So-net無料ブログ作成

Revalidation of magnetic circuit. (Field Coil Fullrange Unit) [AUDIO-2]

In my past blog article, I validated the magnetic circuit of my field coil full-range unit.
In article, I used fixed value for Magnetic permeability of Pole piece and Yoke.
I used this fixed value for calculation.
To express correctly, Magnetic permeability is not fixed value.
Magnetic permeability is the value that is equivalent to gradient of B-H hysteresis curve.
So, it will be changed according to the strength of the magnetic field.
By taking into account the point, I want to refine my article with new calculation method.
About previous article, I want to remain them with correction line.

Basic specs of magnetic circuit
Pole Piece
Round bar shape with 28mm diameter, and 102mm length
Material is Structural steel bar called SS400
Yoke
Structural steel plate (SS400) with 9mm thickness.
Top and Bottom plate size are 125mm*110mm
Side plate size is 90mm*110mm
Others
2 round disks with 3mm thickness are installed so as to sandwich the Top Plate.
(Material is SS400)
With these disks, Total thickness of top plate is 15mm.
(9mm+3mm+3mm)
Gap
Hole size (diameter) of Top Plate is 31mm
Pole piece diameter is 28mm
So, Travel distance of gap is (31-28)/2=1.5mm
Width of gap is 15mm.
 Voice coil diameter is φ29.5mm

Field coil
Diameter of Polyurethane wire is 0.18mm
Length of Polyurethane wire is 4,180m
DCR of Polyurethane is 2,700 to 2,800 ohms (measured value)
Coil turns is 24,265 (Calculated value)

Other important constant
Magnetic permeability of vacuum = 4π×(E-7)=1.25664×(E-6) (H/m)
Saturation magnetic flux density of SS400 = 1.8T (18,000 Gauss)

I made excel sheet for magnetic simulation.

First thing to calculate is Maximum magnetic flux density of Gap.
When incresing field coil current, magnetic flux will follow until saturation.
For this unit, I use the material called SS400 for all magnetic circuit.
That means, first point to saturate is the place with most small size of cross-sectional area.
Check the cross-sectional area size thru magnetic circuit.
Pole Piece
(φ28/2)×(φ28/2)×π/1,000,000=0.00061575(㎡)
Yoke
120*9*2/1,000,000=0.00216(㎡) ※Magnetic flux is devided into 2 stream.
Pole piece's cross-sectional area is smaller than Yoke's.
It means Pole piece is the first place to hit magnetic saturation.
When pole piece hit magnetic saturation, then Gap's magnetic flux density reaches maximum value.

Calculate total magnetic flux when pole piece saturate.
Φ is calculate as a result of maltiplitation of cross-sectional area and Saturation magnetic flux density
Φ=0.00061575(㎡)×1.8(T)=0.00110835(Wb)

This total magnetic flux (Φ=Wb)is constant through the magnetic circuit.
Calculate Gap's maximum magnetic flux density.

Vioce coil size is Φ29.5mm, Gap width (Thickness) is 15mm
Cross-sectional area size of gap at voice coil (Svc) is
Svc= 29.5*15*π/1,000,000=0.00139(㎡)
Maximum magnetic flux density (Bgpmax) is
Bgpmax=Φ(Wb)/Svc(㎡)=0.00110835/0.00139=0.797374(T)
It's about 8,000(Gauss)

This is the maximum value of Gap's magnetic flux density that I can get with this circuit.

Next, I want to calculate field supply power.
This time I use Gap's magnetic flux density value as 0.76(T).
It's slightly smaller than maximum value. It means pole piece almost hit saturation level.

Magnetic flux is created by Magnetomotive force by the supplied power to the field coil.
This magnetic flux is closed circuit.
Consider this magnetic flux as same as current in electric circuit.
Total number of magnetic flux is same through the circuit as long as without branch circuit.

What I want to say is
Magnetic flux of Pole piece = Magnetic flux of gap = Magnetic flux of Yoke

I use this formula to calculate each part's magnetic resistance.

Magnetic resistance of gap (Rgp) is
Rgp=Lgp(Gap width)/(Magnetic Permeability of vacuum* Cross-sectional area of gap)
=1.5×0.001/(1.2566×E(-6)×0.00139016)
=858654.099[A/m]

Gap area is not metal then I can use constant Magnetic Permeability, calculation is simple.
Pole piece and yoke uses metal, then I have to consider Hysteresis curve.
I need to manage to get result with some new idea or method.

I calculated as below.

First thing to calculate is Total Magnetic flux (Φ)
Φ(Wb) =Bgp×Sgp=0.76(T) ×0.001390155(㎡)=0.001056518(Wb)

Calculate each part's magnetic fluxdensity B(T)
Pole Piece (Main Part)
Bpp(T)=Φ/Spp
    =0.001056518/0.000615752=1.715816(T)
※Bpp value is slightly smaller than SS400 material's Saturation magnetic flux density.

Pole Piece (Top end)
※This time I designed pole piece as just a round bar,
but pole piece with wider top end part (T-shape from side view) is very common.
I calculate top part of pole piece.

Bpt[T] =Φ/Spt
    =0.001056518/0.001319469
    =0.800714286[T]

Yoke frame
Byk(T)=Φ/Syk
    =0.001056518/0.00198
    =0.534(T)

Alreadey get each part's magnetic flux density,
then next to get is each part's magnetic field strength (H)

Magnetic field strength H(A/m)
Magnetic Permeability B(H/m)
Relative magnetic Permeability
will be following.

Pole piece (main)
Magnetic field strength
Bpp=1.715816(T)  ⇒ Hpp=8090.816327(A/m)
Magnetic Permeability=Bpp/Hpp=0.00021207(H/m)
Relative magnetic Permeability
=Magnetic Permeability/Magnetic Permeability of vacuum
=0.00021207/(4π×(E-7))=168.7599
Maximum Magnetic Permeability of SS400 is several thousand level.
For Pole piece, Magnetic Permeability value is very small about few hundred level.
It's very far from maximum value.
It means using saturating area, then gradiation of B-H curve's gradient value is small.

Pole piece (Top end)
Magnetic field strength
Bpt=0.800714286(T)  ⇒ Hpt=88.96825397(A/m)
Magnetic Permeability=Bpt/Hpt=0.009(H/m)
Relative magnetic Permeability=7161.9724

Yoke Frame
Magnetic field strength
Byk=0.534(T) ⇒ Hyk=59.2883058(A/m)
Magnetic Permeability=Bpt/Hpt=0.009(H/m)
Relative magnetic Permeability=7161.9724

Now I can calculate magnetic resistance (R) as below.
R=Magnetic path length/(Magnetic Permeability*Cross-sectional area)

Pole Piece (Main)
Rpp=102×0.001/(0.00021207×0.00061572)
  =781116.4319(A/m)

Pole Piece (Top end)
Rpt=0×0.001/(0.009×0.00139469)
  =0(A/m)
(There is no Top end part for this unit)

Yoke Frame
Ryk=216×0.001/(0.009×0.00198)
  =12121.21212(A/m)

Gap's magnetic resistance is
Rgp=858654.099(A/m)
Then, total magnetic resistance (Rm) will be
Rm=Rgp+Rpp+Rpt+Ryk
Rm(A/m)=781116.4319+0+12121.21212+858654.0987
    =1651891.743(A/m)

Calculate the Magnetomotive force (NI).

NI=Total magnetic flux (Φ)*Total magneticresistance
 =0.001056518(Wb)*1651891.743(A/m)
    =1745.2527(A) (Ampare・Turn)

Actual field coil current is calculated as
Actual field coil current Ifc(A)=NI/Number of coil turn
Ifc=0.07193(A)
At this time, Coil voltage Vc is 198(V), Power consumption is 14.23(W).

During unit production, I used temporary setting of Field coil current.
When field coil current is 62(mA), Field coil voltage was 210(V)
How did it operate?

Use excel sheet to simulate this condition.
As a result, when field coil current is 62(mA), Field coil voltage shoud be 170(V)
40(V) difference between actual result and simulation result.

Check the Simulator's parameter settings, and found some reasons.

Number of Field coil turn is just estimation.
Coil resistance is depend on temperature.
And biggest problem is SS400's B-H curve is not acculate,
because I took this data from somebody's website.
It might have big difference to actual material.

To get acculate result, I need to take the statistics with many times of trial.


Small modification of Pre Amplifier (Add Tone Control Circuit) [AUDIO-2]

I spent more than half a year for producing and adjusting self designed field speaker unit.
I felt result is so so Good as my own made.
But I need to evaluate this unit fairly.
I compared this unit with my main speaker system.
Result is I feel frustrated in Sound Scale and High frequency part.
My unit's high frequency sounds bit weak.

Then I decided to modify my pre amplifier to add tone control circuit.
Back panel of my pre amplifier has few pairs of receptacles for spair use.
Extra receptacles are, 1 pair for input line, 2 pairs for output line.
I decided to utilize these receptacles for adding temparary Tone control circuit when I need.

DSCN9774.JPGDSCN9773.JPG

I chose "CR Type" tone control circuit.
Drawing below is the Tone control circuit.

Tone Circuit 20140420.jpg

Criteria frequency for Bass is 400Hz, Treble is 3kHz.
Left half of circuit diagram is Treble part, Right half is Bass part.
I omit details, but basically Time constant Combination of 2 resistors R1=9.1[k ohms] , R2=1.6[k ohms] that you can see in Bass circuit and 4 capacitors makes criteria frequency.

TREBLE Circuit criteria Frequency=3,000Hz
 Incresing Capacitance C1=1/(2πf・R1)=159/(fH〔Hz〕*R1〔KΩ〕)=159/(3000*9.1)=0.005824 [μF]
 Decresing Capacitance C2=1/(2πf・R2)=159/(fH〔Hz〕*R2〔KΩ〕)=159/(3000*1.6)=0.0331 [μF]
BASS Circuit Criteria Frequency=400HZ
 Incresing Capacitance C3=1/(2πf・R1)=159/(fL〔Hz〕*R1〔KΩ〕)=159/(400*9.1)=0.0437 [μF]
 Decresing Capacitance C4=1/(2πf・R2)=159/(fL〔Hz〕*R2〔KΩ〕)=159/(400*1.6)=0.248 [μF]

CR type Tone Control circuit must be used with Low impedance output and High imepedance input.
Luckily my Original pre amplifier uses external Channel Divider, and this channel divider require the same impedance combination. So, I can insert this CR type Tone Control circuit here.
However, This tone control circuit lose gain by about 16dB, I must compensate 16dB.

UTC A20 transformer before Tone control has 2 function, divide signal to 2 channel and impedance change to 150[ohms]. A20 has 6dB of signal Loss. If skip this transformer, I can compensate 6dB back.
Still have 10dB of gain loss, As long as for home use, total gain is enough for me. I completed re-design.

Honestly say, I prepared another idea. I tried this idea, but effect is not remarkable.
What did I do? My pre-amplifier has 2 individual final stage circuit. I wanted t use 2 final stage in series connection. To realize this idea, I'd changed output impedance of one of final stage to high (Originaly 600ohms),
Then High impedance output is connected to the other final stage.
From the Gain point of view, this idea does work, but noise and sound point of view this is not so excellent.

Preamp Circuit 20140420 part1.jpg
Preamp Circuit 20140420 part2a.jpg
Preamp Circuit 20140420 part3.jpg

I tried out my field coil unit with this tone control, then felt still high frequency sound is still bit weak.
So, I decided additional modification with tone control circuit.
Change criteria frequency to middle frequency.
Criteria frequency for Treble 3kHz change to 1kHz,
Criteria frequency for Bass 400Hz change to 1kHz,

To change this criteria frequency, just need to change C1 to C4 capacitance.
For Bass control 400Hz to 800Hz , 800/400= 2, C1,C2 capacitance value need to be double.
For Treble control 3kHz to 1kHz , 1000/3000=1/3 C3,C4 capacitance value need to be 1/3.

Attach final circuit diagram.
Tone circuit2.jpg



500Hz LC Channel Divider [AUDIO-2]

I write the article about 2way channel divider for multiamp.
Circuit diagram itself was already uploaded once in my 6SN7 Pre-Amplifier article.
This channel divider can be work for ALTEC A5 or similar 2way system with 500Hz crossover frequency.
This channel divider is not the network for speaker's low impedance line.
So it's free from speaker impedance, then you can use this channel divider for any impedance type of ALTEC A5.

DSCN9629.JPGDSCN9630.JPGj
DSCN9632.JPGDSCN9634.JPG

My Pre-Amprefiler has 2 stages of 6SN7 amplification, First stage 6SN7 output is connected to UTC A25 transformer and convert impedance to 600 ohms. Volume control is 600 ohms Bridge-T constant impedance type and inserted here.
After volume control, signal goes to UTC A20 transformer to divide siganal to 2 channels and convert impedance to 150 ohms.

This Channel divider has 500Hz crossover frequency, and Attenuation characteristic is 12dB/oct.
I chose Inductors and capacitors for filter elements and realize 12dB/oct characteristic.
Load of this filter circuit is next stage 6SN7's grid resistor. I chose value of this grid resistor as 470kohms.
Input impedance is 150 ohms (A20's output impedance).
Constant value for 500Hz LC circuit is
LC=(Rin+Rout)/((2*pai*f)^2*Rout)
I chose the LC value combination with consideration of cost & size,
L&C Combination that I chose is
L=8.5mH, 
C=12uF

I designed this LC channel divider around year 2006, so I already forget about prices.
But still can find the price for same spec L & C thuru WEB.
8.5mH Inductor without iron core costs about US$167.00, Inductor is very expensive.
If I use 4mH(3.9mH) Inductor, it costs about US$100 only, very reasonable price compare to 8.5mH, and occupy lesser space. It's necessary to use 4 Inductors for this circuit, I can save US$67*4 = US$268.
Of course capacitor price is getting higher, but not so much different compare to Inductor price.
Anyway I already chose 8.5mH for Inductor, continue article with this spec.

My 500Hz channel divider has 3 different crossover freq position as 480Hz, 500Hz, 520Hz.
Crossover frequency moves 20 Hz by increasing or decreasing 1uF capacitance.
(Increase Capacitance, the frequency become lower)

Actually I made divider with 3 posistion of crossover freq, but I couldn't discrimenate clearly with 20Hz difference.Don't need to have crossover point selector switch with 20Hz difference.
It is necessary to have at least 100Hz difference to discriminate.

Diameter of 8.5mH Inductor without iron core is 5.3cm. I needed 4 pcs of Inductors for channel divider, then cannot store into pre-amplifier. So, I prepared independent box for channel divider. I select TAKACHI SL series as same design as pre-amplifier.
This channel divider box has 4 channels input and 4 channels output (L and R, High and Low).
There are 8 canon receptacles in back panel. And this box back panel still have space, I arranged 3 pair of Speaker terminals to select the combination of power amplifiers and speaker units.
For High frequency driver, 20uF capacitors are inserted for protection (When 288 is 16ohms, Cutoff freq = 497Hz).

This 8.5mH Inductors without iron core are easily effected by electro magnetic field of LCR equalizer amplifier, So I needed to separate power supply of equalizer amplifier, Prepared the individual box to store and moved to the place with certain distance from channel divider.

Attach Channel divider circuit and Peripheral.
LC channel devider.jpg

Verification of Magnetic Circuit (Field Coil Fullrange Unit) [AUDIO-2]

I refined this article, please check latest one.

The biggest feature of this self designed unit is that Unit has field coil magnetic circuit.
To be honest, I designed this magnetic circuit without detail consideration of Magnetic flux density.
I only have one time experience that I decomposed my Magnavox field coil woofer and imitate that.
I was very lucky because self designed unit can work properly, I must say this is big Gambling.

I have to verify designe of magnetic circuit.
First of all, Set constant value for calculation
Permeability of vacuum=1.25664E-06
SS400 Relative permeability =2100 (SS400 is used for Pole Piece & Yoke frame)
SS400 Magnetic saturation = 2.30[T]
SS400 Coercive force = 80.0[A/m]
For Field coil, I used 1kg of dia-0.18mm Polyurethane Enamel wire, It has 4,418m length.
Average diameter of field coil is 59mm, then number of coil turn is 23,835.
Number of coil turn(N)=23835.47588
Field coil current=0.07[A]
Ampare turn NI= 1668.483312
Next is calculation of Pole Piece's magnetic resistance.
Diameter of Ple Piece = 28.0[mm].
Cross-sectional area = 0.000615752[㎡]
Length of Pole Piece = 102[mm]
SS400 Relative permeability =2,100
Magnetic resistance of Pole Piece = Rcp
Rcp = Magnetic path length/(Permeability of vacuum×SS400 Relative permeability×Cross-sectional area)
Rcp=0.102/(1.25664E-06×2100×0.00061572)
=62771.87
Next is Yoke case Case is Box-shape, Magnetic lines of force comes from Top of Pole Piece, Pass narrow gap, and spread all directions in front(Top) plate of Yoke. After that, Flow along 2 side plates, goes through rear plate,and finally gathered again in Pole Piece. Magnetic circuit is completed.
Calculate Yoke's magnetic resistance (1 side)
Cross-sectional area of Yoke plate = 110mm×9mm=0.0099[㎡]
Magnetic path length of Yoke =(125/2)mm×2+90mm=0.215[m]
Calculate Magnetic Resistance of half of Yoke is
Ryk1 = 0.215/(1.25664E-06×2100×0.0099)
Ryk1 = 8229.512
The other side Ryk2 is same as Ryk1
Ryk2=8229.512
Combined resistance is Half of Ryk1,Ryk2
Then
Ryk=4114.76
Calculate Magnetic resistance of gap area
Diameter of Voice coil is about 29[mm]
Then Circumference of voice coil L is
L = 29.0×PAI=0.0911[m]
Thickness of Yoke's front plate is 15[mm]
Then Cross-sectional area of gap Sgp is
Sgp=0.015×0.0911=0.001366
Hole diameter of Yoke's front plate is 31.0[mm]
Then Travel path of gap Lgp is
Lgp = (31.0-28.0)/2=0.002[m]
So, Rgp will be
Rgp = 0.002/(1.25664E-06×0.0911)
  =1164611.31
Calculate Total combined magnetic resistance R is
R=Rcp+Ryk+Rgp
R=62771.87+4114.76+1164611.31
R=1231497.933
Flux of gap area φ[Wb] can be calculated by Ohm's law of magnetic circuit
φ=NI/R
NI= 1668.483312
R= 1231497.933
φ[Wb]= 0.001354841
Magnetic flux density of the gap portion B[T] is
B{T} = φ[Wb]/Sgp
B[T]=0.001354841/0.001366
B[T]=0.99140031
Convert into Gauss unit then result is
B[Gauss} = 10,000*B{T}
Then
B[Gauss]=9,914(@ field coil current=70[mA])
My Field coil power supply is designed with variable current function,
When field coil current is 60mA, then B[Gauss] goes down to 8,498
I felt Peace of mind for the time being.
And I noticed that 94.5% of Magnetic resistance comes from gap part.
If I designed front yoke plate hole bigger by 1mm, might not get enough Magnetic flux density.
I noticed one point missing in previous verification.
It's SS400's Saturation magnetic flux density.
I mentioned SS400's Saturation magnetic flux density = 2.3[T]
But I never used this item.
And searching more info thuru internet, Saturation magnetic flux density 2.3[T] should be more small value.
Material of pole piece and Yoke frame is the same SS400.
Pole piece's Cross-sectional area is 0.000615752[㎡], Yoke frame's one is 0.0099[㎡].
Yoke frame's Cross-section area is much bigger than Pole piece.
Pole piece will reach to magnetic saturation situation first.
Then Magnetic lines of force leaks to the air, instead of go inside pole piece.
First, I calculate with Saturation magnetic flux density 2.3[T].
When Pole piece reach to Magnetic saturation,
Flux value = Cross-sectional area of pole piece* Saturation magnetic flux density
Flux value = 0.000615752[㎡]×2.3[T]=0.00141623[Wb]
Suppose all flux spread from pole piece to yoke, gap's cross-sectional area is calculated as
30.0/1000×PAI×0.015=0.001413717[㎡]
Magnetic flux density is
Magnetic flux density = 0.00141623[Wb]/0.001413717[㎡]=1.001783[T] (10018[Gauss])
This means upto 10018[Gauss] , increse Magnetomotive force then Magnetic flux density will follow.
This result "looks" OK. Is it true?
Some info says SS400's Saturation magnetic flux density is lesser than 2.3[T],
Minimum info is 1.5[T], maximum info is 2.1[T].
2.3[T] is only achieved by Permendur (Fe-Co Alloy).
Tentertively use value 1.8[T] for calculation.
Gap's Magnetic flux density = 10,018[Gauss]×1.8/2.3=7,840[Gauss]
My unit may be operating around this value.
And pole piece doesn't have fange,just straight pole.
Flux in the gap wil not be even. So I want to re-design if I have next chance to design
How can this unit achieve 10,000[Gauss] gap Magnetic flux density ?
There are 2 choices to achieve.
1) Use bigger diameter pole piece
2) Use different material with bigger Saturation magnetic flux density
Now I'm very interested in this material with bigger Saturation magnetic flux density
called "Permendur".
Just change material to Permendur, then my unit's Magnetic flux density reaches to 11,039[Gauss] Permendur is the Fe-Co alloy with 2% Vanadium. Some of very expensive unit uses this Permendur. It's cost reaches more than USD$10,000 easily.
Is this so expensive?
I couldn't find the company informs Permendur's price in website.
I tried to guess Permendur price from other market info.
Pole piece has diameter in 28.0mm and length is 102mm,
So Volume of pole piece is 6.28067E-05 cubic-meters.
To easy to imagine, it's 62.8 cubic centimeters.
Specific gravity of iron is 7.86, Cobalt is 8.90
Suppose to make alloy with 50:50 weight ratio, this case ignore about vanadium.
Same weight of iron and cobalt, Volume of these materials are Iron=1: Cobalt=0.883
Specific gravity of this alloy is 7.86×2/(1+0.883)=8.35
Volume of pole piece is 62.8 [cubic centimeters].
Then Weight of Pole piece is 62.8*8.35=524.3[g]
Stereo set needs 2 pcs of pole piece,
then I need 524.3/2*2 = 524.3[g] of cobalt.
Cobalt price is US$14/Oz, 1 Oz=28.35[g]
Total cobalt price is 524.3/28.35×14 = US$258
Iron is not so expensive as cobalt, To make Alloy I cannot imagine how much does it cost.
Now I don't have any plan to make 2nd unit.
But thinking about the way to get 10,000[Gauss] without using Permendur.
Try to Change pole piece diameter to get 10,000[Gauss] for the Gap
I think large diameter voice coil has disadvantage for high frequency range.
At diameter of pole piece =34.0[mm], gap path length = 1.5[mm] ,
SS400's Saturation magnetic flux density = 1.8[T]
Then
Gap's Magnetic flux density is only achieved 8,800[Gauss] maximum.
If SS400's Saturation magnetic flux density = 2.2[T]
Then
Gap's Magnetic flux density can be achieved 10,000[Gauss] maximum.
If Pole Piece's Saturation magnetic flux density = 2.3[T] same as Permendur
Gap's Magnetic flux density can be achieved 12,000[Gauss] maximum.
But Permendule pole piece of this size costs more than US$400 even for material.
Another good idea came to me, change the thickness of Yoke's front plate
In original design, Voice coil bobbin's diameter is around 29mm
Voice coil is 44[turn]/Layer of diameter 0.18[mm] Polyurethane Enamel wire,
with 2 layers of Polyurethane Enamel wire, I get about 6 [ohms] DC resistance.
Winding width is about 7.9[mm]
Design concept is short voice coil, that means voice coil will not travel to out of magnetic gap.
Gap width is 15[mm], then Voice coil can travel upto 3.55[mm] for front and rear.
If I change diameter of Polyurethane Enamel wire to small, I can get same resistance value with narrower width of voice coil.v
Combine 2 ideas, enlarge diameter of pole piece and use thinner enamel wire, I designed again.
Voice coil enamel wire's diameter = 0.16[mm]
Pole piece Diameter = 39[mm]
Voice coil width of winding is 5.2[mm]
with DCR = 12.7[ohms], 32turns/Layer, 2layers winding
I can make Gap width narrower to 12.3[mm], That means I can get minus 2.7[mm] voice coil width compare to original.
Choice of Front yoke plate has limitation of plate thickness,
So I chose 13.6[mm] front yoke thickness.
Then Voice coil travel stroke is changed to 4.20[mm] instead of 3.55[mm]
If use SS400 with 1.8[T] Saturation magnetic flux density,
Gap's magnetic flux density will be 9,808[gauss] instead of 8,110[Gauss]
Almost achieved the target of 10,000[Gauss].
Now It's not necessary to use Permendur.
For comparison, In case of Permendur with 2.3[T] Saturation magnetic flux density
Gap's magnetic flux density reach to 13,350[Gauss]

Design & Production of original-designed full range field coil speaker unit. [AUDIO-2]

I want to upload the article about the Origianal designed speaker unit.

Designing concepts are
1) Original design without using "Speaker parts" for Voice Coil, Magnet, Cone, Frame
2) Full range unit with diameter around 9'
3) Edgeless, Dual spider damper
4) Very light cone paper (Straight cone)
5) With Field Coil magnet
6) Short voice coil & Deep Gap
7) Around 8 ohms impedance
8) Simple and Don't waste money.
9) Make and use Original wire winding equipment

Although I had no experience in disigning of speakers, I picked up some characters based on my past listening experience.
There are a lot of information when I searched through WEB, I can leverage those information.


speaker view.jpg


***Over all Assembly Drawing***
____1) Magnet wire (Polyurethane Enamel wire)
____2) Yoke Front Main Plate
____3) Pole Piece
____4) Spider Damper made by Bakelite
____5) Yoke Side Plate & Rear Plate
____6) Yoke Rear Plate
____7) Voice Coil
____8) Voice Coil Bobbin
____9) Yoke Front Sub Plate
____10) Cone Paper

Unit overall diameter is around 9 inches, Voice coil size is around 1.2 inches.
But detail design is not fixed yet. But To design detail, need to consider cost, easy production.

Items related to fixing voice coil diameter are
____1) Diameter of pole piece
____2) Gap width
____3) Jig for voice coil bobbin
____4) Thickness of voice coil bobbin material

1)About pole piece, considering cost and Relative permeability, I decided to use SS400 round bar. Close to 1.2 inch, I chose the size called diameter-28.0mm.
2) To get high Magnetic flux density, have to make gap inbetween pole piece and front yoke plate as narrow as possible. Narrow gap create difficulty to install voice coil bobbin without touching to pole piece and front yoke. I'd checked several spec exmple thuru internet, narrowest case is 0.1mm, ordinary case minimum gap is 0.2mm and maximum gap is 0.7mm. I tried to find jig with cylinder shape to make voice coil bobbin for dia-28.0mm to 29.0mm. Around dia 30mm, there are a few candidate like aluminum foil center core, small glass bottle for red chilli pepper, etc. But nothing suit. Finaly, I've found dia-28.6 stainless pipe for automobile muffler sold with cutting. I decided to use it. So, bobbin's inner diameter is 28.6. It means gap inbetween pole piece to bobbn is (28.6-28.0)/2=0.3mm
3) Choose bobbin material with certain strength to endure for many years usage.
Thickness of bobbin should be around 200um(0.2mm). Suppose I use Polyurethane Enamel wire with diameter-0.2mm 2layers, then total thickness of voice coil is around 0.6mm, And consider 0.5mm of outer gap, the hole diameter of front yoke will be 28.6+0.6*2+0.5*2=30.8mm. Decided front yoke hole diameter as 32.0mm. Balance between inner and outer gap is not even, I will attach paper with certain thickness to stainless pipe jig for adjustment.

***Design of Field Coil Bobbin***
I already decided the diameter of pole piece.
I tried to find bobbin material for field coil. Core inner diameter needs more than 28mm, and found polyvinyl chloride pipe with 31mm inner diameter. Outer diameter was 38mm. Wind coil magnet wire to this bobbin.
I arranged the bakelite endplate for this bobbin.

Choice of field coil wire (Polyurethane Enamel magnet wire) was depend on the seller's line-up.
Checked seller's web site, 1kg unit wire has many diameter range.

Resistance ratio and copper weight /kg is depend on wire's diameter. Previously, I repaired Magnavox P232, field coil 's resistance was 5,200 ohms. This time, coil wire is dia 0.18mm. current value for this Magnavox P232 is around 70mA. This is just good for Pentode current output. This is first time for me to make origial designed Speaker unit, sharing this power supply for newly designed fullrange unit, because I cannot tell whether original unit can work properly or not. As a result, I chose 1kg of diameter 0.18mm UEW wire (Polyurethane Enamel Type) to use.
Specification of this wire is 757 ohms/km, and 0.24kg/km. So, 4,160m is the length for 1kg of wire. And total resistance value is 3,155 ohms. compare to Magnavox P232 15 inch unit, total resistance value is about 60% only. Original unit is 9 inch diameter, so, I decided to use 1kg of this UEW 0.18 wire as field coil. Field current 70mA case, Power consumption is 0.07(A)*0.07(A)*3,155(ohms)=15.5W.

Consider about field bobbin thickness, width of bobbin is 64mm due to other factors.
Outer diameter of Vinyl chloride pipe is 38mm, this is starting point of winding.
Calculate the thickness of wound wire, weight ratio of copper is 8.96, volume of 1kg copper is 1,000/8.96=111.6 cubic cm without consideration of volume of viod.
Volume of pipe core is (38/2)^2*Pai*64/1000=72.6 cubic cm
Total volume is 184.2 cubic cm. SQRT(184.2/(6.4*Pai))=3.03 cm, Thickness of copper part is 3.03-3.8/2=1.13cm.
Confirmed 64mm width has no problem, continue designing with 64mm width.

Attach field coil bobbin drawing below.

ボビン.jpg
Handmade Bobbin for Field Coil
DSCN3085.JPG

***Original hand made wire winding equipment***
Polyurethane Enamel wire's length is 4,160m, if wind such a long wire to the bobbin, it's necesarry to use wire winding equipment. I'd check auction and found some items around US$50 to 60 price range.
It is not so expensive to buy, but this time I decided to make my original wire winding equipment.
Main parts are
base board, 4 pillars, shaft for bobbins, 4 ball bearings, rotating handle. It cost about US$20.

No drawing exist, all Gauging alignment.
DSCN3090.JPG

***Design of remaining of magnetic circuit and voice coil***
Pole piece diameter = 28.0mm
Voice coil diameter (inner) = 28.6mm
Front yoke hole diameter = 32.0mm
This time design remaining part of magnetic circuit and voice coil.
Concept is "short voice coil" and Deep gap.
This means voice coil will not travel until outer of gap.
Thickness of yoke front plate is 9mm, but I need deeper gap, then add 2 pcs of sub-plate to front plate.
Total thickness of front plate is 15mm.

Tentatively calculate following specs,
Divide this 15mm into 3 parts, front travel stroke=5mm, rear travel stroke=5mm, voice coil width=5mm.
Start design with this tentative value.

1 turn of voice coil wire length is 28.6*Pai= 8.985mm
if use diameter 0.2mm Polyurethane Enamel wire, I can wind 25 turns in 5mm width.
Wind wire 2layers then number of turn will be 50.
Total length of wire will be 4.49m. Resistance of this 0.2mm wire is 607.6 ohms/km
then this voice coil DC resistance is 2.73 ohms, it's too low for my design.
I need to recalculate voice coil width and wire specs.

I finalized design with 7.9mm width, use 0.18mm wire.
Wind 44 turns 2layers, length=7.91m, DC resistance= 5.99 ohms. It may be used as 8 ohms impedance.
Front and rear travel stroke =3.55mm

I want to make this unit without edge, then I needed double damper.
So voice coil bobbin need to have 2 places for 2 spider damper.
Total voice coil bobbin's width is 42mm.
voice coil bobbin design.jpg

***Production of Voice coil bobbin and refine of design***
Pole Piece is dia-28mm SS400 round rod (bar), Hole of front yoke is dia-31mm. So, Gap is 1.5mm. In 1.5mm gap, need to place 2layers voice coil and bobbin without touching.
As a Jig for bobbin and Voice coil winding, I prepared dia-28.6 stainless pipe. I chose "Shibukami" for the bobbin material.
Shibukami is Japanese Traditional Stencil Design washi-paper, originaly use for making Kimono pattern . Very thin persimmon juice painted paper, combined into 3 layers.

I chose the 110um thickness shibukami paper for bobbin. Put 1 layer of copy peper on the stainless pipe to make bobbin's diameter slightly big, and Roll shibukami-paper for bobbin on it. I cut bobbin paper exactly follow the length, no extra length to wrap, to fix the bobbin shape, use very thin 5mm width heat resistance capton tape from inner side.

To make 45 turns voice coil with dia-0.18mm Polyurethane Enamel wire, it takes about 8mm by calculation. But actually need almost 10mm width to wind 45 turns.

Marking the line for voice coil alignment, wind Polyurethane Enamel wire slowly and carefully for 1st layer 45 turns, then spray electro isolation varnish to the voice coil to firm the coil. Wait about 1 hour, then can start 2nd layer.
To wind 2nd layer nicely, 1st layer's alignment is very important. No gap, wind tight is required.

As a result, bobbin width needs 42mm, because spider damper fixing point is 20mm from center of voice coil for each front and rear.
DSCN9199.JPGDSCN9201.JPG

***Design of Yoke and Pole Piece***

Yoke Front Main Plate
トッププレート.jpg
Yoke Front Sub Plate
トッププレート(サブ).jpg
Yoke Rear Plate
リアプレート1.jpg
Yoke Side Plate
サイドプレート.jpg
Pole Piece
センターポール.jpg

***Production process of Yoke and pole piece***

Trial assemble of Yoke and Pole Piece

31mm diameter hole for yoke front plate was processed by professional fabricator, but other holes are done by myself.
Accuracy was very poor, then I needed additional adjustment for alignment.
Disc shape plate with 3mm thickness is purposed to get Deep magnetic gap. Fixed 2 pcs of this plate to Front yoke plate, Total thickness of front yoke is 15mm.

Finaly managed to align center hole to pole piece.
DSCN3543.JPGDSCN3544.JPG
DSCN3542.JPGDSCN3547.JPG

I used 9mm and 3mm thickness SS400 plate with oxide layer. I Painted with Zinc epoxy spray.
DSCN3561.JPGDSCN3562.JPG

***Design and Production of cone paper***
Please see drawing below.
cone cut.jpg

Apex angle is 120 degree.
VC diameter is 29mm, cone max diameter is 228.6mm,
Draw 2 concentric circles, inner circle diameter is 33.5mm, outer is 264mm, cut out sector with 311.8 degree.
Black dotted line is cone's outer end, fold back cone paper 8mm with this dotted line to make cone firm.
Outer circle size including folding back is 280mm.

Half circles dia-158mm and 140mm are cutting line and over wrap width to make cone shape.
Material for cone paper is sshibukami with 160um thickness.
Use circle cutter to cut, and use divider to scribe folding back line.
Glue for cone paper, I chose 3M Scotch "Super multi purpose".
Effective weight of cone paper is 5g, weight of bobbin is 2g. Total weight is 7g only.
For comparison, spider damper weight is 3.5g per piece.
Photos of cone after processing
DSCN9756.JPGDSCN9758.JPG
DSCN9759.JPGDSCN9762.JPG

***Design and production of Spider damper***
Concept of this unit is edgeless, so,there is no support in cone edge. Unit needs more than two dampers to hold cone or voice coil bobbin. I designed double damper to hold both end of voice coil bobbin.
It's not gathered type, spider type. Material is bakelite, 1mm thickness.
Outer size is 100mm diameter, inner hole has 28.6mm diameter, same as voice coil bobbin.
Shape of spider follows one of famous European old unit.
spider design.jpg

Print out the damper shape to copy paper, and attach with starch paste. Use 3.4mm drill bit to cut out shape roughly.
Detail will be finalized by file.
Damper holder is 50mm length of M4 screw without head. Place 3 points on the dia-95mm circle line.
Tentertively fixed damper alone and checked damping function. It looks to hard, cannot get enough stroke distance.
Then file damper until around 0.5mm thickness.
DSCN8835.JPGDSCN8841.JPG
DSCN8847.JPG

***Progress situation of Production 22 Jan. 2014***

1) Made 10 units of voice coils and bobbins including spare
2) 2 pcs of field coil were completed production. Length of Polyurethane Enamel wire for field coil is calculated more than 4,000m. Suppose 1 turn length is 22cm (Ave winding diameter 7cm), then total length of 1kg dia-0.18mm Polyurethane Enamel wire needs more than 20,000 turns. 2 to 3 hours per day, It took 3, 4 days to wind 1 field coil. Have to make 2 units, it was very tiresome.
Lead wire is solderd, wrap coil with isolation paper (same as cone material). This is also good for cosmetic purpose.
Total resistance of field coil is around 2.7 to 2.8k ohms as expected. Resistance value is not stable, due to temperature condition.
3) Changed current out put from 7-Pin connector to terminal block. 7-Pin connector for Magnavox field units moved to exclusive box.

Photo shows field coil with Yoke and spider
DSCN9712.JPGDSCN9711.JPG
DSCN9713.JPG.

***Remaining process***
Fix damper and Bobbin
Fix cone paper and bobbin
Fix Voice coil lead wire to voice coil Polyurethane Enamel wire
Fix terminal Block for signal and power supply

It takes about 8 months, project status is final stage now.

***Last things to do are***
1) Align the voice coil lead wire
Fix voice coil lead wire with capton tape in 10mm distance.
2) Fix cone side damper tentatively.
Front (cone side) damper must be attached before fixing cone and voice coil bobbin.
Make 2 small notches in front damper for voice coil lead wire.
3) Make holes for voice coil lead wire and flexible wire on cone paper.
4) Adhere voice coil bobbin and cone.
for the cone paper, make about 5mm wrap zone at the time of paper cutting.
Use "cemedain C" glue for fixing cone and bobbin.
5) Fix Cone & bobbin assembly to front yoke plate, tentertively. Screw front damper nut.
6) Fix rear (magnet side) damper, tentertively, screw nut.
7) Adjust damper height with screws.
8) Adjust voice coil position in the center of front yoke,
9) Threading flexible wire to the cone paper. And solder voice coil lead wire and flexible wire.
10) Fix terminal block.

All precess is completed.
DSCN9989.JPGDSCN9987.JPG
DSCN0001.JPG
DSCN0004.JPG

Check the static specifications.
DCR of voice coil is 5.6/5.9 ohms OK
DCR of Field coil is 2.7/2.7 k ohms OK (within Expected range for using 1 kg of dia-0.18mm wire)

Connect to the power supply, and measure DC voltage and current.
DC Voltage = 210V both channel
Field current = 62mA both channel

***Sound performance check (Unit alone)***
Re-connected UV-845S-A1 Amplifier from Multi-amp use to single-Amp use for fullrange unit.
Feel bit of over damp in low frquency range, but It operates perfectly, it's almost trouble free .
No smokes, no vibration sound, no voice coil touch.
But I still feel over damping in low frequency.

***Consideration about over damping in low frequency***
Although I don't know about the reason of over damping, when I touch to the cone and tried to move cone paper to stroke.
I still felt some resistance in it's action. It might be reduced.
Checked the dampers pattern, there are 3 points of bridge shape that hold inner circle(bobbin).
Each bridge has 2 branches, cut 1 branch from each bridge for reducing registance for stroke movement.
I felt much smooth movement.

***Sound performance check 2 (Unit alone)***
I'd checked unit again with music. Low frequency sound has more volume, Welbalanced overall.
Cannot play until lowest end, but it's enough for practical use. I can change sound expression with adjusting field supply current. Sound tight or loose, speed high or low.
This check is done without box and just put on the floor, I think that put this unit to flat baffle or rear pressure free box, then I can get better sound.

***Points that I noticed***
1)1.5mm magnetic gap is manageable to fix Voice coil & bobbin inbetween, even for amateur. I will try 1mm gap for next trial

2)Processing accuracy is very important. Don't be stingy. Better to ask professional to drill hole to metal parts. You can get accuracy and will not waste time.

3) Stiffness of damper is still under observation, if still need to make it softer, I will cut out damper bridge and change material.

4) 160um cone paper might be too thin, my unit is still ok, because it does'nt have edge, so far from problem.
If change thickness sound might have changed..

5)This time I designed field coil follow the specs of my Magnavox woofer. If make with thick magnet wire, then easier to wind field coil, because number of winding is lesser.

6) I didn't have any experience to make speaker unit before, but could get result. I'm happy if my article can help audio extinct species who want to design and produce own speaker unit.
This project has ended succesfully, even the person who has no experience designed, processed, purchased parts. Many many information in WEB helped me, I hope I also can help the person who is going to design and produce own hand made speaker.

If I have chance to make field coil speaker for 2nd time, I want to correct and change design for few points.
1) Make front yoke plate thicke to 15mm, remove 3mm sub plate.
2) yoke side and rear plate change to 12mm thickness.
3) Change hole position because of yoke plate thickness change.

***Installation to Rear pressure free box***
This unit has field coil magnetic circuit, it is very old fashioned style.
Aim is light and high speed sound.
Although this is first unit for my experience, I could follow my concepts through designing and production, I satisfied with sound.
Through sound testing, I put bare unit on the floor without box.
I cannot deny the low frequency sound was offset by reflection of the sound from behind the unit.
To prevent this offset, I want to install this unit to either box or buffle board.

There are many reports that Old field unit was installed to flat baffle board,
Flat baffle board needs certain size to prevent back sound reflection.
This is very inconvenience for me, because my main speaker is Altec A5 "type", cannot exist together in my room.
Another idea came to me was free back pressure type, Although position is slightly too high, it still can put on top of Altec 805B. Suddenly image of the barrel came to my mind. There are several choices of Barrel for coffee, wine....
My unit's outer diameter is around 250mm, then I chose barrel with the size of 395mm bottom circle board, height is 490mm.
Price is about US$65 for oilstain finished type.

Cut the 250mm diameter circle to bottom board to fix unit. but there is no frame with the unit because of frameless.
Then need to fix magnet yoke to the barrel instead of front frame, magnet yoke is sandwiched between wooden boards and wooden board is fixed to the barrel.
DSCN9060.JPGDSCN9063.JPG
This is the wooden frame for the unit, I fixed Speaker terminals (for field coil and voice) to
this wooden frame.
DSCN9057.JPGDSCN9064.JPG
Made 250mm diameter hole to Barrel bottom board and fixed wooden base for wooden frame.
DSCN9072.JPGDSCN9076.JPG
Field coil Speaker unit installed to the Barrel
DSCN9081.JPGDSCN9083.JPG
taru13.jpgDSCN9085.JPG
Set up on top of Altec A5 "Type" box (I made spcial 370mm height speaker stand)
And I made special barrel holder, too.

***Overall Sound Impression with Barrel box***
Basically sound impression has not changed compare to bare unit. Very light sound, high speed.
Volume of low frequency sound became rich. I wanted to have high frequency sound more.

***Attach sub cone***
To get high frequency sound, I attach sub-cone to the unit.
Sub-cone diameter is 80mm.
DSCN0497.JPGDSCN0498.JPG
DSCN0499.JPGDSCN0500.JPG

***Add Tone-control circuit to pre-amplifier***
I'll right araticle separately.

***End of the article***
After attach sub-cone and add tone-control circuit, sound is not wide range but very well balanced. Good for 50's-70's music, especially for vocals, Jazz combo.

Design and production of Power supply for Field coil units [AUDIO-2]

Previously I wrote 2 articles about Design and production of Field Coil Unit's exciter (Power Supply) in Japanese.
This time I merge 2 articles and summerize into one in English.

Recently I purchased one pair of Magnavox P232 field coil woofer, I designed power supply unit for them.

Photo of 15" Magnavox field unit (Woofer)
$(KGrHqVHJFQFEMT0KB+2BRf)z3(l5w~~60_1.jpg

No Model number on the frame (Maybe P232 Year1947)
Voice coil impedance = 8 ohms
Field coil resistance = 5,200 ohms
For instrument use

I'd found another info on ebay, that says Field coil needs 360V, 67mA. This info means unit needs 24 watts power. This time I design power supply with 25 watts maximum power. Lower limit of this supply is targeted about 10 watts.
There is a controversy that which type of power supply is good for field coil speaker, constant voltage or constant current. I sided for constant current, because I heard voice coil signal in magnetic fields may affect field coil current. To prevent this affection, constant current power supply with high impedance output has advantage for this issue.

How to get constant current power output? Constant current output needs infinite value of output impedance. Realistically, use constant current diode is the easiest way, but this case I need minimum 70mA current output. Parallel connection of constant current diode is not good idea. And power supply with famous TUNGAR BULB is also not good idea for this unit, because of it's low voltage and high current.Then I decided to use pseudo constant current circuit using pentode's high output impedance.
In the Ep-Ip graph, You can see no much difference in output current, even plate voltage is moved high and low.
This means pentode has high output impedance characteristic.
I decided to use 6L6G that I have.

Reason is
1) 70mA output current is required,and this value is still in the range of 6L6G
2) 6L6G Ep-Ip characteristic shows its very high output impedance.
3) Shoulder voltage is very low (around 60V), I can use very low plate voltage if I need.
6L6G 5 connection.jpg

My idea is "Get constant current characteristic by pentode connection".
But I check triode-connection characteristic first.
Ep-Ip line when Ec1=0V, Output current 70mA point is plate voltage 127V.(You can see the red line's cross point)

6L6G 3connection cross127V.jpg

   This means in pentode connection, if supply 127V for plate, 127V for 2nd grid, 0V for 1st grid then output current is 70mA. Actually 6L6G pentode connection Ep-Ip characteristic graph shows same result. Around Eb=127V output current slope with imaginary Ec2=127V line is very horizontal.It shows very high output impedance.And far enough from sholder area (Eb=60V), I set 6L6G's reference operation point here, for 25W output.

6L6G 5 connection 127V.jpg

This is the basic circuit of pseudo connstant current circuit using pentode. By adjusting the Ec2's voltage, Output voltage and current will follow.

6L6G control.jpg

To supply 25W to field coil, coil voltage needs 364V. So, DC power supply voltage needs 127+364=491V.

Suppose you can adjust field coil power, at the time of 10 watts (minimum), field coil current will be 43.9mA. Current difference between max-min is 70.0mA - 43.9mA = 26.1mA
Use 2pcs of 6L6G, Total current difference is 52.2mA.
Filter circuit's DCR is 266 ohm, at the time of minimum power Ebb will be
Ebb = 491+52.2*266/1000 = 491V + 13.9V = 504.9V
Voltage drop of Field coil is 43.9mA*5200 ohms = 228.3V , then 6L6's Plate voltage is
Ep = 504.9-228.3 = 276.6V
Please See following graph, To satisfy the condition Ep = 276.6V, Ip = 43.9mA
Ec2 must be around 87V.

6L6G 5 connection 127-87V.jpg

Design DC power supply circuit with DC voltage of 491V, Control screen grid voltage from 127V to 87V, then you can control the output power from 25 Watt to 10 Watt.

Think about the circuit that you can control screen grid's voltage 127V to 87V.

Power circuit supplies 491V, and Output circuit uses 491V.
6L6G's screen grid is connected to Ebb (491V) through fixed resistor called R1.
6L6G's screen grid is also connected to ground side, through VR and R2.

If I2 is the current which flows VR and R2, then Current I1 which flows R1 will be I1=I2+Ic2(Screen grid current).
I2*(VR+R2)=Ec2, Calculate the combination with the formula Ec2+I1*R1=491V.

Screen grid current is related with plate current very much.
Ic2:Ip ratio is usually around 0.1 to 0.15, through previous experience.
Tentertively set value of 0.1 for Ic2:Ip ratio as 0.1, continue calculation.

Calculate R1, R2, VR with consideration of VR's power consumption (Heat).
When Ip=70mA, Ic2:Ip=0.1, then Ic2=0.7mA goes to screen grid.

I must find the best combination of R1, R2, VR value. Considering the worst case of Ic2*Ip ratio is 0 or 0.2.

In case Ic2:Ip=0, then Ic2=Ic1, at the time of 25 watt
(VR+R2):R1=127:(491-127)=127:364 => 1:3
In case Ic2:Ip=0, then Ic2=Ic1, at the time of 10 watt
(VR+R2):R1=87:(491-87)=87:404 => 1:4
Ic2 should not exceed 20mA becase of VR's power consumption point of view.
So,R1=16k ohms, R2=3.9k ohms, VR max= 5.2k ohms. is roughly calculated.
You must be very careful to choose value of R1, R2 with considering VR's power consumption.
Otherwise VR value exceed 20k ohm easily. Tokyo Cosmos VR only has max 20k ohms for high power consumption (2.5W) wound type.

VR with more than 20k ohms value only has small power consumption rate.(0.6W or 0.4W)
Most important thing about choosing R1, R2, VR is find out the resistance combination which can achieve necessary power within Wire wound VR Range (<20k ohms).

Make excel sheet to calculate the combination of R1, R2, VR, checking with Max, Min power, power consumption etc.
Result is
R1=16k ohms
R2=3k ohms
VR=20k ohms
From VR 's power consumption point of view, Ic2:Ip ratio exceed 0.2 is the worst case, value above still has bit of margin.

I never mentioned about first grid voltage (Ec1). If use current feedback, then output impedance will be higher.
How to get current feedback? It's very simple just add resistor to 6L6G's cathode without parallel connected capacitor.

Calculate 6L6G's new operation condition with cathode resistor.
Put 30 ohms resistor to cathode, 6L6G's Cathode current is about 1.1 times of coil current. At the time of maximum power, coil current is 70mA, It means 77mA cathode current. 77mA current to 30 ohms resistor create 2.3V voltage drop. That means I get -2.3V self bias.
So, total DC supply voltage at maximum power is 491V+2.3V=493.3V

How to supply 493.3V
Now I know I need to get very high DC voltage (493V) at the time of maximum power.
Current is also required more than 150mA.
It's necessary the transformer with 450V output and more than 150 mA current capacity.
Fortunately, I have 450V output transformer with 170mA current capacity.
But consider voltage drop at rectifier and filter circuit, I have to use low loss rectifier device.
I wanted to avoid to use diode for high voltage output, I chose 5AR4 rectifier tube because of it's low loss and big current capacity character, and made trial circuit.

Drawing below shows trial power supply circuit.

field coil power supply.jpg

Filter is series connected double choke inductor.
As a result, Output DC voltage at maximum power is about 483V. It is not enough by 10 volts.
If I really want to get 493V, then still have way to do like parallel choke etc.
But parallel choke is not the way to get high output impedance.
This 10V can make difference in power by 0.7 watts (10V*70mA), 24.3 watts is still enough to me, so I decided to continue with 5AR4 rectifire tube. Because I wanted to make this power supply with my own parts as much as possible.
DC heater circuit for 6L6G is designed using schottky barrier diode bridge,

I'd checked the my own parts list.
List
1) Power transformer with 450V outut terminal (170mA)
2) Choke coil (170mA) 2pcs
3) Oil Capacitor (10uF) 2pcs
4) 6L6G (Westinghouse) 2pcs
5) 5AR4 (Matsushita)
6) Fuse, Fuse case
7) 3P AC inlet, AC cables
8) Speaker terminal for voice coil (Army) 2pcs
9) Octal Socket 3pcs
10) Power switch
11) Schottky barrier diode bridge
12) Resistors

Parts list need to buy
1) Ampare meter (Maximum range 100mA) 2pcs
2) Voltage meter (Maximum Range 500V)
3) Meter select switch
4) resistors (cement etc)
5) UT-7P Socket 2pcs
6) Capacitors for filter circuit 20uF*1, 6,800uF*1
7) Oil Capacitors hook
8) chassis
9) Lag Terminals

Chassis is most costly thing, If buy nice looking chassis from Big name brand like TAKACHI, It may cost about USD200 or more. I bought Junk-mini-compo tuner chassis from 2nd hand shop only for USD10.
Make 250mm*260mm rectangle hole on top board(cover), and put 1.2mm thickness Aluminium board.
It looks like a power amplifier, because 3 tubes, 1 transformer, 2 choke coil and 2 big oil capacitors and 4 meters on the top board.
Chassis is very small because originaly just for mini-compo tuner, parts layout was very congested.
Distance btwn Rectifier tube and capacitor is too close, then I put hand cut alminium screen inbetween.
Top alminium board is lay on top of chassis with 3mm gap for ventilation
DSCN0372.JPGDSCN0378.JPG
DSCN0388.JPGDSCN0402.JPG
DSCN0397.JPGDSCN0393.JPG
Too dangerous to test this power supply with actual Magnavox speakers which hasn't used for many years.
Power supply unit was connected to 2ch dummy resistor circuit, and check specs like operational range (Max, Min).
As a result I confirmed this power supply achieved original specs (10-25W power range).
No design mistakes and no limitation of operational range were found.DSCN1640.JPG

Before fixing into Altec 825 "Type" box, check sound as bare unit. I'd found one unit has problem.
This unit was the one I'd repaird the field coil connection, voice coil may touch to poll-piece,I can hear rusting noise sound when musical-sound reached to certain loudness.
I loosened 4 bolts next to dumper which is used for fixing Yoke, imput sound to half free voice coil and find the point that voice coil doesn't touch to Poll-piece, and fixed 4 bolts tight.
Main purpose is exchange Altec 515B unit into Magnavox P232 field coil unit.
My system has 515B unit installed in a copied 825 box.
Fortunately, hole pitch of unit's front frame is exactly same between 515B and P232.
Not necessary to make hole again for fixing new unit.
But depth of unit is different, P232 is deeper than 515B, Back board of 825 copied box interfare with P232.
But luckily center part of this back board has speaker terminal unit, just remove this unit, Then back board was successfully fixed without any interference.
If I used original 825 Box, I may not have this happy result.
There are some information about 15" field coil Magnavox P232, most of case unit was mounted to Rear pressure free box or flat baffle board.
But I never found this unit with front-load horn bass refrex box.

DSCN1651.JPGDSCN1649.JPG

DSCN1657.JPGDSCN1654.JPG


I started trial listening with this condition
Music source: LP(Jazz)
Record player: Garrard 301 Ivory (Oil type)
Tone arm: SME-3012R
Cartridges: Ortofon SPU-Mono G Mk2, SPU85 Anniversary + MC transformer
Equalizer Amp: Modified Yoshiba LCR (Separated power supply)
Pre amplifier: Self made 6SN7 with 4channel out put, 600ohms bridge-T attenuator control
Power amplifier for Low Frequency: Yoshiba UV845 A1S Mono-block
Power amplifier for High Frequency: Self made 2A3 A1S Stereo
Driver: Altec 288D
Horn: Altec 805B
Woofer: Magnavox P232 15" field type
Field power supply: 6L6G pentode connection high impedance constant current output
Enclosure: Japan domestic made copied Altec 825B (North American Pine Plywood)

After just installation, without any adjustment, I already felt very light and fast sound from units. every thing is dancing without any constraint, restriction, suppression...
Potential capacity is very high, cross over point at 500Hz is also very natural, I felt nothing is necessary for improvement.
People says, field coil unit shows different character when you increse or decrese current.
P232 might have ability to play until higher frequency because it has smaller voice coil compare to 515B. So, might have different cross-over point in higher frequency area. This old unit is installed to unexpected front horn with bass reflex box, If I remove upper part of back panel to make back pressure free, then may show different performance. And have to wait few months for re-aging, because these units were not in use for many many years. I just enjoy music with these units without trial and error.

Repair of 15" MAGNAVOX P232 field coil woofer [AUDIO-2]

In middle of 2013, I got a pair of 15" Magnavox field type woofer units.
No holes, no dents no signs of repair in cone and edge, It looks perfect in a sight.
But unfortunately, one of them has a problem in field coil wire. It lost contact and shows 11.9M ohms resistance.

I write article about field coil repair.

DSCN9226.JPGDSCN9231.JPG
DSCN0596.JPGDSCN0580.JPG
DSCN9207.JPG

This time I purchased a pair of Magnavox P232 15". There are 4 frames in rear basket. I expected spider damper a bit, but It has usual gathered type damper.

There are many Stamp marks on the frame.
I can recognize 10526, 5200OHMS, 143, 31426?, 232 739, 1D2880.
So, manufacturer code is 232 (Magnavox)
Production year is year of 7 = 1947 (Maybe), 39th week of the year

This unit's field coil has specs of 5200 ohms resistance, actual measurement of mormal unit shows 5300 ohms. Half broken unit shows around 11.9M ohms, but not stable if I touch wire at end of winding. It's necessary to dismantle field coil and pull out copper wire.

DSCN9545.JPG
I dismantled Field coil magnet of the unit.

First, I dismantled black color yoke frame.
You can see 4 pcs of Hex Bolt heads next to the damper.
Record the mounting position and orientation before remove 4 bolts.

DSCN9224.JPGDSCN0597.JPG

If disassemble yoke back plate, then Pole piece come out together with yoke.
Then, coil was still staying to the yoke front plate, but it was free and not hold. DSCN9539.JPGDSCN9541.JPG

Detach coil, then you can see the small plate that is covering voice coil.
Take this small plate out, and expose voice coil.

Starting repair of contactless wire. It was very beautiful unit looks flawless.
I had to cut isolation paper which covers field coil.

DSCN9550.JPGDSCN9552.JPG

I'd checked where the broken point is, and found most inside cable lose contact at the point of coil end.
There was no problem to through away the broken part of cable and shorten it. because this cable has several thousand turns length, 10 to 20 cm has really no effection to the field coil specs.
So, loosen winding for few turns, pull out the wire, cut, and solder to external cable.

(Broken point was at the center top of this photo, Black color clip is oposite side. This side has no problem)
After resolder, Resistor value is changed back to 5.32k ohms as almost same as the other unit. DSCN9559.JPG

Fix the isolation tape and papers back to the original position, repair of field coil was completed.
But unfortunately, outer isolation paper couldn't be fixed nicely as original.
Assemble coil to the yoke again.

DSCN0588.JPGDSCN0590.JPG

I noticed later on, there was a bit of "voice coil touch", I needed to re-position the voice coil center one more time.

VT-25 Class A1 Single-ended Power Amplifier [AUDIO-2]

I introduce my first amplifier that I made in year 1999..
This amplifier is VT-25 Single-ended Class A1 stereo type.
Recently, I use ALTEC A5 "type" speaker driven by multi-amplifiers.
One is UV-845 A1 Single ended, another one is 2A3 A1 Single-ended.
So, this VT-25 amplifier has not been used for few years already.
I decided to lend this amplifier to one of my friend for few years,
Then, now checking this amplifier through N500C Network.
VT25-A1S.jpg
I attach the circuit diagram below.
VT25 amp circuit.jpg

About rectifier tube, you can see 2 types of different tube name, one is 5U4G(direct heating) another one is 5AR4(separated heating), but tube in my photo is 5Y4. Originaly I designed with 5AR4.
For the durability of power tube, I think better to use rectifier with isolated heater, but this time I wanted to use ST shape tube like 5Y4, then I used it after changing transformer tap to 360V.

This time I changed back transformer tap to 400V, and changed rectifier tube to 5U4G.

Charactors of this amplifier are
1) equip input transformer using High impedance tap (10k ohm), not low impedance (600 ohm)
2) Triode connected Pentode drives power tube(6V6G) through interstage transformer.
3) Interstage transformer (TANGO NC9) has limitation for DC cuurent (6V6's plate current), then separete DC load and AC load.
Musical signal goes to transformer and DC is cut off by capacitor inserted to cathode side. DC goes to Resistor which is connected to B+ DC power supply.
4) Output transformer is TAMURA F2005 (Primary 7k ohm) , VT25 needs more high impedance load, then I connected 16 ohm speaker to 8 ohm terminal. Then from VT25, load will be considered as 14k ohm.
5)VT25 has 7.3V filament, I supply heating power in AC because hum noise level was ok for me.
In case my friend cannot take hum noise, then I will make small modification for heater circuit.
May be equip external rectifier circuit.

I used to use Electro Harmonics GT tube for 1st stage 6V6, I found G-tube with black smoked glass, then I changed it. I forgot about the maker of this G-tube 6V6. I attach photo.

6V6G.jpg
VT25 is GE made.
VT25GE.jpg
Even I used very small chassis, equip interstage transformer, choke coil for rectifier circuit,
so, inside of chassis is very very crowded.

VT25S内部1.JPGVT25S内部2.JPG
VT25S内部3.JPGVT25S内部4.JPG
VT25S内部5.JPG
Now using BRIMER 5U4G, this tube used to be used in 2A3 amplifier, once lost contact and retired.
But looks still have enough getter, I tried to repair this tube. Use soldering iron, add some solder to the pins, then tube's function was back to normal. I don't know how long does it work properly, then still under observation.

5U4G BRIMAR.jpg

Output power maybe slightly over 1 watt only. But I can enjoy music and also the glow of the VT-25's thorium-tangsten filament.

6SN7 2 stages Amplification-2outputs Pre-Amplifier (External LC Channel Divider) [AUDIO-2]

I introduce my own design Pre-Amplifier.
Show full-circuit drawing below
Preamp.jpg
This Pre-Amplifier has following charactors and functions
1) 3 input circuit, 1 for RCA, 2 for Cannon 3P connectors
a) RCA high impedance input for Equalizer amp with Peerless 4651 transfomer (10k: 60k),
it has gain of 7.8dB
b) Cannon low impedance (600 ohm) input for CD player without Attenuation
c) Cannon low impedance (600 ohm) input for AUX with Attenuation (15.7dB)
2) Low impedance input (600 ohm) 1)-b), 1)-c), connected to input transformer (Triad 66B-139).
66B-139 is (600: 50k) , has 19.2dB gain.
3) CD player is connected to Triad A-11J transformer (60k:600) before Pre-amp input.Through this transformer, change impedance into 600 ohms, and attenuate 20dB.
4) 1st stage tube is parallel connected 6SN7. Output transformer of 1st stage is UTC A-25. A-25 can be used with 1st stage output DC(max 8mA), and change signal impedance to 600 ohm.
5) After A-25, Signal line is changed into 600 ohm impedance, then insert Bridge-T 600 ohm Main Attenuator here.
6) Main attenuator output is connected to UTC A-20 transformer, get 2 individual output for high and low frequency and change impedance into 150 ohm (6dB attenuation).
7) Once lead high and low frequency signal to channel divider output connectors on back panel
8) External channel divider is LC circuit with 150 ohm input impedance. LC circuit divides signals at 500Hz with 12dB slope. Output Load impedance is 470k ohm. connect back to pre-amplifier.
9) 2nd stage amplification has Low/High individual circuits. Both has parallel connected 6SN7.Output transformer for high frequency is UTC A-25, for low frequency UTC S-11(Max DC=11mA). Both Transformer changes signal impedance into 600 ohm.
10) High frequency output has another Bridge-T attenuator to make balance with low frequency.
(Max attenuation is 16dB)
11) Power transformer is Hashimoto PT-180 fixed in external box.
12) Use 5AR4 rectifier tube and Hashimoto Choke C30-80 to get B+ DC high voltage. Hashimoto choke has 2 inductors inside, I connected 2 inductors in series and make double pai capacitor input filter.
13) Using total 6 pcs of 6SN7, supply heater power in DC. I used SBD bridge diode for rectifier.
14) BOX is TAKACHI SL series,

It's very complecated pre-amplifier, but circuit itself is very old fashioned and simple. And purposely change signal impedance into 600 ohm, where I need to make signal line longer, insert attenuator, etc.

   Pre-amplifier front view      Pre-amplifier rear view 
Pre front view.JPGPre rear view.JPG
 Pre-amplifier internal wiring       Triad 66B-139
Pre内部配線.jpgTriad 66B-139.jpg
       UTC S11         Output 6SN7 and A25,S11 Transformer
UTC S11.JPG出力6SN7とS-11&A25.JPG
      Peerless 4651               UTC A-25
ピアレス4651.JPGUTC A25(OLD)V.jpg


Original design 2A3 Class A1 Single-ended Power Amplifier (All stages connected by transformers) for high frequency [AUDIO-2]

Let me introduce my own design power amplifier
(I made this in year 2003, modified in year 2011)

I attach whole circuit drawing below.
2A3 Amp.jpg

Designing concepts of this amplifier are
1) All stages are connected through transformer.
Equip the power tube for 1st stage and drive 2nd (Final) stage through interstage transformer.
2) Use "AC-2.5V" for heater circuit of 2A3.
AC heating sound nicer and no much hum problem because of low heating voltage.
3) Don't use any Electrolytic capacitors. (As result, I use only one for last part of B+ power supply circuit to reduce hum noise.
4) Operation of 2A3 is following the RCA data sheet (Not 100%)
*Voltage between Plate-Cathode = 250V
*Voltage between Cathode-Grid = -41V
*Load impedance for 2A3 = 6k ohm (Typical use is 2.5k ohm)
Not necessary to get high power for me,
high load impedance has Advantage from the viewpoint of distortion

I chose Tamura PC3003 power transformer. This transformer has 2 circuit of 2.5V output, this is good for 2A3 single "stereo" amplifier. Using 320V high volatage output, Rectifier is 5U4G, 2A3 plate receive DC290V.
Filter type is "Capacitor input pai type" , first capacitor right after 5U4G rectifier is only 5uF from durability point of view.
I used Tamura A4004 choke coil. Input transformer is Tamura TKS50, usually using 600 ohm input circuit only.
First stage tube is triode connected 6F6, interstage transformer is Tamura B5003.

I attach photo of this amplifier.
2A3Amp全景.JPG6F6&2A3.JPG
        Whole view             2A3 and 6F6

5U4G.JPG2A3Amp内部配線全体.JPG
     5U4G(Right end)            internal wiring

2A3Amp内部配線1.JPG2A3ノイズ対策1.JPG
   Mainly use film capacitors   Wiring around transformer

2A3ノイズ対策3.JPG2A3ノイズ対策2.JPG
       Power supply (1)          Power supply (2)

Year 2011, I made modification to reduce hum noise level.
Tamura Choke coil A4004 has 2 different choice of wiring, one is series connection to get high inductance(10H-200mA), another one is parallel connection to get high current output(2.5H-400mA).
Originally I used this choke coil in series way (10H-200mA). I wanted add capaitor to center point to make double pai type. But this trial doesn't work as my idea. This idea based on my experience of Hashimoto choke coil used in my preamplifier and successfully done.
You can see copper shield around power supply and Electrolytic capacitors.

Output power hasn't reach 3.5W, but do not feel any power problem to drive "Altec 288D Driver"