I've just starting looking and researching ribbon speakers and am interested in making one and experimenting with them as well as Magnepan style speakers.
I've read a few threads talking about transformers to run a ribbon speakers, but can't really find any definite answer or guide on how to build them.
Looking at this website... http://hyperphysics.phy-astr.gsu.edu/hbase/magnetic/refload.html I can see the equation for reflected impedance (resistance) at the bottom which says.... R x (number primary turns/secondary turns) squared. So a 20:1 transformer with a 0.01 ohm ribbon gives a reflected impedance of 4 ohms. Am I right so far?
If so, a transformer would need to be made with a 20:1 winding. So you could do 20 on the primary and 1 on the secondary or 40 and 2 or 60 and 3 or 80 and 4 etc. The primary winding would be a very low impedance I would imagine, so does the high pass capacitor need to be calculated on the resistance of the primary winding or the reflected impedance that will be 'seen' by the amp.
Thanks in advance for any help.
I've read a few threads talking about transformers to run a ribbon speakers, but can't really find any definite answer or guide on how to build them.
Looking at this website... http://hyperphysics.phy-astr.gsu.edu/hbase/magnetic/refload.html I can see the equation for reflected impedance (resistance) at the bottom which says.... R x (number primary turns/secondary turns) squared. So a 20:1 transformer with a 0.01 ohm ribbon gives a reflected impedance of 4 ohms. Am I right so far?
If so, a transformer would need to be made with a 20:1 winding. So you could do 20 on the primary and 1 on the secondary or 40 and 2 or 60 and 3 or 80 and 4 etc. The primary winding would be a very low impedance I would imagine, so does the high pass capacitor need to be calculated on the resistance of the primary winding or the reflected impedance that will be 'seen' by the amp.
Thanks in advance for any help.
0.01 ohm... It will be difficult to make the winding resistances much smaller than the loudspeaker resistance.
If you don't, the input impedance increases by the primary winding resistance plus turns ratio squared times the secondary winding resistance. Some of the input power will then only heat up the transformer.
When you have a capacitor between amplifier and primary winding, you could dimension it such that it forms a second-order Butterworth or Linkwitz-Riley high-pass with the magnetizing inductance and the reflected load and winding resistance. The thing is that the magnetizing inductance is inaccurate and quite non-linear unless you use an air gap in the transformer core.
FYI, I have never designed anything like this, so if anyone tunes in who has, please take their comments more seriously than mine.
If you don't, the input impedance increases by the primary winding resistance plus turns ratio squared times the secondary winding resistance. Some of the input power will then only heat up the transformer.
When you have a capacitor between amplifier and primary winding, you could dimension it such that it forms a second-order Butterworth or Linkwitz-Riley high-pass with the magnetizing inductance and the reflected load and winding resistance. The thing is that the magnetizing inductance is inaccurate and quite non-linear unless you use an air gap in the transformer core.
FYI, I have never designed anything like this, so if anyone tunes in who has, please take their comments more seriously than mine.
Last edited:
Start with the Magnepan style speaker first, since the ribbon with transformer is unlikely to go well.
Thanks ray, I will start with the Magnepan speaker, but will also be playing about with a ribbon too.
This is just so I can understand the principal of the transformer for a ribbon. Please can someone tell me if this makes sense...
So, if a transformer was made using 0.5mm diameter wire around a transformer core which was 100mm long per turn and there were 40 turns, I would have 4m of wire, which is 0.34ohms on the primary. If there were 2 turns on the secondary (200mm of 0.5mm diameter wire would give 0.017ohms) and we have a 0.01ohm ribbon, giving a total of 0.027ohms of resistance on the secondary. With this being a 20:1 transformer, we add the primary resistance of 0.34ohms to the turns ration squared x secondary resistance (0.027). So 0.34 + (20 squared x 0.027) = 11.14 ohms.
So would the amp 'see' 11.14 ohms when driving the ribbon speaker? What would happen if the ribbon broke and open circuited the secondary? On a power transformer, the current on the primary increases when there's a load on the secondary, so I assume it's the same here.
I assume 11.14ohms is the value that will be used in determining the values of components for the high pass filter for the ribbon speaker?
Thanks again for any help
This is just so I can understand the principal of the transformer for a ribbon. Please can someone tell me if this makes sense...
So, if a transformer was made using 0.5mm diameter wire around a transformer core which was 100mm long per turn and there were 40 turns, I would have 4m of wire, which is 0.34ohms on the primary. If there were 2 turns on the secondary (200mm of 0.5mm diameter wire would give 0.017ohms) and we have a 0.01ohm ribbon, giving a total of 0.027ohms of resistance on the secondary. With this being a 20:1 transformer, we add the primary resistance of 0.34ohms to the turns ration squared x secondary resistance (0.027). So 0.34 + (20 squared x 0.027) = 11.14 ohms.
So would the amp 'see' 11.14 ohms when driving the ribbon speaker? What would happen if the ribbon broke and open circuited the secondary? On a power transformer, the current on the primary increases when there's a load on the secondary, so I assume it's the same here.
I assume 11.14ohms is the value that will be used in determining the values of components for the high pass filter for the ribbon speaker?
Thanks again for any help
You might want to look at the Wolpert audio transformer design manual for guidance. Somebody in another thread linked to it, I've attached it below. I found it quite useful as I'm thinking of trying to hack up an OPT as a lark.
One thing, secondary current relates to primary current as follows (please forgive my formula typing skills, these are all in Wolpert)
I2 = (I1 * E1) // E2
With secondary voltage relating to primary voltage per your specified impedance ratio of 400:1
E2 =sqrt (E1^2 * Z2 // Z1)
Long story short the current on the secondary will be way way higher than that on the primary, and will depend both on the voltage and current you specify on that primary. As current specifications determine the appropriate gauge of wire, there's no way you can use the same gauge of wire on both primary and secondary.
One thing, secondary current relates to primary current as follows (please forgive my formula typing skills, these are all in Wolpert)
I2 = (I1 * E1) // E2
With secondary voltage relating to primary voltage per your specified impedance ratio of 400:1
E2 =sqrt (E1^2 * Z2 // Z1)
Long story short the current on the secondary will be way way higher than that on the primary, and will depend both on the voltage and current you specify on that primary. As current specifications determine the appropriate gauge of wire, there's no way you can use the same gauge of wire on both primary and secondary.
ranshdow has a point about the wire thicknesses.
Your calculation is correct. The 11.14 ohm would be in parallel with the inductance of the primary winding of the transformer.
For primary, u need to parallel a few enamel wire to create the low DCR. For my ribbon tweeter transformer I uses Litz wire which are individually enameled which made winding much easier to do.
Thanks ghemml. In the images, how many turns of secondary are there as it almost looks as if it simply passes through the centre of the primary winding. It looks the same with these Pioneer ribbon tweeters...
I've wondered how Jenson toroid chokes would work as primaries. Easy enough to put 20 windings on top of the existing windings, figure the primary windings then adjust the secondary for what you want.
Normally I wind 4 to 5 turns, Use Litz wire, they are way easier to wind than trying to wind 5 wires at the same time.
I used to buy them from Korean 80% Nickel Permalloy EI35. I think they still selling in ebay
https://www.ebay.com/sch/sellarzaudiotechnology/m.html?_nkw=&_armrs=1&_ipg=&_from=
Would something like this work? The amp connected to an air core inductor (obviously would need to know the amount of turns), then the secondary wrapped around the primary?
Yeah, I just realised, it would be acting as a current transformer in that kind of configuaration wouldn't it.
I think the orientation of the coils with respect to each other is wrong, so they will hardly couple. I also think you need a core to get good coupling.
I've just been reading another thread...
https://www.diyaudio.com/community/...transformer-for-ribbon-tweeter.184034/page-11
...and someone says they have had good results using Epcos TDK ETD 54 x 28 x 19 B66395GX197...
https://uk.farnell.com/epcos/b66395g0000x197/ferrite-core-etd-n97/dp/1422751?ost=b66395gx197
They are pretty cheap, and you can also get a bobbin for them to wind the wire around. This maybe a good start for experimenting.
https://www.diyaudio.com/community/...transformer-for-ribbon-tweeter.184034/page-11
...and someone says they have had good results using Epcos TDK ETD 54 x 28 x 19 B66395GX197...
https://uk.farnell.com/epcos/b66395g0000x197/ferrite-core-etd-n97/dp/1422751?ost=b66395gx197
They are pretty cheap, and you can also get a bobbin for them to wind the wire around. This maybe a good start for experimenting.
First thing you want to know in with frequency range you want to use the ribbon.
If its for high frequency only you can use something like B64290L0040X830.
I fill it compleet with 1mm wire and calculate the secondair winding lets say 3 windings.
Then wind as much as possible the secondair windings and put a pcb with 2 rings or thick copper wire.
Rob
If its for high frequency only you can use something like B64290L0040X830.
I fill it compleet with 1mm wire and calculate the secondair winding lets say 3 windings.
Then wind as much as possible the secondair windings and put a pcb with 2 rings or thick copper wire.
Rob