I'm building a high voltage tube PSU for a pair of full range field coil drivers.
I am connecting the drivers in parallel. Now my question is: which would be a better topology: LCLCLC, with the last capacitor doubled to each driver for decoupling or would it be better to do LC(RC)LC, with the last LC doubled to each driver. I only have room for 3 chokes in the case, hence the limitation of trying to accomplish it with 3...
Thanks
I am connecting the drivers in parallel. Now my question is: which would be a better topology: LCLCLC, with the last capacitor doubled to each driver for decoupling or would it be better to do LC(RC)LC, with the last LC doubled to each driver. I only have room for 3 chokes in the case, hence the limitation of trying to accomplish it with 3...
Thanks
Each driver gets it's own LC. When you hook a scope to the field coil and feed a signal to the voice coil, you see that signal in the field coil. So the field coils need to be decoupled if they share a supply. I have had best results with a CCS for each driver. Mounted close to the field coil to get the high impedance node close to it's load where it belongs.
I repaired a 1940's Philco radio a few years back -- the field coil was in series with the power supply B+ (i.e. no choke).
That's a standard configuration in radios with field coil drivers. The field coil is the choke. Did it have an extra winding, configured to cancel hum? Lots of radio field coil drivers had these.
There was a humming coil canceler: on coil close to field coil wired antiseries with the moving coil, other used a shorted turn in the field coil. Didn't know that nowadays there are still such a speakers.
Osvaldo - could you explain the mechanism by which "canceler coil" cancels hum?
I thought the opposite was true: a well-regulated voltage source, or one with large output capacitor, provides a short for a signal coming from VC due to VC-FC magnetic coupling.
Since for DC the field coil is just a resistor, and Ohm' s law applies, voltage source should be as stiff as current source.
I use regular tube-rectified supply with no ill effects, as long as it has good filtering.
Interesting discussion of a power supply design, where the load requirements are a mystery. So what's the impedance of this coil, that happens to be wound around an iron shaft, anyway? There must be both an inductive and resistive component to it; I could guarantee it's not "just a resistor". What's the range of current the speaker needs, in order to operate?
There isn't much mystery. In a typical old time FC speaker, which is high sensitivity, the field coil is pretty much static and there is negligible interaction between FC and VC. FC is fed DC current, so there is no impedance here to talk about, it is simple resistance.
How much magnetization speaker needs is a good question. With everything else equal, magnetic field strength in the gap is directly proportional to speaker sensitivity. So, the more the better. But there is a limit, which is determined by magnetic saturation of pole piece or yoke, whichever has smaller cross-section. Therefore, in a FC motor, after the magnetic circuit is saturated, any further increase in FC current will not increase magnetic field in the gap.
Even more critical limiting factor is Joule's heat generated in the FC. 12" 20W Jensen Concert speaker has conservative FC dissipation of about 20-25 W, at which the coil becomes hot to touch. At this power, the magnetic circuit is only about half-way to saturation.
How much magnetization speaker needs is a good question. With everything else equal, magnetic field strength in the gap is directly proportional to speaker sensitivity. So, the more the better. But there is a limit, which is determined by magnetic saturation of pole piece or yoke, whichever has smaller cross-section. Therefore, in a FC motor, after the magnetic circuit is saturated, any further increase in FC current will not increase magnetic field in the gap.
Even more critical limiting factor is Joule's heat generated in the FC. 12" 20W Jensen Concert speaker has conservative FC dissipation of about 20-25 W, at which the coil becomes hot to touch. At this power, the magnetic circuit is only about half-way to saturation.
Sounds like there's room to play dynamically. Is there a sweet spot with the magnetic flux? If so, that would be something a normal speaker couldnt do. If it all sounds the same, only efficiency increases with flux, then you could make a dynamic range expander by varying the field current in proportion to the music loudness envelope. But how quickly could you change the field coil current, before the inductance of the winding starts getting in the way?
Gee, having that field coil where a B+ choke normally goes would increase the current through it, as the amp plays louder. If you were able to time the fluctuations- kinda in a similar way as they time a guitar amp's B+ crush to be in a musically useful rythmic space - it might work as a system where sensitivity goes up automatically with program loudness.
I don't think so, the more the better.
Yes, that's what's going on with series-fed FC speaker powered by Class AB amplifier. It works as dynamic expander. Of course, speaker sensitivity changes in response to average DC current draw by amplifier's output stage, not in response to output signal waveform.