This does not make any sense at all, just on the basis of some anecdotal information, except perhaps in the case of a pathological amplifier.
It is easy to measure distortion produced by a loudspeaker. That would be the first measurement I suggest. There cannot be a difference, only if the amp can't cope with the lowered impedance. The second would be the FR. The worse the amplifier, the lower the SPL, and the bigger the bumps around resonances, would be my prediction. If you take an amp with a sufficiently large damping factor, the two FR responses will be identical.
If you think these measurements will not reveal anything, than at least a controlled listening test should be conducted.
Originally Posted by RNMarshThe question still remains that paralleling at the speaker with a R (50-75 Ohms) makes an audible difference compared to no added R. Why?
I am running RG-59 75R coax cable (all copper no platings) as speaker cables with 75R at each end. I found the coax to be nice improvement compared to fig8 lamp cord terms of clarity and detail and 'ease' and realism in in the sound. Addition of the 75R resistors to terminate the coax resulted in removal of a very low level 'hash' revealing even more detail and allowing distinctly better focus, positioning and depth rendering.
I compared two resistor types ie 1W carbon film and 0.6W MF and the difference is audible with MF sounding 'clean' and the CF sounding warm but coloured. I don't know if the differences are due to chemistry or parasitic inductance or tempco or what and I know this observation seems odd according to theory but one proof is that when I over drove the system and killed one resistor the shift in balance/imaging and change in 'tone' was distinctly audible and this caused me to investigate and find the dead resistor. The amplifier is a high quality Denon receiver and representative of decent home AV systems so of course is not SOTA but it is indicative of load dependence in such typical systems or is it RF susceptability ?. Before the usual objections this is a dirt cheap 50cent quick no brainer experiment and I recommend that you guys try it just to satisfy curiosity if nothing else and if anybody finds it to be beneficial as I have that would be interesting.
Dan.
Last edited:
Sorry, usually at least one of me is right.
You seem confused. Words are not dangerous, their misunderstanding is (or the concepts they stand for). Everybody who wants to put the effort into it can know its proper meaning Electromotive force - Wikipedia. The word ' force' may be misleading to some, but this Wiki explains it well. It would perhaps be clearer if it were replaced by 'induction', but EMF is the agreed word for the phenomenon.
In the case of a bass reflex enclosure, EMF is lowest at the corner frequency of the enclosure. And it is highest at the points where the speaker is in resonance (one the closed enclosure resonance point, the other where pipe and speaker resonance mutually reinforce). None of them are even near the corner frequency.
I don't get a handle on what you mean with your 'Microphonic voltage' etc. Since this is not an existing term, you may wish to define it first, including in what sense it differs from 'EMF'.
This seems to be an interesting case of anthropomorphisme. How does this current know how much it has to 'refill'.
Of course, when you reverse the phase of exitation at resonance, EMF will detract from the impedance and a large current will flow for a brief period of time.
I am not going to figure out what you are trying to say, and not because I don't want to. One thing for sure is that you at least need complex algebra because cone velocity, voltages and currents are all out of phase with each other.
Last edited:
You responded to a post by syn08 where he complained about the (obviously he meant silly) approach to EMC by the 'conformal' coating and shielding of ICs, for a pcb "otherwise perfectly good engineered" with:
"Conformal coating, tin foil. Of course, all the jumpers are still untouched and so is the 20 pin JTAG connector among the others that have been pointed out. What a joke."
Therefore my question as it seems to be an inherent contradiction.
You might call it "semantics", but isn't it in reality not just asking for evidence for some quite brisk statements?
If it were just comments about "it appears to be ....." it would have been nice to include an indication for it.
Nothing wrong with opinions ..... 😉
Last edited:
No. The reference point is current drive i.e. no electrical damping, just the speaker's own mechanical damping. This is DF=0. Then higher DF (lower amp output impedance) means more electrical damping.Joe Rasmussen said:
Maybe you have changed an RF resonance? Such an RF resonance should not be there, of course, but amps are not always perfect in this way because too few audio designers know anything about RF. Maybe the sound only 'changed' because you knew the resistor was present?Max Headroom said:
Sigh. OK, it seems we have to teach a world-class audio designer one of the basics. For DF to be useful it has to be at a low frequency, say 100Hz. It has to be scaled by the nominal speaker impedance, not necessarily 8 ohms.john curl said:
Who cares about HF damping factor? At those frequencies there is quite weak coupling between electrical and mechanical domains in the speaker, so no use for and no need for electrical damping.
Most people reckon that DF of 20 is enough, beyond that you get diminishing returns.
Of course. A speaker design which has sufficient mechanical damping needs little or no electrical damping. However, most speakers are designed for voltage drive so need high DF (i.e. greater than 20).
and completely unnecessary.
Flattens the current phase angle.
See below, while this is a single driver (Red), you can see what happens when you add a significant parallel resistance (Blue), 8 Ohm driver and 8R parallel resistance, so it becomes a nominally 4 Ohm load. Fine for most amplifiers.
This works even better than adding series resistance, but is also more wasteful than using series impedance.
This time series 18R added, but compare with the above and parallel is the better result. Also, voltage swing is severely curtailed with large series resistance.
I know this is going to cause flak, but this is about the amplifier 'reacting' to the speaker's load. The Re (DCR) of the speaker will be about 5-6R and anything above that in the impedance measurement is caused by back-EMF, basically a voltage source.
If you can whistle into the speaker a tone that matches that from the amplifier while measuring the current, and as you increase whistling (this is of course a thought experiment), then you will see a reduction in current from the amplifier.
That back-EMF is also what causes the current phase angle, since the Re part of the impedance cannot. The Re is only resistance and is not reactive, indeed the Re value limits the effect of back-EMF on the amplifier.
This may be difficult for some to grasp, but it sure is not rocket science.
Repetitive nonsense. If you add a parallel resistor to a voltage output, speaker current does not change at all. See R15 stepped from 8ohm to 1008ohm. You are trying to manipulate reality and people. Speaker current goes through R13. Only i(R13) + i(R15) is changed, which is unimportant IF you have an amplifier with high DF and output current.
Attachments
There seems to be too much snake-oil selling going on at the moment from the usual suspects, a little bit is amusing but they've taken over and others are largely quiet, due to boredom probably. Still, hat's off to those who refute it, it's a shame it's necessary.
The other day, when you were talking about 'circulating' currents, was that also about the amp?
Ideally, everybody else should temporary leave and let them sell the snake oil to each other. And FUD each other.
That would be quite a show to watch.
If these tests were not level matched and individually equalized to the same frequency response this is no more than a "fun" exercise no better than the highly criticized "forum ABX". See for instance the Carver challenge.
Yes, 6 ohm and 18 ohm output impedance will have a nice bass-boom boost. Some may prefer it 😀. No dry or sterile sound, then 😀 😀
Oh yeah, and amplifier output current phase is flattened. Why not to use a 1 ohm parallel resistor? Even more flat amplifier output current!!! 😀😀
If I were Joe, I would go for it. Or let's use a 1 milliohm shunt.
If I were Joe, I would go for it. Or let's use a 1 milliohm shunt.
Agreed. Motional back-EMF, mainly at LF. Inductive back-EMF a la Faraday causing that typical rising impedance. And then microphonic back-EMF which can also be caused by resonances etc. A badly designed speaker with high distortion will cause damaging back-EMF that has to be dealt by the amplifier.
If you look at the amplifier requirements as simply a logistics exercise, then you are correct. Absolutely. So on that topic, we are in agreement.
But there is another mechanism at play here. There is something happening and there seems to be a reluctance to consider it.
What that mechanism is, is not related to the capacity of the amplifier to behave in a linear way on the voltage side. On that side of things, if the logistics are in place, beefy output stage and power supplies, there indeed is not a problem.
This [other] mechanism is related to the simple fact that the force is actually the current and not the voltage. The driver is a current device. So yes, you are right that there are frequencies where the load is entirely resistive and you were right to correct him. But we need a more complete picture of how the amplifier reacts to the speaker, the speaker distorts the current of the amplifier and then that altered current affects the speaker. Speaking pictorially, a messy situation.
Mind you, add to the Re of the speaker, you will reduce that, and if the amplifier is held at zero degrees current phase angle, that would be a real improvement. The speaker will still distort, but it will have difficulty is distorting the current of the amplifier.
Parallel resistance is arguably better than series.
Last edited:
That appears to be your definition of back EMF, so it is a tautology.Joe Rasmussen said:
Putting an 8 ohm resistor in parallel with a nominal 8 ohm speaker may reduce any impedance peaks, but it also deepens any impedance troughs. Whether this makes the load easier or harder to drive depends on the details of the amp, but in many cases it will be backwards step. Nothing of any consequence here.
EEs call this 'amplifier output impedance'.
Nice to be patronised by someone who thus far has given us no evidence that he understands complex numbers and hence is capable of understanding impedance.
You show me where you got your opposite claim. Mine came straight from 'first principles' thinking, which I guess is what you claim to be doing except you keep getting the basics wrong.Joe Rasmussen said:
Merely repeating confused thinking over and over again does not make it more likely to be true, but it may pick up a few naive fans. Good for marketing; bad for science.
Well we call it "real" and "imaginary" part, he is just holding us to the strict use of imaginary. 😀
This is wrong, where are you getting it from or are you just making it up?
- Status
- Not open for further replies.