Hiwatt Output Stages, NFB

[bea]

And that's the key point: a repair by the one company nearby i asked would have been more expensive as the Oberton. So the new speaker is actually the repair. Unfortunately the Kappalite and the Oberton do not sound well together (the Deltalite has chances to sound good with either). So i need a 2nd Oberton, and probably build one or later on 2 cabs for the Neos.

Which will allow me some experimenting: it should be possible to run both Neos in really small cabs, especially the 3015. And that would be *really* interesting: small lightweight yet powerful speakers for my smaller amps and the two "large" ones (close to 606es) for the big irons (the Mywatt and a Dynacord G-2000).

And the 2515: waits until i find someone who will replace the voice coil. Everything else should be hopefully ok; the speaker looks as if it were new.

[Mike Sulzer]

I thought the fins are because neo loses it magnetism at such a low temperature, and so it is necessary to provide a larger surface area for cooling. Bea, did this speaker die from a destroyed voice coil, or did the magnets go?

Problem is thar voice coils by tnhemselves handle very little power.
Rod Elliott shows it: a 40mm one (1.5" , what Jensen 12N uses) reaches 250C and burns with mere 38W (old Jensen paper and nail enamel glue burnt over 100C , boiling water temperature) yet we know speakers standing more than that.
Fact is air around moving VC is turbulent and then some (imagine thousands of tiny hurricanes and tornadoes) , is equivalent to having a strong fan blowing, and carries heat away "as if it were solid silver" (what Fane Engineers measured and their own words).

Where to?
The magnet structure, of course, just hundredths of an inch away.

That's why magnets get hot with high power long use and EV once published 2 power ratings: one in fully sealed cabinets and one (larger) in open back or vented ones.

Enter the tiny lightweight Neo magnet ... now you can imagine the end of the story.

Situation is so bad that designers had to add cast finned heatsinks:

Regular Eminence Delta 15


NASA approved Eminence Deltalite II 2515


sorry about your back, but you will never ever have that problem with your excellent looking Obertons

[bea]

Bea, did this speaker die from a destroyed voice coil, or did the magnets go?

I admit i did not check the magnet - but the DC of the voice coil is high and unstable, and moving the cone by hand leads to a scatching noise (actually the speaker still scratches a bit upon a signal).

[Mike Sulzer]

I admit i did not check the magnet - but the DC of the voice coil is high and unstable, and moving the cone by hand leads to a scatching noise (actually the speaker still scratches a bit upon a signal).

I guess the coil melted!

[Jazz P Bass]

What the heck is "unstable dc" as applies to a VC?

The dc reading of the VC will vary if the cone is moved. (it will also generate a voltage.)

If you can hear the coil rubbing on the pole piece, the speaker needs to be reconed.

[J M Fahey]

I thought the fins are because neo loses it magnetism at such a low temperature, and so it is necessary to provide a larger surface area for cooling.

Yes, that's also a factor, of course, although I much suspect either a burnt voice coil, or one that hit the backplate, hard, because of over excursion (also a VC smasher) , quite possible because of poor loading, trying to get too much bass from relatively little cone surface or both.

In my book a single 15" barely matches a drummer and not by far, and is fine for 60/100W RMS ... 150W tops if you don't push the lowest octave.

Personally when I started making my own speakers, over 30 years ago, I tried single 15" vs 2x12" cabinets side by side, same total impedance and driven by 100W RMS, (the standard play live amp) and 2 x 12" beat single 15" on all counts, be it power, punch, definition, SPL, you name it.
Also separate head/cabinet always beat combos of the same size.

For general purpose Rock Music, a largish 2 x 15" is a powerful beast, and can easily handle 400W RMS .

Latest fade: 500W Class D into 2 x 10" ???????

fine if you also love:


hey!!!!! I was talking about the wind blown umbrella !!!!!
As a symbol of blown outside 10" cones !!!!!!

Tsk tsk ..... it's hard to work with so unfocused people

[Mike Sulzer]

If the DC resistance is high, you have a problem!

What the heck is "unstable dc" as applies to a VC?

The dc reading of the VC will vary if the cone is moved. (it will also generate a voltage.)

If you can hear the coil rubbing on the pole piece, the speaker needs to be reconed.

[bea]

I guess the coil melted!

Indeed, also my speculation.
And i would like to understand what happened. When did it melt - after swichting off the Amp? That would have meant that it was high before - but the last day i could use it i was definitely not louder as i usually am, rather less.
And that was the background why i speculated on possibly induced oscillations in the "ultrasonic" range (which for a woofer running fullrange starts pretty early in the audible range).

[J M Fahey]

Not intuitive at all, and drove me **crazy** until I found it, but there is *one* failure mode where the speaker dies *after* being turned off

If I had seen it just once or twice I would have brushed it off as probable observation error, but when repeated over and over I HAD to find the reason.

Which is "differential thermal expansion".

When voice coil overheats, the wire expands, the former and adhesive too, and worst of all, adhesive crumbles, bubbles or at least softens a lot.
Under that situation voice coil may loosen from former, but stay locked to it because former expands more than copper or aluminum wire, if it has a higher thermal expansion coefficient.

You turn amp off, it contracts more than wire, loses adhesion.

Next time you turn amp on, wire flies away, even 1 mm, and pulls connection open: speaker mutes for good.
Maybe even power amp thump on is enough.

I have found no other explanation for: "speaker goes to sleep working (sort of), wakes up dead".

[bea]

Which is "differential thermal expansion".

A mechanism that should be theoretically possible - the thermal expansion coefficient of Polyimides (Kapton) is about twice that of the aluminum used in the voice coil.

At least two of my amps are theoretically powerful enough to cause this damage, not only on the 2515, but also on the 3015 and the 15L400. And the smaller amps (80/110W) should be sufficient to damage my smaller speakers, especially old Fane twin cone with only 100 W which is usually driven by the 80W amp (which should deliver 90W because the voltage from the wall has been raised from 220 to 230 V).

Inspired by the article Juan M Fahey linked previously (Speaker Failure Analysis) i was looking for docs how to estimate both nominal voltage and power of such protection light bulbs. Diffently posed: what's in the JBL control 1? How many do i need for a large woofer?
We have to deal with about 50 V coming from the mywatt (320 W in saturation of the power stage, and the resistance of the bulb shall not become too large - otherwise it might harm the amp.

Is there any rule of thumb?

[J M Fahey]

I'd go for the trivial solution (in the Mathematical meaning) and use 2 speakers

Or Rod Elliott's suggestion: a peak limiter with fast attack and long release, which ends up being the same but much cleaner.

What you'd really need is some kind of limiter which can calculate real time the voice coil temperature and limit signal accordingly in an intelligent way.

Not easy because a speaker is a crazy very complex load, we are interested only in active power (which overheats the coil) and not reactive power (which does not) and to boot voice coil resistance rises with temperature to complicate Math.

Series bulbs do so, sort of, but in a`very wasteful way, if possible `we want to compress/limit before the power amp.

[bea]

I'd go for the trivial solution (in the Mathematical meaning) and use 2 speakers

Not always possible. Especially not with the 3015 - i will not buy a 2nd one - too expensive. That cab shall be driven standalone - its tone is too different from that of my other speakers and does not fit. And too often You need the sound of that amp but the venue is too small for the full stack. This gig - two weeks before the death of the 2515 - is an example:

(Yes, voice & bass duo - we must have been playing Little Wing)

Series bulbs do so, sort of, but in a`very wasteful way, if possible `we want to compress/limit before the power amp.

which can be more or less forgotten if You use vintage amps or vintage amp designs. (Moreover i am a fan of intelligent cheap yet efficient solutions...)

[J M Fahey]

Thanks for posting that picture, which clears a ton of misconceptions we have been working with.

The first one, and *very* reasonable unless corrected, is that you were working in an loud Rock band environment, with a very loud drummer and an obnoxious guitar player with at least a half stack or so, only justification for your using a 200W Tube bass amp, loud enough to fatally damage a heavy modern PA type speaker, rated, at least nominally, at 300W RMS .

Now I'm out of words.

There is no way that a Bass + voice/sax duo, no drummer, playing in small spaces can damage that speaker.

What am I missing?

That brick wall alone, 1 meter or less behind your floor mounted Bass cabinet provides some 3dB boost at low frequencies, go figure.

So, dear Bea,maybe you can add some extra data so we understand better?

Some live footage might help

[pdf64]

'Yes, voice & bass duo - we must have been playing Little Wing'

Is the whole show done as a duo, or were the guitar / drums in the picture brought into play on other songs? ie thereby likely requiring your bass rig to get pushed a lot harder.

[bea]

There is no way that a Bass + voice/sax duo, no drummer, playing in small spaces can damage that speaker.

The photo has been taken on a test gig on an open stage venue. The other equipment is used by other musicians, and partly belongs to the location - actually i have been switching to the tiny Fender Rumble 30 in the background next to my cab later on when we participated in the following Jam session.

Surprisingly I had to play a bit louder than i expected, we also had some trouble with noise induced by the railway immediately near by (that's in a railway station). Like everyone else playing there.
The speaker did not show any problems - and it did survive that gig.


It died shortly later after having been used in our rehearsal room - with a smaller volume setting of the amp. The room has just 17 m^2, and our PA - one of my old 80W tube amp is not demanded in any way - its level meter nearly won't move when Verena sings.

Of course my first idea was that the guys sharing the room with us (young guys playing some kind of metal *really* loud who at that time had only a tiny bass combo) had abused my amp. For me it would be always clear that i would use it with both of my 115 cabs if i wanted to play really loud. That's why i have them. But these guy swear that they did not touch my amp - and i actually tend to believe them if i would see another explanation. But we're always talking of a sustained load well below the 80 or 90 W my next smaller amp can deliver. I actually do not need the 200 W of the amp - i own the Mywatt because of its sound, and a Mywatt 100 would suffice (but it is a lot harder to find used...)

To sum up: the load at low frequencies should have been on the safe side. Remains the load at high frequencies or some induced noise. Which brings me back to my initial question: increase the stability margin of the amp to make it less sensitive against such situations.