Welcome! Need support, you got it. Or share your ideas and experiences.
Return to Planar Speaker Asylum
66.87.7.33
In Reply to: RE: "If we replaced the wire with foil that had 1/4th the mass and 4X the resistance per inch" ... posted by andyr on June 25, 2011 at 05:09:51
Hi andyr,I hadn't really gotten around to discussing actual sizes and weights of foil, or the size of the wire that it would be replacing.
The conclusion I reached was only that _IF_ much-lower-mass conductors could be used, then we could still wire-up a driver with a practical impedance AND much less mass (although it would then require a lot more power to achieve the same current through the conductors, which is what determines the force exerted on the conductor by the magnetic field.
The foil that is available could have a thickness of 0.001-inch or 0.0005-inch, which is .0254 mm or .0127 mm thick.
The 31 ga wire has a cross-sectional area of 3.1416 x (.22606/2) x (.22606/2) = .0401 sq mm.
To get the same cross-sectional area using .0254mm-thick foil the width would need to be 1.58 mm (= .0622 inch = about 1/16th inch).
To get the same cross-sectional area using .0127mm-thick foil the width would need to be 3.16 mm (= .124 inch = about 1/8th inch).
So, when replacing 31 AWG wire with foil, it looks like we'd be cutting things pretty close if we tried to get 1/4th the cross-sectional area, because even with the .0005-inch (.0127mm) thickness it would require a width of only 1/32nd inch (.03125 inch), or 0.79375 mm.
Maybe someone can calculate what rating of FUSE that would make, or how hot it might get.
I just measured the wire on the mid/bass panel of my MG-12/QR speakers and it measured as 0.022-inch diameter, which means it is probably 23 AWG with a diameter of .0226 inch, or about 0.57404 mm, with a cross-sectional area of about 0.2588 sq mm. (It would take a foil width of 10.19 mm of .0254mm-thick foil to get the same cross-sectional area.)
So if we used the .0254mm-thick (.001 inch) foil with a width of 2.547 mm (about 0.1 inch) we'd have a cross-sectional area of 0.0647 sq mm. That's 25 PERCENT as much cross-sectional area as the 23 AWG wire. So it would have 4.000 TIMES the resistance of 23 AWG wire, per unit of length.
I need to go back to basics:
Resistance of an aluminum conductor is R = pL/A, where R is in Ohms, p is 2.6 x 10^(-8), i.e. .000000026 Ohm-meters, L is length in meters, and A is cross-sectional area in square meters. From that, the Ohms per meter would be .026 / (area in sq mm).
So one meter of the .0254 mm x 2.547 mm foil would have a resistance of .026/.0647 = .4019 Ohms per meter. So, to get 3.9 Ohms of resistance it would require 9.705 meters of foil, or 31.84 feet.
But since we'd be replacing existing wire, the length would already be decided.
Let's see... Ohms/meter = .026 / (thickness x width). And (Ohms/meter) x length(in meters) = Ohms [so Ohms/meter = Ohms/length(in meters) and length = Ohms / (Ohms/meter)].
So, with aluminum foil length in meters and foil thickness and width both in mm:
(1) width = (.026 x length) / (Ohms x thickness)
and
(2) thickness = (.026 x length) / (Ohms x width)
So, with those equations, we could calculate the needed width or thickness of the foil, if we specified the thickness or width and knew the length and resistance that we were replacing. (Also, remember that the length is in meters but the thickness and width are both in mm.)
Also:
(3) Ohms = (.026 x length) / (thickness x width)
(4) length = Ohms x (thickness x width) / .026
EXAMPLE:
My MG12s have 32 bass/mid wire runs of about 42.5 inches each, which is about 34.544 meters. The driver I measured had a DC resistance of 3.85 Ohms. (Also note that there are only 26 vertical wire positions, but six of those have two wires.)
If I wanted to replace the MG12 bass/mid wires with foil and get 3.9 Ohms total, and wanted to use .001-inch-thick foil (0.0254 mm), then the foil's width would need to be:
width = (.026 x length) / (Ohms x thickness)
= (.026 x 34.54) / (3.85 x 0.0254)
= 9.18 mmThat obviously wouldn't be a practical width but I wanted to see if it came out to about the same cross-sectional area as the original 23 AWG wire. And 9.18mm x .0254mm = .2332 sq mm, which is close to the .2588 sq mm of the original wire. Note that my measurements of the speaker's wire length were not very accurate.
Since length = Ohms x (thickness x width) / .026, the length of the original wire should have been 3.85 x .2588 / .026 = 38.3 meters
----------------
Now armed with some numbers and equations, and GOING BACK to the example in my earlier post, about using 1/4th the cross-sectional area and using two paralleled drivers of half the original panel size and about 8 Ohms each, let's say that the original MG-12 panel has room for 26 vertical runs of about 43 inches each, or 28.4 meters. Half of that would be 14.2 meters.
To get 8 ohms with 14.2 meters of aluminum, the cross-sectional area would have to be .026 x 14.2 / 8 = .04615 sq mm.
The original 23 AWG was .2588 / .04615 = 5.6X larger.
Anyway, to get .04615 sq mm with .0254mm-thick foil the width would have to be 1.83 mm (.072 inch).
THEREFORE, if we replaced the wires of an MG-12/QR's mid/bass panel with .001-inch-thick foil that was 1.83 mm wide, and wired it as two half panels in parallel, we would get a 4 Ohm driver with only 17.8% as much conductor mass as the original wires had.
[EDITED: Had measured the gauge of the MG-12/QR's bass/mid wires incorrectly. Correct diameter is .0226 inch; 23 AWG. Fixed.]
Cheers,
Tom
Edits: 06/26/11 06/26/11Follow Ups:
What we need is for someone to measure the width of the Magnepan foil, or the resistance of a known length, because per Wendell we should use the same foil width Magnepan found worked best in their R&D effort. This will require some kind of series/parallel arrangement to get the impedance right. But we won't know what that will be until we have some measurements.
Josh,I guess it depends on what you mean by "the Magnepan foil". All I have is a pair of MG-12/QRs. But in case it might be helpful, I measured their tweeter foil:
On my MG12s, the tweeter foil is exactly 0.1-inch wide (2.54 mm).
There are eight vertical runs of foil, with the outer two going to the connectors. There is a space of exactly 0.125 inch (1/8 inch) between each pair of vertical foil segments (so, 0.225" between their centers), except that the last vertical foil segment on the right is 0.250 inch from the one to its left, instead of 0.125 inch like the others.
I did measure the total foil length (8.146 meters), and the resistance (3.60 to 3.66 Ohms), as precisely as I could. The calculated thickness, with measurement variations, came out to be in the range of .0009 inch to .0011 inch. So I'd have to assume that it is actually a .0010-inch thickness.
I measured the length of the MG12 tweeter foil from connector to connector, as 8.146 meters (320.7075 inches). [No, my measurement precision wasn't .0001 inch. I just kept the four digits from the conversions of 16ths etc to decimal fractions, until after I added everything up.] (The lead-ins from the connectors totaled 16.8125 inches. The length of each of the eight vertical sections is 37.6875 inches. There are six horizontal runs of 0.325 inch each and one of 0.450 inch.)
I unplugged the tweeter wires and measured the resistance, with my Tektronix DMM916 multimeter. Of course I also measured the resistance of the meter's probes and wires by shorting the probe tips, and subtracted that from the tweeter resistance measurement, each time. I repeated the entire process until the measurements stopped changing (five or six times). I got 3.6 Ohms, for the one tweeter that I measured (only one sock is off).
I also tried measuring the resistance of the tweeter foil using my TH2821A LCR meter, and got 3.66 to 3.65 Ohms, using frequencies from 100 Hz to 10 kHz.
To calculate the cross-sectional area, I used the equation A (in sq mm) = [0.026 (for Aluminum) x Length (in meters)]/Resistance (in Ohms). I then divided the Area by 2.54 mm to get the thickness in mm. Then I divided that by 25.4 mm/inch to get the thickness in inches.
Cheers,
Tom
Edits: 06/30/11
Turns out the MMG and 12 use the 1 mil foil. Despite the fact that his replies are in red so they won't be missed by idiots like me, I managed to read right past it. Probably paying too much attention to the dancing girls.
But, hallelujah, this gives us exactly what we need! With your measurements of the 12 it remains only for me and JBen to measure our new and old MMG's respectively. I've been unusually busy for the last week, but I'll make it a priority to do mine within the next few days.
No sooner had I begun to uncover one MMG than my wife warned we had visitors on their way Sunday afternoon. So, it all got moved up, again. However, I kept thinking that I must have saved my early measurements, at least the ones that I needed for the PLLXO-based bi-amp calculations.TaDahhh! I just found them and I also rechecked the impedances from outside each MMG. They match what I measured long ago. At the time, years ago:
a. Both tweeters measured 2.61 ohms. This was 6 runs, 38 inches long each + 7 inches in bottom traces to the terminals. Aprox total=235+ inches.
b. The mid/bass panels measured 4.1 & 4.14 ohms, respectively. There are 24 runs, of which 14 are double. Therefore, the equivalent of 38 runs, each 37.75 inches long (extending 38" when as twin loops). There are about 16.25 inches of additional connecting traces at the bottom. In total, roughly 1451+ inches.The actual lengths are probably just slightly longer because I don't think I added the short horizontals at each end. I do remember measuring at the looping (U) center between two mid/bass panel traces, that should be close enough.
In fact, I did a quick proportional extrapolation with the numbers that Tom_G supplied earlier about the 12s. Tom is not clear if he included the traces to the terminals but there's enough thoroughness to suggest he would not have missed it. Electrically, the mid-panel numbers are right on. Using his figures backwards, mine should be 4.1~4.16 ohms, which is a virtual match.
The tweeters are also close enough. Backing up from Tom's numbers I get that my tweeters would have been 2.64 ohms, I measure 2.61 ohms...so yeah, ballpark.
Therefore, it seems that my old MMGs and Tom's 12s are made of the very same type of wire and foil.
The newer MMGs probably are, as well. If so, it then means that their longer tweeter runs (vs old MMGs) would have around 3.5 ohms impedance, close to the 12s. OTOH, the newer MMGs mid/bass would be able to match the old MMGs impedance by using 22 rows of wire, 16 of them doubled for a total of 38. Is this the row count? If it is, you should see around 4.1 ohms there.
In other words:
- New MMGs with 8 tweeter runs and 22(38) mid/bass runs?
- Old MMGs with 6 tweeter runs and 24(38) mid/bass runsCuriously, it appears that the 12s use less wire for mid/bass than the MMGs.
Anyway, for QR on my mid/bass panels, it looks like I'd have enough "run length" for 24 runs of 38 inches (plus about 16" for horizontal & terminal runs). Using the same tweeter QR material, I would get a total impedance of about 10.5 ohms. Should I just thank God that I bi-amp at line-level or are we getting more suitable QR strips from Magnepan?
Edits: 07/12/11
Excellent, I'll measure my own MMG's tonight or tomorrow at the latest and we'll see how they compare.
I think, with a DCR of 10.5 ohms, we're talking about paralleling the two halves of the woofer panel. That way, DCR will be 5.25 ohms, which is a bit high but not horribly so. The newer MMG's should be very slightly lower, owing to the missing woofer loop.
Huh! In parallel, how did you wind up with 5.25? If I split the runs in 2 equal, my finger calculator says I'll wind up with less than 3 -- ohms not fingers : - ))
LOL sorry I must have been half asleep that night. Or maybe distracted by the dancing girls?
Cool, thanks, some figures at last! The only question is, does the 12 use the same foil Magnepan is offering us? Once we have that, we can figure out what kind of series/parallel arrangement we need to get the resistance right with the new foil.
Wendell also wrote to say that while 1 mil foil would sound great on the woofer, it would be too inefficient, and that he's guessing that 2 - 2.5 mil foil would be right for the woofer. I counted 22 (doubled, he reminds me) wire runs on my MMG and 8 foil traces. So I get 2.75 times the thickness of the tweeter to keep the resistance/unit length and mass changes proportional. If we used .5 mil foil on the tweeter, that would come out to a single 1.5 mil trace on the woofer. But since there may be other factors that I'm missing, I'd say Wendell's guess is a better starting point.
Since the total available foil length should be known (based on the panel where the wires are being replaced), I derived an equation that can be used to find, from a desired total length of foil, the cross-sectional area needed to get two identical parallel foil runs that result in a 4 Ohm load:Area = thickness x width = length x .001625
where length is in meters, Area is in square millimeters, and thickness and width are in millimeters.
If the desired width or thickness were known, then that could be rearranged to one of the following:
thickness = length x .001625 / width
or
width = length x .001625 / thickness
Or, if you had foil with a known thickness and width, you could calculate what total length would need to be divided in half to get 4 Ohms when the two halves were run in parallel, with:
length = thickness x width / .001625
The following conversion formulas can be used:
mm / 25.4 = inches
inches x 25.4 = mm
meters x 39.37 = inches
inches / 39.37 = metersExample:
If the optimal width that Wendell mentioned turns out to be 0.1 inch, i.e. 2.54 mm, we can calculate what total lengths would give 4 Ohms when divided into two parallel runs, for each of several different available thicknesses:
.0005" (.0127 mm) x 0.1" foil: .0127x2.54/.001625 = 19.85 m (65.13 ft)
.001" (.0254 mm) x 0.1" foil: .0254 x 2.54 / .001625 = 39.7 m (130.25 ft)
I'll stop there, since .0015-inch by 0.1-inch foil would need 59.55 meters split into two parallel runs, to get 4 Ohms.
-------------------------
With only a tiny bit more work, I was able to derive the equations for n aluminum foil runs in parallel:
(1) Rn = .026 x Ltot / nnTW
where Rn is the resistance in Ohms of n parallel foil runs, each with length Ltot/n (in meters), and Ltot is the total foil length in meters, T is the foil thickness in millimeters, and W is the foil width in millimeters. Note that nn is n x n which is "n squared".
The derivation for n parallel foil runs was easy, if we can assume that we want the same current through all runs, and therefore we have to assume that their resistances are equal. For equal resistances, the parallel resistances equation [Rpar = 1 / (1/R1 + 1/R2 + 1/R3 + ... 1/Rn)] boils down to Rpar = R/n.
(We could also use shorter lengths of thicker foil, for example, to get the same current, which would give the same force. But then we'd be accelerating a different mass, so the acceleration would be different, even though the current was the same. We could calculate the difference in mass needed for a different current, to get the same acceleration from a different length, but then other aspects of the behavior of the membrane would probably still be changed. Maybe someone should figure that out, though, because it would open up an infinite number of options, if it could be done that way while still retaining satisfactory characteristics.)
If we assume that we want to end up with 4 Ohms, that gives us the following different rearranged versions of the same equation:
(2) Area = T x W = .026 L / 4nn
(3) T = .026 L / 4nnW
(4) W = .026 L / 4nnT
(5) L = 4nnTW / .026
Examples:
1) What total lengths of .0005 x 0.1 inch foil would give 4 Ohms when divided into different numbers of parallel runs?
TW = .0127 mm x 2.54 mm = .032258 sq mm
n = 2: L = 4 x 2 x 2 x .032258 / .026 = 19.85 m (whew, same as before)
n = 3: L = 4 x 3 x 3 x .032258 / .026 = 44.665 m2) What thicknesses would work to get 4 Ohms from n parallel runs of 0.1-inch-wide foil that totaled 28.4 meters in length?
T = .026 L / 4nnW
n = 2: T = (.026 x 28.4) / (4 x 2 x 2 x 2.54) = .01817 mm (.0007153 inch)
n = 3: T = (.026 x 28.4) / (4 x 3 x 3 x 2.54) = .008075 mm (.000318 inch)It doesn't look too good for the 28.4 meters needed for the MG12s, SO far.
3) What widths of foil would work to get 4 Ohms from n parallel runs totaling 28.4 meters, for a couple of different thicknesses?
W = .026 L / 4nnT
T = .001 inch (.0254 mm):
n = 2: W = .026 x 28.4 / 4 x 2 x 2 x .0254 = 1.817 mm (.0715 inch)
n = 3: W = .026 x 28.4 / 4 x 3 x 3 x .0254 = .8075 mm (.0318 inch)T = .0005 inch (.0127 mm):
n = 2: W = .026 x 28.4 / 4 x 2 x 2 x .0127 = 3.634 mm (.143 inch)
n = 3: W = .026 x 28.4 / 4 x 3 x 3 x .0127 = 1.615 mm (.0636 inch)---------------------------
It's not looking overly-convenient for the MG12s, yet. But I guess it depends on what optimal foil width would be.
We want the current through all runs to be equal, I assume. So if we can't get what we need with n identical parallel runs of foil, we're out of luck? It looks that way, to me, at the moment. The next-best thing would probably be to change the desired resistance to something slightly different than 4 Ohms. And I guess another option would be to sacrifice some power and add a resistance in series or parallel with the whole shebang.
Cheers,
Tom
Edits: 07/02/11 07/02/11 07/02/11
That's an interesting question, actually. I'm still assuming we can't parallel foil as you can the wire. So you can't forex double just the central section, as in some of the early models. But that doesn't mean you can't run more current through say the central area of the diaphragm to produce the same effect.
Anyway, as I see it, we'll be using foil with half the mass in the tweeter, so if it's the same width we just run two sections in parallel to restore the nominal 4 ohm DCR. And if we do the same in the woofer, that is, half the mass, that's a simple parallel run as well. The foil's been made available to us, so whey doesn't someone who's doing the mod just get some? Then we can measure it and find a source of the bass foil.
OMG after all the work you did figuring that out you put a Discalimer at the end ? hahaha "Do Not Attempt Professional Driver Closed Course" I promise not to sue you if It's wrong. :)
Actually, I ran out of time and was pulled away but I suspected that there was an error, because when I calculated what the replacement of the MG12 bass/mid wire would be, with the same length and resistance, I think that I should have gotten foil dimensions that gave the same cross-sectional area as the original wire. But I didn't get anywhere even close. I figured I'd go back and fix it so I put the disclaimer in.
I have since discovered that I measured the MG12's mid/bass wire diameter incorrectly. I guess I happened to place my micrometer on one of the double-run wire pairs with a little exta glue, instead of on a single wire, and got .05 inch. I went back and measured the single wire stub that was protruding from the bottom connector and got .022 inch, which is probably .0226 inch, i.e. 23 AWG wire.
I corrected everything that referenced that wire size and then extended the post to include the calculations needed for changing an MG-12/QR's mid/bass panel into two wired-in-parallel foil panels of 8 Ohms each, per my previous post about that. It ended up having only 17.8% as much conductor mass as the original.
I'll have to give a little more thought to what effect that might have on the power requirements. The same voltage would now give only half the current. But half as much current would still accelerate the membrane up to 2.8X as much (neglecting the mass of the mylar) as the full current did with the original wire.
Cheers,
Tom
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: