I repeat. You don't think Ohm's law applies to you. If the amp can not deliver
peak currents, you can measure that with a scope. You will find that is not an
issue.
G²
This is beginning to look like trolling.
Dude comes here asking a forum for advice, receives the required information regarding choice of resistors but is warned that he won't get the result he wants: more power into his 5 ohm speakers, as 30W isn't loud enough.
He then gets increasingly irate with the various members who politely attempt to point out his fallacy, asserting that he has studied the subject (and ignoring the fact that contributors to this forum have, collectively, hundreds of years experience and work or have worked in electronics engineering) and that we're not understanding how electricity behaves.
Then when it begins to look like we're unanimous he moves the goalposts and starts blathering on about transients, and how we're only describing DC with P=I^2R. He then goes on to throw a hissy fit and accuse members of ad-hominem attacks. I'll give him the benefit of the doubt and assume that Latin isn't his first language and he doesn't know the meaning of the words.
I don't know why we bother.
Attachments
Yeah, there you go.
If you want more power, you add more speakers.... then the total current will be {(current through the first speaker) + (current through the second speaker)}.
Next is to understand the ideal voltage source. That will help with your thinking around power amplifiers and multiple transistors connected in parallel... always think about power supply ability; the power supply needs to provide an instant infinite current into a 0-ohm load, without sag of the DC supply rails.
The main reason manufacturers parallel the output devices is to obtain low output impedance... (to approach ideal 0-ohm internal resistance), so that the amplifier output stage works as close as possible to the ideal voltage source.
To everyone who are uploading pictures with transparent background, especially images containing dark text,
schematics etc. makes it hard to read them in combination with the forum theme color set to Dark, it will look as follow.
Another example found on DiyA.
The same concerns use of font color, take into account the visibility when forum theme is set to dark.
DiyA forum Style chooser themes are located at the bottom of the page.
Here's one way to change images with transparent background to a solid color using Gimp (a free powerful image editing SW that runs on Windows, Linux, Mac) as follow:
And here's the result.
schematics etc. makes it hard to read them in combination with the forum theme color set to Dark, it will look as follow.
Another example found on DiyA.
The same concerns use of font color, take into account the visibility when forum theme is set to dark.
DiyA forum Style chooser themes are located at the bottom of the page.
Here's one way to change images with transparent background to a solid color using Gimp (a free powerful image editing SW that runs on Windows, Linux, Mac) as follow:
- From menu bar: Image > Mode > RGB
- Set background color to white, or any other color of preference, click on either field to select foreground or background color whereafter a pop-up window will appear where one can set any color of choice.
- From menu bar: Layer > Transparency > Remove Alpha Channel
- From menu bar: File > Export As (from here select preferred options, such as changing image format from PNG to JPG etc., renaming file.. and save).
And here's the result.
Last edited:
Yes. Wikipedia states it well for fixed current sources: "When connected to a load resistance, the current source manages the voltage in such a way as to keep the current constant." A current source will increase the voltage across the load until the desired current is reached. What may be missed is that, in real life, the current source cannot increase the voltage beyond the power supply rails (clipping). It is no longer able to "force" additional current into the load. There may be additional confusion for some as LTSpice defaults to "ideal" current sources that do not have this limitation. They can force the current regardless of the voltage source to which they are attached, even when reverse biased or have a ground connection at the other end. This cannot happen in real life.
Oh no, I'm shamed again!!!
That example was my post! In my defense, I was new to posting at the time. I have since learned my lesson. If only I can remember it! Thanks for posting a general method for fixing the background. Often, the application that generated or edited the figure/picture can do this as well. You know, I could be snarky and suggest that it is easier for you to change your theme. Just kidding. It's the considerate thing to do.
No, it's just that, now some 90 posts in -- a number of which have Absolutely Nailed the point that I think you're missing -- several of us are trying to find a way to convey it, but you continue to 'stonewall'.
It's easy to see that you aren't foolish -- one needs only observe that your 'Reaction Score' is in the 20% range!
Cheers
It's easy to see that you aren't foolish -- one needs only observe that your 'Reaction Score' is in the 20% range!
Cheers
Last edited:
First I'd like to say I too have argued my wrong ideas in many threads, and for sure will again. If that makes you feel any better Tony.
So the idea of paralleling amps is not totally a crazy idea. The fact we are trying to convey is the current does not double up unless you also halve the load impedance. But it seems you are only using the same set of speakers, so the output current will not increase. I think if you had acknowledged that at the beginning, members would both keep advising against it (I would), but also many would support the idea and even I would look forward to hear about your results.
So the little lost in translation about the current capacity and delivery pulled this thread off course.
Yes you will get some more output at the limit of the amp's power b/c there will be lower Zout. (if the amps turn out stable - including no high freq oscillations - when parallel) But it'll be so marginal you might measure it, but not hear it. There are IMO too many risks and not enough practical gains for it to be a good idea.
The available current is according to specs limited b/c of something else... (hint: power supply). The specs for the amps is 45W into 8ohms and 90W bridged into 8ohms. When bridging the voltage doubles, and this requires double the current, so the potential output power when bridged should be quadrupled and the power should be 4*45=180W. The fact that you 'only' get 90W indicates the power supply is not up to the task of doubling the output current.
So if we try to get the thread back to something useful, how about trying to accept that you will NOT get any more current into the same speaker set.
That doesn't mean the experiment isn't worth doing,
So what to look for when paralleling?
The first thing I would do is measure the response of the two amps that are to be paralleled. If they come out identical I would try with 1ohms in series on each output. Then reduce to a lower value in steps. Perhaps ending up with nothing in series. Monitoring with a scope all the time and applying a step input, or at least a series of square waves.
If the amps do not come out identical, say there is 100mV difference at full amplitude (ca +-27Vp for 45W/8ohm) - something very difficult to actually measure and detect with certainty. Then what will the two dissimilar outputs do? 'Fight' each other? Help each other? Are these voltage feedback amps or current feedback? I would try with the same steps as if they are identical, but be a bit more vigilant as the series resistors are lowered.
What can happen and how to do it with out risking destroying the amps is what should be discussed here. Whether the sound is improved or whatever is subjecitve anyways. Personally I don't think it's a good idea, I would bridge them if needing more power. But for the sake of experiments go ahead. Just dont say you will get twice the current and quad the power from only paralleling two amps.
So the idea of paralleling amps is not totally a crazy idea. The fact we are trying to convey is the current does not double up unless you also halve the load impedance. But it seems you are only using the same set of speakers, so the output current will not increase. I think if you had acknowledged that at the beginning, members would both keep advising against it (I would), but also many would support the idea and even I would look forward to hear about your results.
So the little lost in translation about the current capacity and delivery pulled this thread off course.
Yes you will get some more output at the limit of the amp's power b/c there will be lower Zout. (if the amps turn out stable - including no high freq oscillations - when parallel) But it'll be so marginal you might measure it, but not hear it. There are IMO too many risks and not enough practical gains for it to be a good idea.
The available current is according to specs limited b/c of something else... (hint: power supply). The specs for the amps is 45W into 8ohms and 90W bridged into 8ohms. When bridging the voltage doubles, and this requires double the current, so the potential output power when bridged should be quadrupled and the power should be 4*45=180W. The fact that you 'only' get 90W indicates the power supply is not up to the task of doubling the output current.
So if we try to get the thread back to something useful, how about trying to accept that you will NOT get any more current into the same speaker set.
That doesn't mean the experiment isn't worth doing,
So what to look for when paralleling?
The first thing I would do is measure the response of the two amps that are to be paralleled. If they come out identical I would try with 1ohms in series on each output. Then reduce to a lower value in steps. Perhaps ending up with nothing in series. Monitoring with a scope all the time and applying a step input, or at least a series of square waves.
If the amps do not come out identical, say there is 100mV difference at full amplitude (ca +-27Vp for 45W/8ohm) - something very difficult to actually measure and detect with certainty. Then what will the two dissimilar outputs do? 'Fight' each other? Help each other? Are these voltage feedback amps or current feedback? I would try with the same steps as if they are identical, but be a bit more vigilant as the series resistors are lowered.
What can happen and how to do it with out risking destroying the amps is what should be discussed here. Whether the sound is improved or whatever is subjecitve anyways. Personally I don't think it's a good idea, I would bridge them if needing more power. But for the sake of experiments go ahead. Just dont say you will get twice the current and quad the power from only paralleling two amps.
True. The voltage is controlled by Vcc. If an amplifier is asked to provide too much current at its maximum power ( less than Vcc ) then it will be current limited and the voltage will plateau ( or momentarily sag on a peak ).
Adding a second amplifier in parallel provides more current so the maximum voltage is maintained into load that require lots of current.
I guess that using amplifiers in parallel works best for low impedance loads.
Thanks for a reasonable response.
Yes, I did notice that the power supply is the limiting factor in these amps as bridging them only doubles, not quadruples the power.
My concern about bridging them is the impact on the damping factor, however, if I run minimonitors with limited bass response ( -3db at 45 Hz or so ) I might not have to worry too much about that?
And, since I'm sitting on a bunch of amps that I got for close to a song, I figured I'd play around with them.
My first idea was to put a resistor in series that would be just double the output impedance, but as you note, and the literature notes, they all recommend much higher levels and at such levels I need to start looking at the power dissipation over the resistor. If the speaker is nominally 6 ohms, and I add 1 ohm, then that output resistor will dissipate ~14% of the power... is that right? 1/ (1+6) = 1/7 = ~0.14. So whatever gains I might be looking at, they are likely going to warm the house... and that 1 ohm also affects the output impedance the speaker sees ( huh? )... so there goes the damping factor.
I might be better off putting the Parasound Z stuff in the closet and drag out the B1K and A5/60s instead.
There's a certain lack of symmetry in that drawing that bugs me: vide the placement of U101.
BTW: my amps are v1 and v2, I don't think I got any v3. The latter version is more powerful.
I would bet if you're only paralleling 2 channels, your output impedance with parallel will be higher than BTL due to the current sharing resistors.
Jesus H. Christ... BEFORE you start insulting me, you might want to look at my profile.
I asked a couple of questions about running an amp in parallel ( specifically the effects on output impedance ) and a few of you IMMEDIATELY started to tell me not to bother, that it was not worth it, that according to Ohm's Law, blah, blah, blah... then, you all went off on ad hominem attacks.
Look, I know all of that... I just got a bunch of amps and I want to play around with them... Capisce? ( Oh, btw, that is Italian.. and no, I am not Italian ).
You never even BOTHERED TO ANSWER MY QUESTIONS.
If anything, it was YOU who was trolling with your non sequitur answers. Indeed, a number of you all fed off from each other in an echo chamber.
To top it off, you go off insulting.
Nevermind, dude, I got some answers from others. I got half a mind to ignore you from now on... because as I asked:
PLEASE READ THE FULL POST BEFORE ANSWERING.
Otherwise you run the risk of going with non sequiturs.
Oh no, I'm shamed again!!!
Hey, now the world knows you were the other culprit!
;)No worries, all well.
Cheers Michael
- Status
- Not open for further replies.