Attenuator question

[shreditup]

Hey,
So as an experiment i built an attenuator using this schematic (http://freestompboxes.org/members/so...erMiniMASS.gif).
To simplify it, i only used the 50ohm rheostat and the 16ohm speaker(without line out,impedance switches and treble caps). As for the speaker, i bought a 200-5000hz random speaker, and tried to strip it down. I figured i just take off the cone, spider and dust cap (http://cdn.head-fi.org/b/b8/b88b5caa_speakerparts.jpeg), leaving plates, magnets, pole and the coil, hoping it won't produce sound. My lack of theory showed, and even in this configuration, it produced sound.
The thing is, the attenuated amp sounded amazing and very natural, but i then tested the limits, and the coil burned down. So I'm going to build another one, and have 2 questions:

1) What should i take off the speaker so that it won't make any sound, but still have the characteristics of why it's there in the first place? Looking at what they use it seems like a there's a coil, magnet and spider. Or, what they call it, voice coil motor assembly. I've googled it, but didn't seem to find any answers.

2) Does it matter what kind of a speaker i use here? I don't mean the wattage or impedance, I mean the frequency response.
In other words, say we have 2 speakers of 16ohms, but with different frequency response. If we run the first at 50w, then add the second in series before the first one and pump the power to 100w(so that both will get 50w), will the first speaker sound change?

[Tom Phillips]

The spider holds the voice coil centered in the magnetic gap. Without the spider the voice coil will be bouncing around, rubbing and I'd expect it to fail after a short time. It may even jump right out of the gap. All this is going to make noise of course right up until it smokes.

The purpose of using actual speaker components for an attenuation load is to more closely simulate a real working speaker load. That means that the spider needs to be left in place and the voice coil needs to be allowed to move. You would need to glue the voice coil firmly in place so that it cannot move at all if you don't want it to make any noise. This will defeat the purpose of using real speaker parts. If you want the attenuator to be totally silent then you might as well just build a reactive dummy load using passive components (resistors, inductors & capacitors). I think that will result in a load just a good, or better, than a stripped down speaker.

Cheers,
Tom

[shreditup]

What i don't understand then, is given that the MASS doesn't produce any sounds(I'm not 100% sure, it's what people on forums say), I suppose the coil isn't moving? Is it just a fixed value inductor then? If so, why not use a an inductor of that value that will no doubt be cheaper?

Still, does it matter what characteristics(frequency response) of the speaker i'm going to cut are?

[shreditup]

Ok, i've googled some more, and the motor in the MASS indeed does produce sound, especially when driven hard. So i'll just cut the cone and dust cap off of some high-power speaker, and i'm good.

The only thing left is: Does it matter what speaker i'll use for this? Will it affect the sound on my actual speaker, if the motor inside the attenuator is from a bass driver(20-2000hz) or if it's from a tweeter or whatever?

[Tom Phillips]

I would think that the type of speaker you use would not inherently make a lot of difference. However, you do need something that can dissipate all the power from your amp and it ideally should move enough to generate the back EMF similar to a real speaker. Therefore, a base speaker or a standard musical instrument speaker should be be used.

[The Dude]

Why not just use something like THIS (or combinations depending on your power rating and desired impedance)?

[Joe L]

Check out this thread on my research on the same subject. I have received the speaker and when I have time, I am going to post my results. I don't plan to modify the speaker. My first tests will involve remote location, and internal 4X12" location of the speaker as a main load of the amplifier's output.

[bob p]

> If you want the attenuator to be totally silent then you might as well just build a reactive dummy load using passive components ...

Tom, are you referring to something like the TW Airbrake, or do you know of a better model with passive components?

[Tom Phillips]

> If you want the attenuator to be totally silent then you might as well just build a reactive dummy load using passive components ...

Tom, are you referring to something like the TW Airbrake, or do you know of a better model with passive components?

I was just commenting about the general technical approach. Didn't have any specific commercial model in mind. My feeling is that by the time a speaker is cut down to just the "motor" or mounted without a cabinet load it is going to interact differently with the amp than a normal musical instrument amp speaker setup. Maybe a reactive load built with Ls Cs & Rs would produce a sound that is fine. There have been discussions covering this and lots of opinions. Some people even like the purely resistive load. All the approaches will work so the one the user likes best is the one that is best for them.
Tom

[bob p]

Thanks for your reply, Tom. I guess I wasn't clear, when I said "model" I wasn't thinking about any specific commercial model -- I was thinking of an LCR equivalent model of a reactive load circuit that closely approximates a speaker. So instead of asking for your recommendation of a specific commercial product to buy, I was asking about an equivalent circuit model that's designed to emulate a generic speaker. I was thinking that that would make an interesting starting point for tinkering.

FWIW I have tried building "ladder" type attenuators (the kind that you'd find in a radio engineering handbook) that preserve a constant Z load as the level of attenuation is varied. These are lots more complicated than the typical TW Airbrake circuit. The ladder-type attenuators probably provide the best overall performance of the purely resistive variety of non-reactive loads, with the trade-off being complexity, parts count, and the price to build them. Although the overall performance of the ladder type attenuator is about as good as it gets for a purely resistive load, I'm thinking that it would be interesting to model a reactive load type of circuit, without actually using any speaker components. I was hoping that someone might have a good theoretical model for something along those lines.

thanks.

[Tom Phillips]

Hi Bob,
I too would be looking back through my records and searching out available models so I don't have any specific circuit model in mind. I've thought about this for a long time but I have never gotten around to building a reactive speaker load circuit. I remember starting with the information in Ch 20, Loudspeakers, of the Radiotron Designer's Handbook. They show several equivalent speaker load circuits. However, they qualify the design by saying "at low frequencies" which, based on accompanying figures I take to mean <500 C/S. There are, of course, other circuit models posted on the net and in musical instrument amp books and discussions.
Tom

[shreditup]

Sorry for not posting for long, i was playing around and testing this stuff.
So in the end i did this and put it in place of the voice coil motor assembly in the weber schematic, and put a 4,7uf cap parallel to it.

The results are very good. I am not sure if it preserves tone as some expensive attenuators do, but even at extreme attenuation(which is how i use it) the tone is outstanding and i have not unplugged it since.
Next i would like to experiment with some reamping stuff. Anyone have any good schematics for that?

[Tom Phillips]

...The results are very good. I am not sure if it preserves tone as some expensive attenuators do, but even at extreme attenuation(which is how i use it) the tone is outstanding and i have not unplugged it since...

If it sounds good to you then you achieved your goal and that's great. Thanks for reporting the results.