Thanks, I wasn't completely sure what to expect either 🙂.
12-14g seems to be the general range I found when researching 12 inch guitar cones. Whether the cones will sound dull is still to be determined, of course. Seems promising though. The thickness and general 'feel' of the cone doesn't seem too far off.
It will definitely be less than that. The original voice coil assembly (coil and former; 16 ohm) from the P12P I have is only about 2.7 grams.
The classic Celestion speakers are approximately 75Hz (lead cone) or 55Hz (bass cone).
Do you happen to recall what the limiting factor was (or, at least, was assumed to be)?
What for?
There even is NO music down there, lowest Guitar string is 82Hz.
And this is not a "woofer" anyway.
MOST important is mid/high frequency response, and the lightest possible voice coil is the best.
No big deal, just manufacturer´s lazyness and unwillingness to experiment , as in: "oh, we did it this way since forever, what are you trying to invent?"
For him "good" was 25 grams and an unpressed thicker than a cookie with the back like the Lunar surface even better, as in "it has more Bass".
I do not understand people´s obsession with Bass, no matter what, no matter if sound is muddier than Hell.
Best Guitar speakers are very light, snappy efficient, jangly, break up all over the place, have a 10dB 😱 or so peak between 2500 and 3500Hz, a dip around 1500Hz, drop like a rock aboge 3000 to 4000Hz, etc.
They are not flat/HiFi reproducers by any means but strong equalizers.
My technique was very simple: the cone molding machine has a manometer showing pressure differential between front and back of the sieve, which is an indirect indicator of cone mass or thickness: at the beginning pressure is low because screen mold is unclogged, free floating fibers swim towards the unclogged areas and pile up there; the process is self leveling.
The more fibers pile up, the more clogged it is, and pressure rises.
Sorry, should have made it clear earlier, fiber "soup" goes through the sieve sucked by a centrifugal pump, not relying "only" on moving the sieve through liquid like on "Art" paper.
I simply turned pump OFF at an earlier point, experimenting until I found the best point giving me 14 grams average, 12 grams minimum.
So here I turned pump OFF at different indicated pressures, pressed and dried the different "cakes", drew a weight/pressure graph and then could predict beforehand any weight/thickness I needed,at least on that machine..
I doubt it has much to do with if there is music down at around 50Hz but a lot to do with how that lower resonance cone reacts, sounds and feels compared to say a 75Hz cone which will be quite different.
I've used the odd 55Hz Celestion in amps where clients have requested them and they do sound and feel quite different to the 75Hz ones. Some people love them, some don't. A lot depends on the amp too,
as it really is a marriage between the two. Type of OP tubes, transformer design, how much feedback (if any) will all come into that equation. Then there is whether you want to drive crazy distortion levels, just have
nice breakup or mostly clean. IME folks that prefer clean to moderate breakup often prefer low to med power Alnico's. At the other end of the spectrum are the metal crowd that predominantly like more aggressive
higher power ceramic Celestions. In the middle are classic rock crowd that often like Greenbacks or similar.
TCD
True, they do sound different.
But the main point is not resonance itself , but that there is different cone thickness, weight, breakup modes, frequency response, all in the mid/high areas which is where the action is.
Ah, I couldn't imagine what the technical reason might be, hence my question.
My favorite speakers tend to be the ones targeting the classic British sound, which pretty much fits that description. Some of the old American speakers are nice too. I have a current-production Celestion G12M-25 and an Eminence GB128 to compare my cones to. I'll post measurements after I get one of my cones installed (may be a little while still).
Good to know. My guess, having only seen some brief videos, was that a measured volume of slurry was pumped through. Pressure differential makes a lot of sense. The correct value for a given cone mass would depend on the surface area, flow rate, and properties of the pulp (higher percentage of fines and more refining [more fibrillation of the fibers] would mean higher flow resistance), but if these are controlled the process should be repeatable.
My scheme forces all of the slurry through the sieve, so the dry weight of the cone is basically just the dry weight of the fiber you put in. Easy to control 🙂
Nice trick 🙂
And yes, "you can find everything in the Net" is a FAT lie; zillions of things never ever show up or at best in a most incomplete form, specially Technical stuff; videos always skip or miss essential steps, etc.
And yes, "you can find everything in the Net" is a FAT lie; zillions of things never ever show up or at best in a most incomplete form, specially Technical stuff; videos always skip or miss essential steps, etc.
Well, after a couple more tests I found that this problem was consistent and seems unavoidable without making some changes. Fortunately, I think there's a simple solution: a vacuum press. I don't need lots of pressure, so I think I can just use a wet/dry shop vacuum as the vacuum source (or possibly a venturi-type vacuum pump, if required). I have not tried actually making a cone using this technique yet, but the idea seems resonable. I made another sieve (out of aluminum window screen, again) with the help of a vacuum storage bag (intended for clothing) and it did indeed seem to press the screen quite evenly onto the rigid mold.
Question (mostly for @JMFahey, since he probably knows): is the wire used in guitar speaker voice coils typically single build or does it have thicker insulation?
I ran the numbers for an 8 ohm voice coil a while ago and found that for the dimensions I was targeting (1.53" ID, 2-layer, 0.35-0.39" [9-10mm] winding width), 32AWG single build comes out a little low in terms of DCR (around 5.9Ω for 10mm width) while 33AWG is a bit high (8.1Ω for 10mm, 7.5Ω for 9mm). 33AWG heavy build is about right at 7.5Ω for 10mm width and 6.8Ω for 9mm width. The gap height in my P12P is about 8mm and the original voice coil winding is 9.5mm.
Using heavy build wire would mean slightly lower BL (and lower efficiency, higher Qts), I imagine. Probably not really an issue in my case. Any other reasons not to use heavy build wire?
Single build is used, because available space inside tiny gap is strictly limited, thicker insulation takes more space in exchange for nothing.
In fact, by what you mention , double and triple build "cheat", it isn´t that they are "higher resistance" because of the coating, simply they use thinner wire inside to keep outer diameter constant.
Don´t obsess with keeping DCR, say, strictly 70% of nominal impedance or something, those are all nominal values, first because they are not that precise, there´s a lot of elasticity involved, and second because us mere mortals must work with what´s available off the shelf. we can´t order 1 Ton of custom drawn wire in some non standard diameter to get exact desired DCR.
I have measured as low as 5 ohm DCR (even one with 4.9 ohm) in some "8 ohm" Eminence speakers, specially some 80´s ones (found in Acoustic and similar amps), a side benefit is that speaker pulls a little more power and is perceived as more efficient.
Only voice coil wires which use an extra coat are self-bonding ones, which add one layer made either of a semi-cured adhesive which is activated with alcohol or MEK or something similar to Nylon, which partially melts and sticks either in an oven or passing current through the wire, think improved hot melt pistol mix.
In your case, 32 AWG is perfect, even if you wind it 9 or 9.5 mm high.
In fact, by what you mention , double and triple build "cheat", it isn´t that they are "higher resistance" because of the coating, simply they use thinner wire inside to keep outer diameter constant.
Don´t obsess with keeping DCR, say, strictly 70% of nominal impedance or something, those are all nominal values, first because they are not that precise, there´s a lot of elasticity involved, and second because us mere mortals must work with what´s available off the shelf. we can´t order 1 Ton of custom drawn wire in some non standard diameter to get exact desired DCR.
I have measured as low as 5 ohm DCR (even one with 4.9 ohm) in some "8 ohm" Eminence speakers, specially some 80´s ones (found in Acoustic and similar amps), a side benefit is that speaker pulls a little more power and is perceived as more efficient.
Only voice coil wires which use an extra coat are self-bonding ones, which add one layer made either of a semi-cured adhesive which is activated with alcohol or MEK or something similar to Nylon, which partially melts and sticks either in an oven or passing current through the wire, think improved hot melt pistol mix.
In your case, 32 AWG is perfect, even if you wind it 9 or 9.5 mm high.
Thanks.
The heavy build is lower DCR. The copper is the same diameter (so same resistance per unit length), but the coil ends up with fewer windings because the insulation takes up a little extra space. The 33AWG single build is specified as 0.0077" diameter while the heavy build is 0.0084". The nominal diameter of 33AWG wire is 0.00708" (0.18mm), so the insulation thicknesses are 0.31mil (7.9µm) and 0.66mil (17µm), respectively. 32AWG single build is 0.0087".
Last edited:
Ok.
Anyway, what´s useful is copper there, anything else is wasted space, so go for "only as thick as necessary" insulation.
In the old days, Jensen definitely used single build.
And again, shoot for as close as possible DCR of course, but you have wide tolerance there and in any case must bow to commercial availability-
I have measured Yamaha 8 ohm voice coils having 7.5 ohm DCR, go figure, but then they can custom order anything they want.
FWIW with just 3 wire sizes: 0.16 - 0.18 - 0.20 mm I can wind a TON of Guitar speaker voice coils out there.
Not into Hi Fi woofers, not into tweeters, so in general don´t need anything thinner or thicker.
Anyway, what´s useful is copper there, anything else is wasted space, so go for "only as thick as necessary" insulation.
In the old days, Jensen definitely used single build.
And again, shoot for as close as possible DCR of course, but you have wide tolerance there and in any case must bow to commercial availability-
I have measured Yamaha 8 ohm voice coils having 7.5 ohm DCR, go figure, but then they can custom order anything they want.
FWIW with just 3 wire sizes: 0.16 - 0.18 - 0.20 mm I can wind a TON of Guitar speaker voice coils out there.
Not into Hi Fi woofers, not into tweeters, so in general don´t need anything thinner or thicker.
I have always wanted to do a hyperboloid of revolution, with a seamless transition to the voice coil at the neck.
A nice structural shape, easy to reinforce and easy to parameterise.
Richard Lee (member here and expert from when the British were a power in speaker manufacture) said he did experiments and it didn't make too much difference.
But I'd still like to try it.
Have you considered to add chopped carbon fibre strands to stiffen up the cone?
My compliments on your work so far!
Best wishes
David
Do we really need a stiff cone in a guitar speaker? I think cone breakup is a quintessential part of the geetar sound.
Best regards!
Best regards!
Yeah, I've noticed pretty wide variability overall. Jensen has a couple 8Ω nominal speakers in the 5Ω DCR range (the Mod 15-120 is spec'd at 4.94Ω), while some of the 8Ω WGS speakers actually have a DCR over 8Ω (the Reaper is 8.2Ω according to their specs).
That's 34, 33, and 32 AWG, so it looks like I'm definitely in the right range 🙂. Thanks again for your expertise. There aren't many people around with actual experience in this stuff that also share information so generously.
Interesting, thanks.
My understanding is that using fibers that are not cellulosic generally requires the addition of binders (waterborne epoxy, for example) to the pulp. Otherwise the non-cellulose material will not bond to the cellulose and will mostly just act as a filler. At least for now, I'll stick to simpler mixtures.
Echoing what @Kay Pirinha said, consider that guitar speakers typically strongly color the sound in specific ways. Addition of carbon fiber, Kevlar, etc. may well make better HiFi cones, but for guitar you actually want a cone that's 'bad' in just the right way 😉
This is worth a read, it covers quite a lot.
https://audioxpress.com/article/voice-coils-a-tutorial
WRT insulation, when I used to work at a transformer manufacturer we used Polyester double insulated for the more critical apps and especially where inter winding voltage gradient was higher.
WRT thickness of double layer insulation, from memory I don't recall this being an issue as it would affect the 'fill factor' but the wire date sheets will have your answer. WRT higher temps, I don't recall using modified Polyester.
This double insulated Polyester is incredibly robust. As an example I wound an EL84 / 30W OP transformer with Bifilar primaries. I proof tested the primaries up to just over 1100V winding to winding and wasn't game to go higher
which is quite amazing, especially for such fine wire.
TCD
100%
I tried some high efficiency, Hi Fi Altec 414C, 12" cloth surround, Alnicos driven by my AC30 and they basically sucked. 🙂
It was a very interesting experiment -> flat, undynamic, compressed and just not 'electric guitar like'.
Weird alchemy is the electric guitar sound. All that non linearity, cone breakup, FR peaking, etc etc is something we seem to like.
I think bmco will find this speaker design will be 95% mod / listen / mod / listen / mod .............................
TCD
That's impressive but seems overkill for a speaker.
Since a speaker uses ~100 V max. I have wondered if aluminium ribbon could just use an anodized insulation layer?
The turn-to-turn potential difference is only a volt or so, and I would trust anodize to handle that.
Very thin so better fill-factor, better temperature resistance and better thermal conductivity, for what that's worth.
It seems an obvious idea, has any commercial manufacturer done this?
Best wishes
David
Yes, "just anodizing" has been commercially used in high Tech very high power speakers, not sure now but Gauss-Cetec might have been one of them.
Yep, interwinding voltage gradient very low on speaker. Was just pointing out how good this insulation can be, it's highly dependent on how you wind and varnish.
WRT Al ribbon, yes have thought about that but not aware of any. Prolly wouldn't be appropriate for 'gita sound' 🙂
TCD