SY said:
Hmmm, it isn't Jerome Lemelson. Jack Kilby is dead. The guy whose name I can't remember who invented ARPANET?
Or maybe the guy who invented the CD demagnetizer?
The guy who will make commercial spaceflight a reality. The man who built the commercial equivalent of the space shutle in 2 years with 20million US $. I need 200k to set aside for my flight. Any sponsors?😉
Jan Didden
Joseph K said:
I did just that, it was I guess 1969 - 70 when I built my first serious DC coupled complementary symetry transistor amplifier. 2 -3 years after building my first serious amplifier of any kind (10 W 6BQ5 pushpull with NFB).
I dropped vacuum tubes then and never looked back.
Rodolfo
Re: Mauro
Absolutely. I rised this issue on several occasions . Given the nasty loads to be encountered in the real world, that an output stage will cope is not to be asumed lightly.
Rodolfo
Jorge said:
Absolutely. I rised this issue on several occasions . Given the nasty loads to be encountered in the real world, that an output stage will cope is not to be asumed lightly.
Rodolfo
Trying to hammer it again
Again, there is nothing mysterious on nature with respect to back EMF. Once the electromechanical model is reasonably accurate, we are simply left with a complex impedance and a complex transfer function.
No matter if it is a physical loudspeaker and output sound pressure, or lumped passive components and electrical output signal.
The rules are the same, the end result is the same. If we cast not well substantiated doubts on this (I mean significant phenomena not encompassed in the model), we are not going to make progress.
Rodolfo
Again, there is nothing mysterious on nature with respect to back EMF. Once the electromechanical model is reasonably accurate, we are simply left with a complex impedance and a complex transfer function.
No matter if it is a physical loudspeaker and output sound pressure, or lumped passive components and electrical output signal.
The rules are the same, the end result is the same. If we cast not well substantiated doubts on this (I mean significant phenomena not encompassed in the model), we are not going to make progress.
Rodolfo
janneman said:
No. If you have other elements across the speaker, such as a
crossover network or another driver, current can flow through
them, but then the speaker is not seeing a real current source.
If you drive just an ordinary voice coil driver with a true current
source, the EMF that might appear on the speaker's terminals
does not induce current, and so does not alter the forces on
the voice coil, neglecting the few pF of stray capacitance and
megohms of stray resistance.
I notice in F2, NP is putting 15ohm parrarel inside to the amp's output. Is this for providing path for back EMF that doesn't go into the amp itself, making better sound? But this amp has no feedback.😀
Is providing solution for any back EMF is just as simple as putting R parrarel inside the output node, providing a circular path for the speaker+back EMF so it doesn't get back into the feedback loop?
This can be a clue. Any tube power amp (with output transformer) has a circular path for the back EMF, that is inside the secondary windings itself. Is this also one aspect that makes tube amps sounds different than transistor amp (transistor amp doesn't have circular path for back EMF, forcing it enters the feedback system, unless you put R parrarel with output like F2)
lumanauw said:
Could you pls explain this a bit?
And 15 ohms is way too much to dampen any back EMF...
Hi, Jorge,
My assumption is the connection between power amp and speaker is always there (like Janneman said). So if the speaker wants to discharge it's back EMF (in back excursion motion that is not part of music program), it will give the back EMF to the power amp, because the speaker never loose connection with the amp.
If there is a R parrarel with the speaker, this back EMF will go circular path with the R, not necessarily entering the feedback system of the power amp.
Maybe this quote from NP himself can help give a picture?
My assumption is the connection between power amp and speaker is always there (like Janneman said). So if the speaker wants to discharge it's back EMF (in back excursion motion that is not part of music program), it will give the back EMF to the power amp, because the speaker never loose connection with the amp.
If there is a R parrarel with the speaker, this back EMF will go circular path with the R, not necessarily entering the feedback system of the power amp.
Maybe this quote from NP himself can help give a picture?
Or you can download the OM of F2 in firstwatt.com. NP wrote that the 15ohm is for changing damping factor. But I think it does more than that in that position.
Nelson Pass said:
Dear Nelson,
even if the speaker slows down after a quick impulse?
in this case the current source stopped generating the current,
but the coil still moves!
and the current from the coil maybe is low, but could impact in some way on the current source (maybe only produces a little heat in it?).
Yes, exactly the "micro" analysis like you do. If between t=0 and t=1, the amp gives current to the speaker to move from 0 excursion to 5mm excursion. At t=1, the amp stopped giving current. The speker's cone will go back to 0 excursion (because it is the nature of "spring", it will go back to 0 excursion when there is no outside force forcing it to excursion).
In the way back of the speaker from 5mm to 0mm, it will give voltage at the speaker's input terminals, because it is electromechanical device, like microphone. If this voltage has no where else to go, it will enters the feedback system.
If the amp has no feedback, the speaker's nature back excursion will be slower than if it is not connected to anything at all, the voice coil and the magnet will peform somekind of "slowing action" than the nature's speed of that speaker going back to 0 excursion.
You can experiment with subwoofer box (sealed). You move the cone with the input terminals open and input terminals shorted. You will feel it is more difficult to move the cone when the input terminals are shorted.
lumanauw said:
Here I'm opening a can of worms - a BIG one!
First, NP made these amps for full range speakers - and no others.
Second, these full range drivers will have to be placed in a specially designed enclosure, or bass will be very boomy.
Third, while I fully agree with the First Watt theory, the amps have been designed to be euphonic - controlled ammounts of low order harmonics to 'sweeten' the music.
In the past I've built something much less sophisticated, but I bet as good as - a triode voltage amp driving what's basically a chip amp.
It will give the same sonic signature - with any normal speaker.
I'm posting a photo of it.
And the 15 ohms resistor doesn't have any unknown property besides improving damping, as NP said.
80 ohms Zout is way too much for ANY box!
Attachments
lumanauw said:
and this is why I do not use neither global nfb nor nfb from the output stage!
😎
Nelson Pass said:
Indeed, it doesn't it see the current source only. It sees the current source (the amp) in parallel to whatever is connected to the amp terminals, including itself.
Nelson, I beg to disagree. The EMF ALWAYS leads to a current. It takes the same path as that current from the amp. If there was no path, your amp output current also would have no place to go. The EMF simply is a voltage source connected to some impedance and will cause current. Changing the amp from voltage output to current output changes the paths available for the EMF-induced current, but not the principle. There is always at least ONE path available, both for the amp current and the EMF, and that is the voice coil. It is clear from the examples I included in my earlier posts. It is quite counter-intuitive that the speaker can be both the 'mover' and the generator that causes current through itself, but it is exactly the same situation with electric motors where the EMF-induced voltage or current opposes the drive voltage or current through itself.
Jan Didden
guys the speaker isnt a generator
any induced emf is proportional to cone movement,rate of current
change,inductance of coil, magnet, etc etc...........
any induced emf is in phase with the signal, however as ive
said the speaker isnt a gen and any opposing current has
a tiny opposition to the main signal
theres no distortion there
yes damping helps but the returns are tiny
cheers
any induced emf is proportional to cone movement,rate of current
change,inductance of coil, magnet, etc etc...........
any induced emf is in phase with the signal, however as ive
said the speaker isnt a gen and any opposing current has
a tiny opposition to the main signal
theres no distortion there
yes damping helps but the returns are tiny
cheers
Re: Detrhoning back EMF
Very educative thread.
Voice coil acceleration equals sound pressure.
Part of the article about MFB:
"A comparison of the speaker's frequency response measurements, (fig.10) obtained by using a measurement microphone and the accelerometer (fig.13), gives compatible results (within 1dB deviation) in the frequency range of interest (20 to 300Hz). Therefore, it may be safely concluded that the accelerometer is directly applicable as a sensing element, in which case no additional correction circuits will be required. In addition, as Novak [7] points out, for a speaker with a diameter less than one-third of its maximum operation frequency, its far-field sound presure level will be proportional to the cone acceleration."
Regards,
Milan
Very educative thread.
ingrast said:
Voice coil acceleration equals sound pressure.
Part of the article about MFB:
"A comparison of the speaker's frequency response measurements, (fig.10) obtained by using a measurement microphone and the accelerometer (fig.13), gives compatible results (within 1dB deviation) in the frequency range of interest (20 to 300Hz). Therefore, it may be safely concluded that the accelerometer is directly applicable as a sensing element, in which case no additional correction circuits will be required. In addition, as Novak [7] points out, for a speaker with a diameter less than one-third of its maximum operation frequency, its far-field sound presure level will be proportional to the cone acceleration."
Regards,
Milan
Attachments
- Status
- Not open for further replies.