[Paul Hynes]

Hi Folks,

There is some interest in powering field coil loudspeaker drive units so I have opened a new thread specifically for this subject. As usual all contributions to the knowledge base will be welcome. More to follow.

Regards
Paul

[engela]

hi Paul,

these are the specifics of the PM6a lowthers.

Resistance of Voice Coil  Re 7.4 ohms
      Free Air Resonance Fo 61.980 Hz
      Emission Piston Area Sd 0.021M²
      Equivalent Volume Suspension Vas 47.192 Ltrs
      Suspension Compliance Cms 753.6 u M/N
      Mass of Cone Mmd 7.0 g
      Mass of Moving Parts Mms 8.75 g
      Force Factor Bl 7.27 T-M
      Mechanical Q Factor Qms 1.578
      Electrical Q Factor Qes 0.477
      Total Q Factor Qts 0.366
      Voice Coil Inductance @ 1 kHz Le 148.2 uH
      Reference Efficiency No 2.278%
      Sound Pressure Level SPL 95.593 Db
the Lowthers have been modified to a field coil design, impedance of the electro-magnet will follow.
Alexander

[engela]

Hi Paul,

The specifics of the electro-magnetic are as follows:

fc=18.2 ohm, inductance 330mH, I=1Amax. U=19V. Power requirments is around 20W
max.

Great to hear your conclusions on the ideal power supply.

Alexander

[serengetiplains]

Paul, I thought I would shuttle over here to ask a question related to your SR8 supplies.  I'm building a pair of full-range field-coil multi-driver OB speakers.  Would you recommend powering each field coil separately?  My hunch is they should be powered separately, but I'd appreciate your opinion on this.

Btw, here is a design I'm toying with.  Drivers will be mounted on a rectangular rear board, which will be bolted to the frontispiece, which is an art-deco design.  Front and rear boards will be aluminum, separated by a sheet of lead.  Side wings will complement the V-shaped front to equalize mean baffle widths along the length of the driver array.

[Paul Hynes]

Hello Tom,

Nice project, almost a line source array! I like the retro look. It should work well. I have several friends now using shortened open baffle line source arrays with 4 x Visaton B200 per side. They are very happy. I considered line arrays using field coil drivers a while back but the cost was above my available funding for the 16 drive units at the time. Still is! I think I mentioned it to Jon at Lowther America but he did not take the bait. Knowing how you work on your system I am convinced this will be one hell of a speaker.

As for how to configure the power supply I will need to know the power requirements of the field coil drive DC coil to assess what is possible with the shunt regulated power supplies.

You will also need to consider signal power amplifier issues. I assume you have no intention of compromising the sound by wiring the drive units in series parallel to give an easy amplifier load. This is so detrimental to the sound quality as the drive units are not properly under control being driven through other voice coils. For the best sound quality the drivers signal input coils must be wired in parallel and this can be an awkward load for most amplifiers.

Regards
Paul

[serengetiplains]

Hi Paul, thank you.  You know, you sold me instantly on your speaker design when I first read about it.  I had by then concluded that an OB speaker must be the ideal, a box being a compromise one would better do without.  FR drivers are coherent.  And multiple drivers reduce distortion, and move air.  I can't think of anything better than an array of FR drivers in OB.  It then occurred to me your power supplies must be quite something if your audio preferences go the direction of the speaker you built for yourself.  That's when I started using your supplies. 

IIRC, the field coil current draw per driver is nominally 1A, with a voltage range of 6-17VDC thereabouts.  I'll get confirmation of those figures. 

I'll take your advice and drive the drivers in parallel.  My amp (being constructed) is a SIT amp with two SIT devices on the output operating in class A.  The amp will operate in what I prefer to call parallel SE, as the SIT devices will operate in reverse phase to each other, but drive separate drivers.  Each SIT will thus drive two drivers in the above array.  Given the SITs' reverse-phase output, they will be connected inversely at the drivers to allow power supply noise cancellation at the drivers' summed acoustic output.  That be the theory so far.  We'll see if it works!

The load presented to each SIT will be 16/2 = 8 ohms nominal, which is an acceptable impedance for the SIT devices (I'll have SIT operating voltage soon for the shunts I asked you to design).

DAC is a TotalDac with four channel output operating in differential stereo.  I haven't figured out the coupling yet between DAC and amp.  I have Feastrex Finemet transformers I'll use if I can swing a proper design for biasing the SITs, as their use would allow some bias-supply noise cancellation if the SITs are biased via the centre-tap of the transformer with the same supply.  The only problem here is that each SIT requires a different bias voltage, though that could be accomplished through separate resistor dividers.  Volume control will be digital, for starters (in the DAC).  If that doesn't work well, I'll figure out some Lightspeed arrangement.

[needsp]

Hi

Wouldn't it be best to power the field coils from constant current sources?

Paul N

[Paul Hynes]

Hi Paul,

The DC coil power supply would need to be able to cope with reverse current from the coil back emf. A series reg would drop out of regulation under these conditions and I suspect a constant current source would too and it would provide little in the way of coil damping either. A shunt reg with an output impedance of less than one milliohm throughout the operating bandwidth can deal with both of these issues.

Regards
Paul

[needsp]

Hi Paul

At the risk of getting intellectually duffed up, I'm going to argue my case....

The field core needs to maintain its magnetic flux density constant, and this flux is a function of the current in the coil only, if the coil is rigidly constructed. Temperature variations in the coil, and magnetic fields from the voice coil will to vary the field coil current- a constant voltage regulator can't keep the current constant, only a CCS

As for its design, that would be your problem! 

Best

Paul N

Stands back to discover what he's missing

[serengetiplains]

Paul, the current draw for these drivers is 1A, with a voltage range of 4.8–16VDC.  Magnetic force saturates at 14V.  I think I will simply pick a voltage at the upper range of magnetic flux generation---a kind of set it and forget it approach.  13V strikes me as a good number. 

[Paul Hynes]

Hi Paul,

I’m open to further education on the issue of driving field coils.

From the information I have acquired so far, on the one hand, the usual method of providing power is a voltage source, and the overwhelming consensus of opinion is that the better the voltage source the better the sound quality. On the other hand there is hardly any information about current source drive to provide for a balanced assessment. This does not mean that current source drive is not an option. Just that it is outside my current experience.

Magnetic system theory is not a subject I have spent much time with, but I do know that there are certain fundamental differences in the way magnetic circuits operate compared to the electrical systems I am used to. Unfortunately at present I have no free time to R&D this, as I have to catch up with backlogged orders for customers due to various materials acquisition delays for a number of my products over the last 6-8 months.

To provide a definitive answer to this issue would require that I acquire a pair of field coil drivers and spend a fair amount of time on R&D to arrive at a working result that is fully assessed empirically as well as theoretically. With my current commitments, this will not be happening this year and I am not, at this point, prepared to invest a large amount of money on the drive units either. I am prepared to work this through should anyone be interested in funding the R&D effort and the drive unit acquisition although it will have to wait until I have stabilised my outstanding work commitments.

With the above in mind Field Coil Driver users are getting excellent sound quality using Voltage Mode drive and the SR8 will still be amongst the best solutions available for the currently accepted method of Voltage Mode power for these drivers.

Regards
Paul

[Paul Hynes]

Hi Tom,

See my reply to PaulN above before you consider how you want to proceed with the Field coil power supply for your loudspeaker project. At this point in time I cannot say for or against current source drive, and it could be a while before I can resolve this.

If you want to get the project running in the mean time with good voltage power, I can provide the new SR1MHD high power shunt regulator modules set up for your drive units rather than fully built power supplies. I am charging £150 for this module. The four modules required for this project weigh in at around 2.7 Kg packed and insured carriage and packing will cost £66.

I have checked the drive unit spec and based on your proposed application I would suggest you use one shunt per pair of drivers, wired in the same manner as you are wiring the signal coils as this will mean the current drawn from the regulator should be relatively constant. With the regulator current set to 2.5 amps, the regulator dissipation will be around 55 watts with no load connected and 29 watts with the pair of DC coils connected. You would need two SR1MHD regulators per channel and a heatsink rated at 0.5 degrees C per watt per pair of shunt regulators for one channel will give a heatsink operating temperature of around 55 degrees C above ambient without a load, which is safe should you inadvertently disconnect the load while the regulator is powered, and around 29 degrees C above ambient with the load connected. Changing the current setting is easy to do but setting it higher would require a larger heatsink. The output voltage can be adjusted via a preset potentiometer on the module so you can vary the voltage to adjust the loudspeaker damping.

Regards
Paul

[serengetiplains]

Paul and Paul, would driving the field coils with a regulator on each terminal help (push-pull, err, without the pushing and pulling)?

[Paul Hynes]

Hi Tom,

This would not work as you think. Using one reg to run two coils as I described earlier will behave like your proposed wiring for the signal coils if that is what you are thinking about.

Regards
Paul

[serengetiplains]

Thank you, Paul.  Would there be any benefit feeding these shunts with a series reg?

[Paul Hynes]

Hi Tom,

The supply line rejection of the shunt reg will be between 100 and 120 dB from DC to 100 KHz. Adding a series pre reg will add approx 80 dB over the same bandwidth. I have done this on my SE1 power amps, but with series regs in both positions, and there has been a noticeable, across the board improvement. Low-level information is significantly clearer. So I would say yes to using the pre reg. The SR1MHD module has both AC and DC input terminals so this can be done with a few modifications to the shunt board. A PR3 module can be set up to feed one shunt reg but you will have extra dissipation in this reg to deal with so a larger heat-sink will be required.

To keep the temperature manageable each heat-sink will have to be rated at 0.25 degrees C per watt for a 30 degree C rise above ambient with two shunt regs and two pre-regs fitted which will power two drive units.

Regards
Paul

[serengetiplains]

Great!  I'll go with 4 AC-fed series regs and 4 DC shunts set to 13V.  Let me know the ACV for the series regs, and I'll get the appropriate transformers.  I have heat sinks in waiting ...

[Paul Hynes]

Hi Tom,

The new PR3HD has full wave centre tapped Schottky Barrier rectification and this will require a mains transformer with two 18vac secondaries for each PR3 module. This give a performance improvement over a single bridge rectified secondary. So you will need 4 transformers. 250 VA will be good. Larger would be better but make sure they are would for low noise operation.

The DC input shunt regs will cost less at £130 each and the new PR3HD modules are normally £120 each I can give you a 10% discount on the 8 modules. This works out at £900. The weight goes up so the insured carriage and packing goes up to £99, unfortunately, making a total of £999.

Regards
Paul

[serengetiplains]

Thank you, Paul.  Payment's sent.