Amp attenuation scheme discussion

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
yldouright: do an A/B test of a system as you suggest compared to a direct connection of the speaker to the amp. If you can not hear the difference, well than...., if you can find a speaker that "does not like" the damping, let us know!
This thread is going in circles and I am getting dizzy! E
 
mickeymoose
I guess moose like to chase their tails around. The only thing circular in what I have stated is the shape of the pot wiper. Seriously, I've read a tighter speaker should be okay with no damping ratio. I've even heard some people say they prefer it because it sounds less compressed, is this incorrect?

Michael Bean
I am investigating the pros and cons of attenuating an amp at its outputs. It seemed to be a way to avoid amplifying noise. I know it has been unpopular and I would like to know if there are new methods or technologies that can make it a practical design choice.

janneman
When I opened this thread, I invited the possibility of any type of attenuator but since the discussion was veered to the common pot, let's examine this first. Doesn't the assertion that current is flowing through the pot to ground assume that the pot is connected as a shunt? Isn't there a way to tie in the pot without it wasting the output to ground?
 
This is where the noise of the pot would be of the smallest relative value to the signal, right? What are the issues that make this practice undesirable?
and
When I opened this thread, I invited the possibility of any type of attenuator but since the discussion was veered to the common pot,
I'm certainly willing to disengage my contribution then.
 
OK, putting aside all of the arguements against your idea for being impractical, you cannot improve signal to noise ratio at the back end of the signal chain. As was pointed out in an earlier post, decent quality modern amplifiers don't normally add significant noise, and attenuating the signal at the output won't do anything to reduce noise that has already mixed into the signal, it just lowers signal level and noise together. If you really want to lower noise level (or improve S/N ratio) in you system, you must do it at the front of the signal chain. I highly recommend you take another look at the link that Jan Didden provided in post #15 about Gain Structure, lots of good info there. Frankly, what you want to do, and how you're trying to do it just won't work.

Mike
 
sofaspud, mickymoose, Michael Bean
This thread was opened to explore amp attenuation. Theoretically, it is better to NOT amplify noise so I started out looking at the rump end of the amp first. You have chimed in your negative comments with only sideswipe justification of your positions. The issue of power dissipation in the attenuator was left unresolved as was the issue of loudspeaker damping and output impedence. If a person cannot accept a blanket dismissal of an idea without logical reasons to do so, does that make him stubborn? It may turn out to be truly impractical to go this route but that does that mean we shouldn't look more closely at it?
 
AX tech editor
Joined 2002
Paid Member
[snip] Frankly, what you want to do, and how you're trying to do it just won't work.

Mike

Ohh, it'll work, no reason it shouldn't.
Lets get practical: an 8 ohms speaker connected to the wiper of a pot which is driven from the output of the power amp.
Normally you would chose the pot resistance about 1/10 of the load so that the load doesn't influence the pot curve. So, we need a 0.8 ohms pot. Hmm. That IS impractical...
OK lets fudge the rule of thumb, lets take a pot only 1/2 the load, so we get a 2 ohms pot.

Right. Getting there. Now, we would want say 50W in the speaker with the pot turned all the way up. The amp sees, at max level, a 2 ohms pot parallel to an 8 ohms speaker, which is 1.6 ohms total. For the 50W in the 8 ohms speaker you need 20V RMS, which means there will 10A RMS through the pot and 2.5A through the speaker. The pot will dissipate 200W, and 50W will go to the speaker.

Now turn the pot way down; nothing goes to the speaker but you have still 10A through the pot so the pot still dissipates 200W. At intermediate settings, the pot dissipation will vary but will always be quite high.
But nothing stops you from doing it.
Just get a 2 ohms, 200W pot, mount it such that it doesn't cause a fire hazard, get an amp that can drive 20VRMS into 1.6 ohms continuously, and you'r done. Piece of cake :D

jan didden
 
Apparently, what we have here is a failure to communicate.

First, you said it yourself,"Theoretically, it is better to NOT amplify noise", but it isn't merely theory, it's a fact. What you're not seeing is that once it's come through the "rump end" of the amp, the noise HAS been amplified and you cannot remove it with an attenuator. Think of it this way, suppose you are in a noisy environment trying to talk on the phone. If the person on the other end can't understand what you're saying, you have two options, either talk louder (improve S/N ratio) or go to a quieter environment (noise reduction), the person on the other end, the "rump end", cannot do anything to improve the situation by turning the volume down on thier end, they can only adjust the level of the unintelligible garble as a whole. Did you revisit the link that was provided about Gain Structure?

Second, we have not chimed in our negative comments with only sideswipe justification of our positions. I provided you with diagrams and simpified calculations to show you how attenuators work, did you read it and do the math? Again, if you put power through a resistor (attenuator) it turns that power into heat, and with more power you'll get more heat, and it will affect loudspeaker damping. It's physics, not a blanket dismissal of an idea without logical reasons.

Third, it has been looked at more closely already and it turns out to be truly impractical to go this route. I recommend you go back and re-read what has been already posted here and study the info more carefully, you might gain some insite as to why what trying to tell you is true.

Fourth, nobody here is trying to beat you up, we're trying to help you understand, but until you're ready to let go of your misguided preconceptions, this will go nowhere. It seems the old saying that "you can lead a horse to water, but you can't make him drink" applies here.


Mike
 
Jan, I think he's trying to achieve noise reduction with an output attenuator, so in that respect what he wants to do won't work. I also tried to explain to him that while you can use an attenuator that way, it is terribly impracticle, just because you can doesn't mean you should, and that's why we use preamps with volume controls.

Mike
 
Michael Bean
If you had read what I had stated about the power dissipated being a function of the current passing through it you would have noted that your math would have been invalidated if there were a way to avoid losing the power to ground. You did not directly address this possibility and I daresay that you sidestepped it with the math you provided. Lesser minds will typically fall back to elementary explanations when their mind is forced to bend into an unfamiliar contortion but let's stop sniping and asking each other to reread things we have already stated we read and continue this exploratory discussion. If the attenuator is adding noise of 30uV which is amplified 20X, then the whole system minus the attenuator at the input must be comparatively less noisy at the output because of the much larger S/N ratio, no? Okay, let's recount the positions; the most annoying problem of attenuating the output signal is the power dissipation according to both you and janneman. You both reject any possibility of the attenuator being a garden variety pot and both agree that this type of attenuator at the output cannot be used without burning power into ground. Have I understood both of your positions clearly? Now, does this same problem exist for all other types of attenuation?
 
Last edited:
OK, i'm confused. I thought you were talking about attenuating the signal between the amplifier output and the speaker, but in your example where you say "the attenuator is adding noise of 30uV which is amplified 20X" I guess you're talking about an attenuator placed before the amplifier? I need clarification on this point.

Mike
 
sofaspud
I hope I'm not giving the impression that I'm averse to learning. I just don't want to learn something wrong or incompletely. I assure that I've no intention of ignoring anything useful. I hope others reading this agree.

Michael Bean
Don't be confused. I was showing you why an attenuator after amplification might have a sonic benefit by comparison.

janneman
What prevents us from having an attenuating pot in series without a ground at the output?
 
OK, so the attenuator is after the amplifier. So please explain to me how "If the attenuator is adding noise of 30uV which is amplified 20X" is true if it's after the amplifier, it would have to be before the amplifier for that to happen. Your statements are contradictory.
Also, a potentiometer that is not connected to ground as a voltage divider (diagram 3 in post #11) is called a rheostat or variable resistor (refer to diagram 2), and will drop votage depending on how much current it's conducting. As was already pointed out, adding resistance to the circut will alter the damping factor, and will change with different settings. Power will be dissapated as heat in any case.
I mean no disrespect, but it seems you don't understand Ohms Law. Google it and study, and the answers you seek will be self evident.

Mike
 
Michael Bean
You are absolutely right, you are confused. I can only assume this is because you don't understand the comparative example I presented to you. Please reread my earlier post and tell me if you get it now. While you are trying to become more familiar with the English language, I will try and become more familiar with Ohm's Law. I will repeat that I never said there would be no power dissipated in an attenuator at all times when placed at the amp output. I proposed the maximum power dissipation would be roughly half the amp power output because no current will flow at full attenuation and there is nearly zero resistance at full throttle. I asked if this position was refuted and still haven't received a direct reply. There is nothing contradictory in anything I have stated and the aim of this thread is still straight even though it has been buffeted by comments aimed to throw it off track.
 
I was only trying to help. There is a huge store of knowlege and experience available here free for the taking. If you want further assistance from anyone else I suggest you drop the attitude and insults, stop questioning other's credentials and intelligence. Good luck, I'm outta here.

Mike
 
mickeymoose
I can't think of a traditional pot with a wiper that can work like this. No high power amp driving a typical speaker can be controlled this way but the situation could change if we had a line of speakers like in a stadium setting. Okay, I'm ready to rule out wiper pots what about switches?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.