Eminence Flux Density Modulation

[J M Fahey]

Hi Steve, great post.
Much more than what I expected.
I always stated (still do) that ear is the final judge, but any technical measurements backing (or not) what we hear (or think we hear) make for a solid foundation.
In this particular case I am delighted (and that's an understatement) by having real world measurements backing so closely what I predicted.
Going to your graph , I see 12 actual curves, corresponding to attenuation stages 1 to 12 below.
They seem to be color coded on the original, grayscaled in your image, but anyway can be matched:
full magnetic power (no attenuation by definition) must be "Memory 1", which shows the lowest Q (lower=better/louder) ; its impedance peak must be the 26 ohms one, the 40 ohms one strikes me as somewhat high.
As you rise the attenuation (lower magnetic flux) , Q rises steadily, as I predicted, and the peak impedance rises in step, reaching an incredible maximum of 96 ohms.
So far it matches Theory perfectly.
At some point, although Q continues rising, the impedance peak lowers, because the speaker motor becomes too weak to generate a counter electro motive force (CEMF) or induced voltage to counter what the amplifier sends.
Would love to see that chart re-posted in colour to be able to match curves to specs better.
You say those are -2dB steps? Fine with me, the rise in Q from Memory1 to memory10 is incredibly smooth and regular.
A very interesting data line is Memory 11 ; the speaker no longer has a resonance peak !!! ; the measuring software searches everywhere and finds a peak at over 5 KHz, which is certainly there (can be seen clearly on the impedance curves) but unrelated to what we are interested in.
Thanks again for posting.
PS: the phase curves above are also interesting, they show behaviour of a tuned system (the speaker) above and below the resonance frequency.
It can be seen that the amount of phase shift varies a lot depending on attenuation/magnetism, confirming the Q variation.
Nice when different pieces of data back each other.
PS2: please tell me something about the measurement system/software you used. Is WT3 similar to LMS, used by Eminence?

[kg]

Hi kg. Interesting questions.
As far as tube longevity, the tubes are being driven full power, they don't know whether they drive a real speaker, a power resistor or a very weak magnet speaker, I guess they get "spent up" the same with any method, that's the price you pay for full power sound after all.

well, this kind of gets at the heart of the matter.

IF, as you say, the tubes don't "know" what they're driving, then we should be able to substitute an electrically equivalent black box in place of the speaker load, and expect the very same response, right? at least in terms of the amplifier's output voltage/current characteristic?

The impedance curve of a fully unmagnetized speaker must be its DC resistance plus whatever loss/parasitic inductance the voice coil has.

snip

Note: a magnetless speaker will be flat; a weak magnet one will not; the HF part will always be about the same but the resonant peak and the "Q" will not.

indeed, the chart provided by mr coil backs that up.

The voice coil will get hotter, you are right that motion cools it; in fact I already commented here that one employee of mine, probably inmersed in a Jamaican dream or floatin in a sea of beer, has on occasion mounted an unmagnetized speaker mixed with regular ones, the lower SPL was not that much noticeable (logarithmic ear response, etc.) but the cooking epoxy smell gave it away.

hopefully the "reduced magnet strength" method manufacturers (and i say that because it's both fluxtone and FDM) take that into account and either de-rate their voice coil wattage ratings, or add ferrofluid, or some other thermal management.

Don't think the screens will be affected, because there are two "errors" which tend to compensate each other: impedance will be lower, because the reactive component will be lower, given a weaker cone movement (the object of this attenuation); yet as you say, the voice coil might get somewhat hotter, raising the DCR; both effects act in opposite ways and will somewhat cancel or thereabouts.
Tube plates and screens should not be much affected.
I leave that test to you or whoever wants to do it.
Good luck.

i agree they are somewhat counteracting. hopefully they'll cancel each other out!

in my experience replacing reactive speaker loads with purely resistive ones has resulted in premature output tube failure. however, this is not exactly the case here as there is still some inductive component to the load.

[Mr.coil]

JM..I am happy that you found this useful.

Just one thing though..the 96 ohms is the non attenuated reading..you got that 180 out.
As the attenuation goes up the ohms go down...the flatter the lines are = less SPL.

So the most attenuation = just a bump at 130 hz.

I will try to find the original around here somewhere....ya I looked that was on the old puter...sorry

WT3 = woofer tester three

MC

[J M Fahey]

Fine.
I should only like to correct a small point, maybe I didn't spell it as clearly as needed:

Originally Posted by J M Fahey
Hi kg. Interesting questions.
As far as tube longevity, the tubes are being driven full power, they don't know whether they drive a real speaker, a power resistor or a very weak magnet speaker, I guess they get "spent up" the same with any method, that's the price you pay for full power sound after all.

Posted by kg
well, this kind of gets at the heart of the matter.

IF, as you say, the tubes don't "know" what they're driving, then we should be able to substitute an electrically equivalent black box in place of the speaker load, and expect the very same response, right? at least in terms of the amplifier's output voltage/current characteristic?

I only refer to "load" characteristics and health/durability, not sound here.
Your words "response" and "voltage/current" characteristics *might* be misunderstood by someone else not following this discussion, or quoting this somewhere else out of context.
I know it's clear for us, just wanted to state this for the benefit of third parties.
Your finding of reduced tube life on full power + attenuator setups, sounds very reasonable to me.
Fact is, fully overdriven tube amps are very hard on screens.
As of somewhat derating speakers when magnetically attenuated, *I* would test that and post my findings on the label or user manual, but I'm not building them, never will, so I leave that to who's directly concerned.
Personally I'm into "live use" amplifiers, and strive for maximum sound output possible, within realistic production cost constrains.
Bedroom rockers, may God bless them, scratch me the wrong way.
Mind you, I provide a "3AM" switch on the back of my (few) tube amps, which lowers power output to around a still loud (by 3AM bedroom/livingroom standards) 0.1 to 0.5W , but that's provided more as an anti-divorce method than anything else.
On my SS amps, the ones that put the bacon on the table, that's unnecessary.
Good luck.

[J M Fahey]

Hi Steve, thanks.
Waiting for that full color chart, sure will make things clear.[]

[kg]

I only refer to "load" characteristics and health/durability, not sound here.

Your words "response" and "voltage/current" characteristics *might* be misunderstood by someone else not following this discussion, or quoting this somewhere else out of context.
I know it's clear for us, just wanted to state this for the benefit of third parties.
Your finding of reduced tube life on full power + attenuator setups, sounds very reasonable to me.
Fact is, fully overdriven tube amps are very hard on screens.
As of somewhat derating speakers when magnetically attenuated, *I* would test that and post my findings on the label or user manual, but I'm not building them, never will, so I leave that to who's directly concerned.
Personally I'm into "live use" amplifiers, and strive for maximum sound output possible, within realistic production cost constrains.
Bedroom rockers, may God bless them, scratch me the wrong way.
Mind you, I provide a "3AM" switch on the back of my (few) tube amps, which lowers power output to around a still loud (by 3AM bedroom/livingroom standards) 0.1 to 0.5W , but that's provided more as an anti-divorce method than anything else.
On my SS amps, the ones that put the bacon on the table, that's unnecessary.
Good luck.

i agree. i'm the crazy asshole who uses a 600w tube power amp. it does have a triode/UL switch, which knocks it down a tad, but it's still a couple hundred watts.

the best solution i've ever heard to attenuating an amp is to build a large soundproof box to put the whole thing in! or a pair of earplugs...

[Mr.coil]

So KG,

What is the tube complement in your 600 watt amp?

Does it require 220 VAC plug?

Does it weigh 190 lbs?

I am guessing...12 each 6550's?

24 ---6l6's

40--- EL84's

a pair of 813's
one 3-500?

Come on really? 600 watts rms?

[Chuck H]

B A G A

The Big Ass Guitar Amplifier | ken-gilbert.com

[J M Fahey]

I am guessing...12 each 6550's?

Congratulations, you nailed it pretty well: 12 x KT90's.
Too much for me , I don't find *much* justification for an over-50W amp in everyday life, but hey, to each his own.

[Mr.coil]

I like it !

[kg]

it's 100lbs flat, in its aluminum rack mount chassis. basically all the weight is iron! three plate trannies, one filament, two hammond 1650w outputs.

still going on the original kt90s (well, most of them. i had a screen blow out on one and i think i dropped one once.) tried switching over to the sovtek 5881, and they were tough enough to take the voltages, but the sound was flat.

the pisser is that the kt90s are no more. unobtanium. cannot source them, at least for any kind of money i can afford. i got them at cost years ago when i worked for a shop. i think i paid like $20 each. i kick myself for not buying enough for a lifetime supply. stupid me thought they'd be around forever.

the limiting factor is the 750va toroidal main plate tranny. it's from a ss amp. i rewired the secondaries in series to give me 240vac. that, in turn, goes into a full wave doubler. idle plate voltage is ~630vdc. it's biased pretty cold. never heard much of a difference running them hotter. there is a silly amount of filter caps crammed in there... two sets of 4x500uF@500vdc LCRs, stacked one on the other, netting 1,000uF@1kvdc.

since i built the amp, i've actually added some series resistance to the plate supply, which cuts down the continuous/rms output power a bit. about 50w iirc. however that headroom comes back in once the b+ charges back up. basically a big version of power supply sag that only affects the output tube screens/plates (the rest of the amp is driven through two additional separate plate trannies). the compression keeps things reasonably dynamic by letting through the transients.

the other thing the series resistance did was reduce the ridiculous stress on everything whenever i'd power up. there's a 15A slow blow fuse in the mains that sometimes opened. i'd also trip magnetic circuit breakers in the building circuits! blew up a few 1n5407 diodes too in the doubler circuit. ended up paralleling them up to keep them in one piece. iirc the secondary of the main plate tranny was something like 5 or 6 ohms, and i added a 10r resistor. it's been a while.

the amp still cranks! though i have to say that if i get back gigging again there will be a lighter version made... i'm too old now!

[rf7]

I just read some of the patent stuff. It's funny how you can bring back an abandoned technology (Field coil speakers) to do something that would be much harder to do with the current technology.

[Diablo]

I just read some of the patent stuff. It's funny how you can bring back an abandoned technology (Field coil speakers) to do something that would be much harder to do with the current technology.

Most of this forum revolves around obsolete and the mostly abandoned technology of vacuum tubes. The abandonment of one technology for another is usually driven by the lower cost of manufacturing.

[balijukka]

just a coil

I still wonder would it work if you just make a movable coil, perhaps with some core parts, that you can just put on any speaker like a cap. Dial in some current from the power supply and attenuate the original speaker. If you don't need any attenuation, just leave it.
This way you could attenuate your priceless vintage amps without distroying anything.

jukka

[J M Fahey]

I still wonder would it work if you just make a movable coil .... that you can just put on any speaker like a cap

Fine as an "out of the top of my head" idea, but digging into it, the big problem is that speaker magnetic circuits are *closed* circuits, with only the Voice coil minuscule gap as an interruption.
In a way similar to a transformer , where an "external" coil would have no effect.
To affect magnetic flux , in both cases, speaker and transformer, you should wind the auxiliary coil you suggest around the center "leg"; whether it's the speaker pole piece or the transformer core.
In the speaker case, it would not be an "ad-on" at all.
If you wind something in that pole piece, just make it a Field Coil and use it as such.
Good luck.