It doesn't take much .
Even psychic vibes can be a medium for entrainment .
Entrainment In Action | healingsounds.com
Listen to some Lou Reed he never performed the same piece twice (the same way)
Funny I had the same discussion with someone over a famous guitarist, his idea of note identical performance had a lot of latitude to say the least.
I would disagree macroscopically and microscopically, but somewhere between the two I would probably agree. Macroscopically, tempo is never precisely called out, only approximate ranges. Microscopically, micro-timing of musical performance is something classical musicians with good teachers learn a great deal about, although they don't call it the same thing I am.
An interesting explanation and demonstration of micro-timing in musical performance can be found here:YouTube The particular section begins about 12 minutes into the talk, although more insight may come from watching the whole thing.
There is very little in the way of explanation as good as that one out there, so I am happy to have found a resource that people can be referred to.
Last edited:
Agreed.
But if you use round base conductor, the nested wire will not add any height, it will fill the interstitial without protruding above the base drive coil.
base coil / pickup coil
#16 #29
#18 #31
#20 #33
#22 #34
#24 #37
Your method would be great on an existing vc, assuming you had the extra room between the coil and the front plate. You would need to co-wind a teflon insulated wire with the 4 mil wire to create a uniform spacing, and remove the teflon wire after cure, or at least after set so you could smooth down the epoxy while it was still pliable but holding the 4 mil wire in place.
totally agree.
But then you also need to span the gaps with something with strength. Because the epoxy would not have good tensile strength, especially if it started getting hot.
As to soldering aluminum, man that is a PITA. If the ends were copper plated, maybe. But definitely not worth the effort.
As to 4 turns/low output...the whole idea of nesting is to make the exact same coil so there was no additional signal processing needed at the amp. Just a second wire pair (coax or twisted pair), a connector, and a resistor into the feedback loop.
Setting the bypass resistor from the output to the feedback means that the amplifier can be designed to work on a speaker with no dvc feedback, but if the owner wanted to, just run the connector and second wire to the speaker, the amplifier would automatically operate in the newfangled mode.
We're talking about a field modification that would take less than half an hour. Break the fb loop, add a resistor, add a connector, solder the wires from connector to the fb/ground.
jn
Last edited:
Thanks for the clarification, now I got it, basically, that is
So, we subtract the velocity voltage (a copy of it, via coil 2) from the total coil voltage of coil 1, resulting in the voltage actually impressed on the transfer impedance which makes a driving current from the voltage accross it, after all.
This would still give nonlinearities in the current as it isn't a stable impedance. If we regulate upon this voltage we'd still have some similar errors terms as with standard voltage drive but we would have an SPL transfer curve (and damping characteristic) that is almost equivalent to current drive as net effect is the same: the output current isn't measured explicitly but still we have a close equivalent of it in form of a voltage needed to develop accross the quasi-constant transfer impedance and which is measured by cancelling out the velocity terms. Which results in the same behavior as current drive to first order but with probably higher distortion?
Where is the actual benefit, what am I missing?
If they are wound front to back, they will not see the same coil inductance through all positions. They will be opposite, so Le nonlinearity would not be compensated. If they are bifilar, that will work but you have double the mass or half the turns of the volume for drive.
The Le is dependent on position. As it increases, the magnetic field sucks in energy from the vc. As it returns, the energy returns to the vc. That non linearity is seen by a co-wound coil exactly as the drive coil. So all non liearities attributed to magnetic stuff is read exactly by the pickup coil and subtracted.Thanks for the clarification, now I got it, basically, that is
So, we subtract the velocity voltage (a copy of it, via coil 2) from the total coil voltage of coil 1, resulting in the voltage actually impressed on the transfer impedance which makes a driving current from the voltage accross it, after all.
This would still give nonlinearities in the current as it isn't a stable impedance. If we regulate upon this voltage we'd still have some similar errors terms as with standard voltage drive but we would have an SPL transfer curve (and damping characteristic) that is almost equivalent to current drive as net effect is the same: the output current isn't measured explicitly but still we have a close equivalent of it in form of a voltage needed to develop accross the quasi-constant transfer impedance and which is measured by cancelling out the velocity terms. Which results in the same behavior as current drive to first order but with probably higher distortion?
Where is the actual benefit, what am I missing?
What remains it the exact voltage component that is in phase with the drive voltage. So in essence, yes we are pretty much driving as if it were a current mode. But with compensation of magnetic non linearities.
jn
Last edited:
> If they are wound front to back, they will not see the same coil inductance through all positions.
> They will be opposite, so Le nonlinearity would not be compensated.
> If they are bifilar, that will work but you have double the mass or half the turns of the volume for drive.
So what for an initial ' 1rst order approximation ' to see if it works at all ?
> They will be opposite, so Le nonlinearity would not be compensated.
> If they are bifilar, that will work but you have double the mass or half the turns of the volume for drive.
So what for an initial ' 1rst order approximation ' to see if it works at all ?
Bingo! Except for Bl(x) and final Bl drop when the coil leaves the gap as you've mentioned. Are there other non-corrected terms, for example does it really also cancel the reluctance force (as per Klippel's definition)? Reluctance force isn't addressed by current drive so if your approach did that additionally could make it favourable to current drive. Thermal issues wouldn't be addressed though, right?
Dead on. It doesn't worry about BL(x) and final drop.
I'll look up Klippels reluctance force, I may have covered it but it's best to be sure..
And, no, thermal issues are not addressed. In fact, my setup also will suffer power compression as I'm not compensating the dc resistance at all.
He defines reluctance force as relating to dL/dt time current squared, which is interesting as it appears to be units of "energy per unit time".
Oh, Yes, mine does. from the eq V = L dI/dt + I dL/dt. The second term is what I believe he is referring to in his reluctance force. I subtract that out, as the voltage that is superimposed on the drive coil from that is entirely dependent on the flux going through the loop, and the pickup has the exact same flux. So it should be automatically cancelled.
(well, at least I think so...)
jn
Last edited:
Hi John,
Who's the dude with the dirty socks?
Now, looking at a low volume production you have time to match the devices in order to extract the promised performance. I've been repairing examples of your design almost my entire life. Having a lot of experience getting top performance from this setup gives me some insight as to what will happen if you try mass production. You would have to use dual, matched transistors. Too bad LM394 doesn't have a compliment! Of course the price difference for those parts makes the product a lot more expensive. Again, you are pointed to the high end, low quantity market.
-Chris
Who's the dude with the dirty socks?
Well, of course. Mass manufacture of your input stage isn't going to be affordable except as a top tier product. A small monitor amp really is a horrible application for your circuit! I can completely understand why your circuit was turned down in two different departments.
Now, looking at a low volume production you have time to match the devices in order to extract the promised performance. I've been repairing examples of your design almost my entire life. Having a lot of experience getting top performance from this setup gives me some insight as to what will happen if you try mass production. You would have to use dual, matched transistors. Too bad LM394 doesn't have a compliment! Of course the price difference for those parts makes the product a lot more expensive. Again, you are pointed to the high end, low quantity market.
-Chris
The more I think about the more I tend to fall back to square one. MFB works best with a completely isolated independent sensor that has a greater excursion range and better linearity than the motor we want to improve. Working sensor designs we find eg. in Backes&Müller speakers, among others. A velocity sensor for woofers isn't that hard to build and install, it could be done for almost pennies in mass production. A vertically installed humbucking coil (figure-of-eight) in a field generated by small Neodymium magnets and associated pole pieces, the coil electrically shielded (but without LF-eddies in the shield, therefore a slotted shield).
Absolutely. Me, I'd just be happy if the thing didn't oscillate...
All good techniques. What kind of bandwidth has been covered using those techniques?
While a sensor is cheap, what about the electronics? And in the event of a sensor failure or wire break, the electronics would have to be smart enough to not destroy the driver.
jn
Dance notation
She's mock conducting and they're staring intently at the score.....
Good enough for a typical midwoofer. A real world example is the "Delphi" speaker by Silbersand where they also implemented doppler distortion compensation wich can be relevant for a 2-way. MFB is especially useful to extend the distortion-free frequency range to way below the resonance frequency at higher excursions. The upper corner frequency is limited, no way around that, once cone (or sensor) starts to break up... game over.
In any competently designed bullet-proof active speaker you need all sorts of protection schemes anyway, so having to add a few more for MFB-specific failure scenarios doesn't really spoil the party, I'd say.
Last edited:
Do they publish distortion specs vs frequency before and after? (actually, I'd be also interested in modulation with lots of bass excursion and maybe 500 hz to 3Khz superimposed.
Pickup within the coil is not limited in frequency, it should reach anything the driver can do. Yes, cone breakup is a stopper always.
The units you speak of, what do they cost?
jn
Last edited:
- Status
- Not open for further replies.