Take woofer resonance. It is damped by a combination of the speaker parameters, the box and the amp. These are an arbitrary combination so I chose my words to preserve the option.
As an example, when faced with a random combination of valve amplifier and speaker I might experiment with output taps to find one that gives a good level of bass, but when possible I prefer to be more organised of course.
All it requires is that the load appear resistive, constant magnitude and phase. Amps will consume power that the speaker won't see, regardless.
I agree, and I put the same effort into my supplies as the amps themselves. The supply should not be modulating the amp. I don't see that as a hazard.. it's simply broken if it does.
Back to damping. An underdamped speaker rings. Allowing too high a Q factor through carelessness could be seen as an error.
On the other hand, bass perception requires time and takes more than a cycle.. Room modes are also resonances.
Ofcourse. Nonetheless it has to deliver that power otherwise the response will get a dip. The more stable the amp can deliver a stable voltage on changing impedances aswell as on phase changes, the less it will change the sound (or -> frequency response).
Yes, ofcourse. But that does not mean they can't sound good, it just means you should use them with a more stable amplifier.
Well, that's not what makes the bass perceived as 'squishy' or 'firm', it's the impulse response and the ability to dampen an overswing.
Hi, this is an interesting article which I think you will find helpful when making your decision about the pros and cons of various output impedances, even negative ones
Variable Amplifier Impedance
I agree, it actually does change a lot. That's a very easy thing to try out for yourself, use a woofer or sub directly connected to the amp with just the cable in it. Listen to it. Then insert a 0,2 - 0,5 Ohm resistor in series and listen again. It instantly gets a bit boomier and the resolution goes down aswell. Or try it with a simulation:
The woofer loses overall spl but between 45 and 90Hz only around 0,5dB while above it loses up to ca. 2dB. Or in other words, you get a 'loudness' effect of ~1,5dB. If you want to simulate it yourself, use a higher Qts driver (this one got a Qts of 0,47) to see it clearly, the difference is smaller on lower Qts drivers.
The cable itself is not much of a concern but the transition resistance (often ~0,2 Ohm) actually is! That's also one of the reasons why a new, 'better' cable is often perceived as improvement sound wise because the old one became a bit oxidized on the open air and thus got a higher contact resistance while the new one isn't (yet). There's a solution to that too, just use good, high pressure crimped cable shoes or snip a few cm off the ends every couple of years to get 'fresh' wire ends for unoxidized copper surface.
Attachments
Well, you can't 'fix' the impedance peak on the fs/fb, at least not with justifiable financial investment, the parts, especally the coils are extremely expensive. Let alone 'fixing' the phase of the speaker. Can you see the problem now?
I'm not saying you should. But there are reasonable examples where it does make sense. Look, not every DIYer who builds amps also develops their speakers too.
Here's the link to Nelson Pass' article on using true current sources with high efficiency woofers.
http://www.firstwatt.com/pdf/art_cs_amps.pdf
IMHO, most good tube amps are not current sources. They're just not as ideal voltage sources.
http://www.firstwatt.com/pdf/art_cs_amps.pdf
IMHO, most good tube amps are not current sources. They're just not as ideal voltage sources.
Last edited:
Well, I'm not sure how it will be of any help if you do as if it doesn't exist. The usual speakers do not have 'fixed impedance'. So the issue actually applies in practically every case, it's just the question if it's neglectable (in most cases) or not. Ignoring the existence of such facts isn't exactly 'creative'. Creative is to find ways to make amps (and speakers) work despite non-ideal circumstances. If you ignore that, you'll sooner or later end up with mismatched pairings of amp and speakers, so please don't get upset if I mention existing possible issues. Remember, we are talking here in a thread where exactly that is discussed, about mismatching or matching amps and speakers. Again, claiming the problem does not exist may probably apply for your amp and speakers but that's not universally a rule.
(BTW, I have a nice collection of lighting ballasts and transformers that I paid nothing for
Well, nice for you. Thank you for contributing to the topic.
This is a diy forum. Many speakers are not usual. If passive XOs are used, care can be taken to minimize impedanxe variations. Joe Rasmussen’s latest revs of the elsinore are an example. FRs often prefer amps with high output impedance. If one is multiampling the higher frequency drivers rarely have non-flat impedance.
So while you argument might play in the commercail world of hifi it carries much less weight here.
dave
It is my understanding of the bigger picture. The apparent mandate of any amplifier output impedance is a fallacy specific to speaker building. Just as the automotive industry refers to condensors which we stopped doing almost a century ago, it propagates without review. It suits the commercial sector (and many DIYers) because it is a convenience. I feel I'm arguing semantics, but if it passes the same current it's part of the same circuit.
Last edited:
That is so not true. None of it. A lot of DIYers build speakers on a budget, to compensate the impedance at the bass peak(s) costs often close the complete drivers itself. And FRs don't 'prefer' high output impedance amps. If they do, you failed to develop them amplifier uncritical. Your claim about the flat impedance tweeters on multiamping only applies to coil-less drivers or already compensated ones.
Sure, which is why I said "What is definitely true is you can find a lot of amps whose adherents LOVE their sound, but the amps "measure bad." I had a couple of tube amps I liked, one of which I loaned to a friend who freaked out because it "was talking to him" in the night (tube resonance I presume, ha ha, it was pretty funny)
...it is very much like vinyl versus digital*: vinyl features many known problems and distortions, but some prefer it's sound. Same for DSD/SACD: multibit can be shown to measure better, but some prefer SACD. Are the preferences due to euphonious distortions? Psychological self-hallucination? Aspects of the sound we have not discovered yet? Unknown.
[And tough to say in the case of media, since to get true comparisons is almost impossible. For example, what sessions have been recorded simultaneously in both DSD and multibit? Or direct to digital and direct to vinyl at the same time? I believe there are a few titles like that, but most all multi-format titles are NOT, and therefore not really valid comparisons of the technology)]
I suspect some tube amps may have less or less unpleasant transient distortions, and clip more gracefully, but I don't recall any research into this. Meanwhile Nelson Pass says he can duplicate a tube sound in transistors using more modern transistor devices and IIRC less or was it zero negative global feedback. Of course, tube amps also have output transformers, which add in another layer of dynamic behavior.
*and here we have to specify well-recorded digital, not the "let's dump the LP-eq'd master onto a CD" or the "let's remaster this and compress it to neutron star density" etc
I beg to differ. Our electronics wiz built a couple variable TransAmps (transimpedance amp with output impedance from near zero to near infinity). Each different FR speaker we tried worked best with a different setting on the amplifier.
Anecdotal evidence that a person’s FR speaker sounds poor until they hook up a higher impedance amp (tube or SS).
diy allows us to play with designs that can take advantage of the lower distortion possible with amplifiers that don’t have output impedance approaching zero.
dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.