Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 02:06:36 PM

Login with username, password and session length

Recent Posts

 

Speaker Attenuator Idea with Reactive Load

Started by rowdy_riemer, April 13, 2012, 05:52:53 PM

Previous topic - Next topic

rowdy_riemer

Ok, let me post a drawing (PLEASE FORGIVE MY CRAPPY DRAWING). Then, I'll explain.

rowdy_riemer

Ok, this was just a spur-of-the-moment idea, and I figured I'd run it past y'all to see what the problems with the idea were.

The idea behind this is that R1, RS1, The capacitor I forgot to label (let's call it C1), and Q1 form an active load. R2 allows us to bias Q1, but is high enough to not provide a significant shunt to the signal fed across C1. The signal applied to R1 is also fed to whatever I use to attenuate the signal. The output of that is fed to a power amplifier, perhaps a simple source or emitter follower. The power amplifier feeds the speaker, which is in series with RS2. The voltage between the speaker and RS2 is amplified by a factor that is the reciprocal of the factor by which the attenuator attenuates. The amplified voltage is then compared to the voltage between R1 and RS1. The output from the comparator is tied to the gate of Q1, so that the conductance of Q1 is increased/decreased to reduce/increase the load from R1/RS1/Q1.

The idea behind this is that the load caused by the active load mimics the speaker's load. The impedance of this load should be about the same as if you plugged the speaker directly into the amplifier.

Any ideas?

rowdy_riemer

Of course, this idea might not be so applicable to solid state power amps and might be better posted on a tube amp forum, but then again, maybe not. It used solid-state components.

Also, even if there's some reason I'm not quite on the right track here, shouldn't there be some way of providing a reactive load that mimics the reactance of the speaker via feedback?

A somewhat related idea I recently had was to use some digital means of providing a reactive load as well as vary the frequency response of a small-signal amplifier for a line out to use as an active-load speaker emulator. Perhaps an arduino board doing some frequency counting could do the job. You could then programatically change the behavior of the active load and frequency response of the line-out as you wish to mimic several different speakers.

rowdy_riemer

Damn, just saw a glaring error. From an AC perspective, Q1 is parallel to Rs1, not R1.

rowdy_riemer

A little rearranging of things fixes the problem. I'll post more after doing some simulation.

rowdy_riemer

Here's a more developed version of the idea. R7 and R8 simulate the speaker load. R4 and R5 are supposed to be dual-ganged pots. In multisim, this basically works, though there are issues to work out. The Reactive load increases and decreases as I change the R7/R8 simulated speaker load. But the current is close to double what is going through the simulated speaker load at 8 ohms and about half that at 100 ohms. Still, it works well enough for me to think this is doable. BTW, don't read too much in my source follower output stage. I'm just testing out a basic concept. I know not to run DC through a speaker. Just keeping it simple at the moment.

rowdy_riemer

Ok, some significant improvements. The inputs of the op-amp comparing the voltages across the sampling resistors now have a midpoint bias upon which the ac voltages of the sampling resistors are imposed. This prevents the dc current of the active load from messing up the comparison.


rowdy_riemer

The above changes significantly improve the performance of this idea. The current through the simulated load closely follows the current through the simulated speaker load. There is still a problem, though. If I attenuate the signal to the output amplifier, things fall apart. I'll figure it out though. :) I guess I'd better. If it doesn't attenuate the output, the whole thing is a bit pointless.

rowdy_riemer

BTW, if it looks like I should have seen many of these problems from the beginning, understand that I don't have a lot of time to spend on this. I've got WAY more physics homework than I can possibly finish, and I've got an Essay for another class due soon. I'm working on this here and there when I get a chance.

phatt

#9
Not sure I get it but the UA (Ultimate Attenuator) might give some clues as it actually uses a S State amplifier to drive the speaker. supposed to be transparent reproduction at reduced SPL
I vaguely recall Teemu mentioned how it worked some time back??
Might have been another forum?
Phil

rowdy_riemer

I'll try to find it. Did it use a reactive load that mimicked the actual speaker's reactance? I'm sure there's reamping attenuators that do, but I haven't come across a description of one yet. I should look in Teemu's book.

rowdy_riemer

Ok, I've had a little time to look at this again, and I realize I made a pretty retarded mistake with my last revision. I accidentally hooked the output of both sampling resistors to the same input on the op amp, putting them in parallel. No wonder the voltages across them matched so well.  :duh