Hey guys, i'm preparing to build a ribbon microphone and i've decided on winding my own transformer.
I've done the math and come up with the appropriate wire gauges and turn counts, and i will be using a smallish iron powder toroidal core, but i was wondering if anyone potentially has more knowledge when it comes to audio transformer construction. My biggest question currently has to do with the "layout" of the windings, more specifically, should the primary and secondary windings be wound into each other, or separated along the circumference of the toroid? i'll attach an iffy paint drawing to illustrate what i mean if it's unclear 🙂
(also another thing, but does the orientation of the windings in relation to each other, i.e. if one's would Clockwise and the other counter-clockwise?)
I've done the math and come up with the appropriate wire gauges and turn counts, and i will be using a smallish iron powder toroidal core, but i was wondering if anyone potentially has more knowledge when it comes to audio transformer construction. My biggest question currently has to do with the "layout" of the windings, more specifically, should the primary and secondary windings be wound into each other, or separated along the circumference of the toroid? i'll attach an iffy paint drawing to illustrate what i mean if it's unclear 🙂
(also another thing, but does the orientation of the windings in relation to each other, i.e. if one's would Clockwise and the other counter-clockwise?)
Maximum flux linkage when primary and secondary are in their own layers and each winding is tightly packed (both side by side turns and layer on layer if any). This minimises (annoying american spell checker) losses and maximises fidelity. However having the correct core material is critical.
What sort of input and output impedance do you expect? Winding a high impedance torroid transformer can be a bit tricky to do.
So, like the left drawing? And what effects do different core materials have? I'd like to use as much stuff that i just have laying around as possible and that's mostly ferrite and iron powder toroids, as well as a couple ferrite bobbin "e-shape" kinds of core i could also use. Cheers!
For the input i wanted it to be very low on account of ribbon motors being very low output, so i went with 26 turns of 0.8mm wire which comes out to about 0.02 ohm, the output had to be somewhere in the 50-200 region so i counted out 988 turns of 0.1mm wire which should be about 64 ohm. The main goal is a low primary Z, secondary in the range i mentioned, that comes out to a 1:30-40 ratio
The resistance of the wire on the primary will not matter so much.
You need to know the minimum inductance needed on the primary side.
It helps to know the permeability of the core used. Some cores can have a permeability of 1 up to maybe 100. It will affect how much inductance per turn.
Also remember inductance will equal some constant times number of turns squared. A more permeable core will give you more inductance per turn but it will also saturate easier.
Do you happen to have an inductance meter? If not, you may be able to find inductance using a signal generator and a low impedance source as a driver.
At least a microphone transformer should be easy to build considering the low signal level.
You need to know the minimum inductance needed on the primary side.
It helps to know the permeability of the core used. Some cores can have a permeability of 1 up to maybe 100. It will affect how much inductance per turn.
Also remember inductance will equal some constant times number of turns squared. A more permeable core will give you more inductance per turn but it will also saturate easier.
Do you happen to have an inductance meter? If not, you may be able to find inductance using a signal generator and a low impedance source as a driver.
At least a microphone transformer should be easy to build considering the low signal level.
I don't have an inductance meter nor a signal generator, i am however pretty sure that the permeability is about 75, if the coloring of the core checks out
That's in your browser. Possibly you can change it. You can on desktop Firefox and Chrome. Chrome has seven flavours (sic) of English aside from the obligatory "US".
Attachments
I do not think you will get the required inductance from 26 turns on the primary. You will likely end up with low frequency loss.
Looking up a ribbon microphone, I see the impedance is 200 ohms.
Assuming a 40 hz cutoff I come up needing 786mH. L = R / 2(pi)f
I think 26 turns might get you into the 10mH range though a way to measure your core inductance is needed.
You may be able to measure inductance with the headphone output of a computer if you find a signal generator app. You can put a 200 ohm resistor in series with your primary and see what frequency gets you even values between load resistor and the inductor.
One catch is you need to measure the inductance at a fairly low voltage matching what you would get from the microphone.
Measuring at headphone level will likely saturate the core and give lower inductance numbers.
Looking up a ribbon microphone, I see the impedance is 200 ohms.
Assuming a 40 hz cutoff I come up needing 786mH. L = R / 2(pi)f
I think 26 turns might get you into the 10mH range though a way to measure your core inductance is needed.
You may be able to measure inductance with the headphone output of a computer if you find a signal generator app. You can put a 200 ohm resistor in series with your primary and see what frequency gets you even values between load resistor and the inductor.
One catch is you need to measure the inductance at a fairly low voltage matching what you would get from the microphone.
Measuring at headphone level will likely saturate the core and give lower inductance numbers.
Well, ribbons tend to output in the microvolts so that won't really be easy, maybe i should just wait until i have the rest of the thing built and mess around with the transformer until i get something good?
Anyone know if the ribbon microphones come with an internal transformer? That would change the input transformer a bit.
Looking at this link it suggest a user changed the transformer to hopefully improve performance.
https://www.sweetwater.com/store/de...-F-EGC_r9tfF9k3VUOMaAodGEALw_wcB&gclsrc=aw.ds
Looking at this link it suggest a user changed the transformer to hopefully improve performance.
https://www.sweetwater.com/store/de...-F-EGC_r9tfF9k3VUOMaAodGEALw_wcB&gclsrc=aw.ds
Both sides are VERY low impedance. 150Ω on the secondary and typically 0.05Ω on the primary.
Over-winding and interleaving are pointless at low impedance. Capacitance is not a problem.
Sufficient inductance IS needed, as with any audio transformer. It may be teeny-Henries but insufficient inductance is poor bass.
Stray primary resistance WILL destroy the system S/N.
My old ribbons used #16 primary lead-ins for low resistance; paired for hum reduction.
You have to compromise inductance against resistance; you can't just wind many turns of fine wire.
I have never seen a first-timer wind a satisfactory ribbon transformer.
I have "heard music" through a 3V to 120V power transformer tapping. Not full range and rather hissy.
Over-winding and interleaving are pointless at low impedance. Capacitance is not a problem.
Sufficient inductance IS needed, as with any audio transformer. It may be teeny-Henries but insufficient inductance is poor bass.
Stray primary resistance WILL destroy the system S/N.
My old ribbons used #16 primary lead-ins for low resistance; paired for hum reduction.
You have to compromise inductance against resistance; you can't just wind many turns of fine wire.
I have never seen a first-timer wind a satisfactory ribbon transformer.
I have "heard music" through a 3V to 120V power transformer tapping. Not full range and rather hissy.
Iron powder cores are typically for RF, they have low inductance and thus not great for audio as they have lots of distributed gap (which is why
they are low-loss at RF)
Silicon steel (aka magnet steel) laminates are used for high inductance low frequency magnetics, or sometimes amorphous iron or sometimes nickel alloys.
they are low-loss at RF)
Silicon steel (aka magnet steel) laminates are used for high inductance low frequency magnetics, or sometimes amorphous iron or sometimes nickel alloys.
If you want to use a toroid, opt for a ferrite one instead. It is far from ideal, but it will be a great improvement over the powder one: many types of ferrites used in common-mode chokes for example have a µr in excess of 5,000, and their poor saturation limit doesn't matter in the least here. In addition, ferrites have a good initial permeability thanks to their low hysteresis.
Btw this is great info, but one thing i'd like to add, i've based all my numbers on commercially available transformers, for instance the RTX-35 from bumblebeepro.com and the RTP-35 from diy audio components, both of which have similar DCR on primary and are specified at 5mH on primary, and from what i'm hearing those are typically pretty decently regarded and the kits they come packed in seem to sound pretty good, is there something larger i'm missing here?
There are some nano crystalline material small toroids available which could be very good. They are intended for common mode choke applications but could
be suitable for audio use. They have high permeability and good maximum flux density limit. Their spec will give an Al value of inductance which if multiplied by the
intended number of winding turns will give you the final inductance.
be suitable for audio use. They have high permeability and good maximum flux density limit. Their spec will give an Al value of inductance which if multiplied by the
intended number of winding turns will give you the final inductance.
So, i've rounded up the different appropriately sized toroids i had laying around as well as my guesstimation of the core material based on the color coding (which seems to kinda be a thing? albeit not very well documented)
They are filthy on account of them being salvaged out of old PSUs and other hot enviroments. Any suggestions as to which would fit best would be appreciated!
They are filthy on account of them being salvaged out of old PSUs and other hot enviroments. Any suggestions as to which would fit best would be appreciated!
Attachments
If your MnZn is actually made of that material, it is the one to go for.
To make sure it is, test its resistivity by probing it with an ohmmeter and two sharp spikes as test probes: you need to push them hard enough to pierce the lacquer layer and make contact with the actual ferrite material.
You should read something, typically in the kohm or tens of K range, maybe hundreds of K.
If it is in the megohm range or infinity, it is not good: in fact, the lower the better. µr and the resistivity are closely related.
You should devise a way to measure the inductance/Al, because without hard engineering data, it is difficult to properly design something, and you need to know the minimum required number of turns for your target LF cutoff frequency
To make sure it is, test its resistivity by probing it with an ohmmeter and two sharp spikes as test probes: you need to push them hard enough to pierce the lacquer layer and make contact with the actual ferrite material.
You should read something, typically in the kohm or tens of K range, maybe hundreds of K.
If it is in the megohm range or infinity, it is not good: in fact, the lower the better. µr and the resistivity are closely related.
You should devise a way to measure the inductance/Al, because without hard engineering data, it is difficult to properly design something, and you need to know the minimum required number of turns for your target LF cutoff frequency
Sorry Friend but then you can not even dream on designing transformers on unknown material "mystery" cores.