overheat the resistor or even potentially cause a fire. The one pictured appears to have been pretty warm.

I would agree with Bill that this speaker was working way harder than it ever should have been.
Although good design practice dictates not working past 50 to 75% of the power rating of the resistor, a 30 ohm 10W resistor, with proper convection cooling should handle 17 volts across it all day long.
This kind of voltage across a piezo array is stinkin' loud.

Moving up to a 20 W resistor will only increase the maximum voltage to around 24v, vs 17v.

I would inspect the piezos, If this resistor is burnt, it's likely the drivers are damaged too.

[Blueskink]

I know that zener diodes can't protect speakers from voltage spikes but can they help with this kind customer by limiting what voltage can get through? I only have very limited knowledge of electronics.

[Bill Fitzmaurice]

I know that zener diodes can't protect speakers from voltage spikes but can they help with this kind customer by limiting what voltage can get through?

No, because they can't take the power. If it were that simple every speaker would use them.

[byacey]

Zeners are also a very non-linear device, their clamping action would create highly objective distortion.

The old school, lowly light bulb is still one of the best protective limiters for this purpose. At lower power levels the filament is a very low resistance, virtually transparent as far as audio quality. As voltage increases across the bulb the filament resistance goes up and limits the current to the speaker driver. If someone tries to be really stupid, the bulb simply burns out and results in no audio.

I can't comment on the Dfend product; I haven't seen or heard one yet.

[Bruce Weldy]

The old school, lowly light bulb is still one of the best protective limiters for this purpose.

I was running sound last year on some JBL JRX boxes.....sounded like ass. Once it got dark I saw flashes coming out the ports.....oh crap, it's on fire!

Ran up there and looked in - it was the bulb on the crossover. I've mostly seen those in a lot of lower end stuff, so I guess it was right where it belonged in the JRX boxes....real pieces of crap.

But, they do work. Opened up a monitor that was supposedly dead....made a trip to the auto parts store for a dome light bulb - good as new.

[Bruce Weldy]

The old school, lowly light bulb is still one of the best protective limiters for this purpose.

I was running sound last year on some JBL JRX boxes.....sounded like ass. Once it got dark I saw flashes coming out the ports.....oh crap, it's on fire!

Ran up there and looked in - it was the bulb on the crossover. I've mostly seen those in a lot of lower end stuff, so I guess it was right where it belonged in the JRX boxes....real pieces of crap.

But, the bulbs do work. Opened up a monitor that was supposedly dead....made a trip to the auto parts store for a dome light bulb - good as new.

[jimbo7]

Got a metal jack dish? Mount the resistor to that for heat dispersion.

[Bill Fitzmaurice]

The old school, lowly light bulb is still one of the best protective limiters for this purpose. At lower power levels the filament is a very low resistance, virtually transparent as far as audio quality. As voltage increases across the bulb the filament resistance goes up and limits the current to the speaker driver.

It won't work with piezos. When the filament heats its resistance increases, so more voltage is dropped across the lamp, less across the driver, if it's a driver that presents a resistive load. Piezos present a capacitive load.

[byacey]

I think we've had a similar discussion before regarding this.

The 30 ohm resistor in parallel with the piezos is used to simulate a constant resistive load for the crossover network, otherwise the crossover point would change wildly all over the map with frequency, due to the natural inverse impedance curve that piezos exhibit.

A light bulb inserted in series ahead of the crossover network will form a resistive voltage divider with the 30 ohm resistor. If the bulb drops more voltage across the filament, less voltage will be dropped across the 30 ohm resistor, and the piezos will receive less voltage resulting in lower audio output.

The capacitive reactance of the piezos can essentially be ignored because it's swamped by the relatively low value of the 30 ohm resistor.

[Bill Fitzmaurice]

I think we've had a similar discussion before regarding this.

My conclusions are based on having actually tried it. It doesn't work, the load impedance being too high compared to the bulb resistance.