After you build your ribbon, the next task is to figure out a way to drive it.  With ribbon impedances in the tenths of ohms (or less), you will not be driving them with most standard amplifiers. Your choices are:


1. Design a ribbon that has a higher impedance.

2. Design an amplifier than can efficiently drive an impedance of much less than one ohm.

3. Use a matching transformer to raise the impedance seen by the amplifier to something it can comfortably drive (like 4-8 ohms).


I already had my ribbon design, so “1” was not an option. I was not ready to take on the amplifier design, so “2’was not an opotion. I chose “3”.


The ribbon matching transformer is a special beast. Something you are not going to be able go to Madisound or Parts Express to  order. I came up with a satisfactory design after a fair amount of research (see the original thread for the blow-by-blow description: Another DIY Ribbon Thread).


I decided to build a toroid transformer based on the fact that there are lots of good sources for toroid cores, both new and surplus. I found that the best material is a high flux ferrite. You may also find powdered iron, Sendust, Molypermalloy (and probably other) types of toroids. These other types are not suitable as they can carry very limited magnetic fields before saturating. Here are some or the toroids I’ve bought and wound in trying to come up with an acceptable transformer.



Below are the calculations for the first toroid with which I had real success; the transformer I used on the prototype ribbon. It is second from left in the above photo. On the far left is the transformer  I’m using in my final, full size ribbon design.

Proof of concept ribbon toroid:

Model FT240-77 (2.4" OD, 1.4" ID, .5" thick - Type 77 ferrite)

u = 1800
mh/1000 turns = 2740
Max flux density of type 77 = .46 tesla
Here we go:

Primary Reactance = "CV" Rule of thumb 5-10 times desired impedance

Xl = 2Pi * f* L

L = XL / (2Pi * f)

L = 8 * 10 /(6.28 * 1000)
L = .0127 ~ 13 mh

Xl = desired impedance in ohms
2Pi = 6.28......
f = crossover frequency = 1000hz
L = inductance in Henry

Primary Turns = 1000 SQRT (L / toroid mh per 1000 turn)

L = Primary Inductance in mh

Primary Turns = 1000 SQRT (13 / 2740)
Primary Turns = 68 Turns

Primary impedance = Turn Ratio ^2 * Ribbon resistance
Turns Ratio = SQRT (Primary impedance / Ribbon resistance)
Turns Ratio = SQRT (8/.028) = ~16
Secondary Turns = Primary Turns / Turns Ratio
Secondary Turns = 68 / 16 ~ 5 Turns (rounded up)
A = Toroid Cross Section = 1.25cm *1.25 cm = 1.56 cm ^ 2
A = .000156m^2
N = V / (4 * F * A * B)

N = Primary Turns
V = Max RMS Primary Voltage
F = Xover Frequency
A = Toroid cross section in M^2
B = Max flux density in toroid in Teslas
4 = constant (something to do with converting peak V to rms?)

V = N * 4* F * A * B
V = 68 * 4 * 1000 * .000156 * .46 = 19.5 V rms
P = V^2 / R
P = 19.5 ^2 / 8 = 47 watts

P = Max power applied to primary without saturating core
R = Reflected impedance at primary

Primary = 68 turns (I'll probably use 16AWG wire)
Secondary = 5 turns (I'll use 14 AWG wire)
Impedance @ Primary = 8 ohms
Max power at Primary = 47 watt (more than enough)


You  will find that the power the transformer can transfer is proportional to the square of “B”, the maximum flux density (this is why you want a high flux density ferrite toroid), and “A”, the toroid cross section (this is why you want a physically large toroid). The higher you crossover frequency is, the higher the power you can transmit without saturating your core. You will find that the number of turns you use will be determined by the properties of the core and the primary inductance you need.











Designing a Matching Transformer