Output attenuator getting too hot

[Laker1980]

Hi. I'm new to electronics - I thought I'd have a go at a simple attenuator build to get me started. I followed a build from guitar magazine - a 100w 8ohm Lpad with and input and output in a box.
I plugged the attenuator in and it does work, however it started to smoke. I've read similar posts online and I Know that attenuators do get hot but this isn't what I was expecting. My amp is a Laney VC50 - so a 50 watt output. The Lpad is rated to 100 watts so I'm a bit confused... I've read a few possible solutions, one being to add a few 100watt  8ohm resistors before the Lpad to reduce the load on it.
If I added 3 100watt resistors before the Lpad (as suggested), how will the resistance be affected? Does the resistance need to be 8ohms to coincide with the speaker? Or will adding 3 8ohm resistors affect the current flow?
Any help would be greatly appreciated. Ta

[Rob Strand]

I followed a build from guitar magazine - a 100w 8ohm Lpad with and input and output in a box.
I plugged the attenuator in and it does work, however it started to smoke. I've read similar posts online and I Know that attenuators do get hot but this isn't what I was expecting.

Unfortunately the ratings of those things aren't as clear as they should be.  Typically the 100W rating applies to the system power when the LPAD is used on a tweeter and fed from a high-pass crossover.

You can see from this graph the power a tweeter sees with a crossover is a lot less than the system power,
https://sound-au.com/tweeters.htm

It makes sense since the tweeter is only fed with a small part of the spectrum

If I added 3 100watt resistors before the Lpad (as suggested), how will the resistance be affected? Does the resistance need to be 8ohms to coincide with the speaker?

Adding series resistors increase the impedance seen by the amplifier.   The resistance seen by the amp is the sum of all series resistances and the speaker impedance.   It also changes the damping of the speaker.

So one way of correcting the impedance seen by the amp is to use a two resistor divider, like this,
http://www.sengpielaudio.com/calculator-Lpad.htm

(There's a few forms of this idea.  For example another type would move R2 to the amp terminals. The calculator link only applies to the case shown on the page.  You need another calculator when R2 is moved.)

EDIT: I should say the form with R2 at the amp terminals is probably a little better for a tube amp.  You can do better with complicated dividers using inductors and capacitors, like the Marshall products.

[Elijah-Baley]

Hi. Can you show us the schematic you followed?
I'm interesting in this kind project, though I don't have a tube amp.
I read somewhere that with a poor and simple project of this there's something awfully wrong with the impedance of the speaker that could even damage the amp. Someone measured 1,6 Ohm or so. And I guess amp must never go under 4 Ohm.

[Rob Strand]

I'm interesting in this kind project, though I don't have a tube amp.
I read somewhere that with a poor and simple project of this there's something awfully wrong with the impedance of the speaker that could even damage the amp. Someone measured 1,6 Ohm or so.

You can get OK results with simple attenuators.  There's no reason for the impedance to go very low value, that would be a specific design issue not an inherent problem with resistive attenuators.

[Elijah-Baley]

Indeed, it could be a problem of the design. For this reason I ask for the schematic.
I can't find the page where some guys built an attenuator until someone found out this weird issue, that could cause problems with the impedance.

[...] a 100w 8ohm Lpad with and input and output in a box. [...]

Reading this it seems a really really simplified version, in the web fortunately there are some a bit complex designs of this kind of attenuator. Some schematics includes caps, but I guess caps need for change frequencies response. Some schematic includes a peculiar high power resistor or whatever.

Anyway, I'm can't say why the build of Laker1980 doesn't work properly, but could be something unconsiderated, perhaps?

[Ripthorn]

What kind of resistor did you use? Did you attach the heatsink to something? Smoke isn't good, but you have to make sure that you are getting rid of the excess heat appropriately as well.

[Rob Strand]

Anyway, I'm can't say why the build of Laker1980 doesn't work properly, but could be something unconsiderated, perhaps?

The whole problem stated because the Lpad is called 100W but it probably isn't.

If you drive enough power into any resistors it will eventually get very hot.  I've got some very cooked looking 5W resistors I used for load testing.    A 5W resistor dissipating 5W gets upto 200C on the surface so going over  5W is getting pretty darn hot.

[Rob Strand]

Reading this it seems a really really simplified version, in the web fortunately there are some a bit complex designs of this kind of attenuator. Some schematics includes caps, but I guess caps need for change frequencies response. Some schematic includes a peculiar high power resistor or whatever.

There's a whole lot of issues to consider when designing an attenuator for a guitar amp.  Even if you match-up all the impedances it might not match-up in tone.  It would take a lot of writing to explain all the details.

The complex circuits with inductors, caps and resistors, even transformers, try to match-up everything but the penalty is a very large costly circuit.

On the other side of the coin some circuits try to reduce the number parts and simplify switching.  This is a minimalistic "what you can get away with" solution.

It is possible to do an OK job using only resistors.