Hi
I'm going to be winding my own inductors for my speaker build.
I'm trying to figure out the best dimensions in terms of the length of the coil and the internal diameter.
Looking at Solen Inductors (which is what Caps I'm using), they seem to stick to 1:2 length:diameter eg 25mm length and 50mm diameter.
Jantzen seem all over the place as do most others.
Then I found this article: An introduction to the air cored coil
which suggests trying to make the cross section of the windings as square as possible. This seems to back Solen's measurements.
The best example of a Solen inductor is 1mH which is 16x32x64 lxdxD.
Does anyone have any other info to help?
In the article in the link above, it suggests a circular cross-section would be the best but they are difficult to manufacture. Considering I'm doing everything from scratch, I'm sure I could get close if not a rounded-edged cross section.
Cheers
I'm going to be winding my own inductors for my speaker build.
I'm trying to figure out the best dimensions in terms of the length of the coil and the internal diameter.
Looking at Solen Inductors (which is what Caps I'm using), they seem to stick to 1:2 length:diameter eg 25mm length and 50mm diameter.
Jantzen seem all over the place as do most others.
Then I found this article: An introduction to the air cored coil
which suggests trying to make the cross section of the windings as square as possible. This seems to back Solen's measurements.
The best example of a Solen inductor is 1mH which is 16x32x64 lxdxD.
Does anyone have any other info to help?
In the article in the link above, it suggests a circular cross-section would be the best but they are difficult to manufacture. Considering I'm doing everything from scratch, I'm sure I could get close if not a rounded-edged cross section.
Cheers
There's quite a lot of different ideas about what choke profile produces 'better' sound in an air choke and you'll have to make up your own mind but the design is slightly different between low and high freq too, just to make it more complicated.
If you look at the basic Brooke's design, you have the ratios of a : b : c and the equations to build to a specific inductance.
As you see there, using the middle dimensions in fig 2, a=3c/2 and b=c. but a later work (N Theile, 1960's) found that if you make c=2b/3, (c = 0.67b) you get a 'better' sounding inductor and it still produces the same inductance when you use the design equation, depending how well you stack the windings.
Most of the commercially available inductors/chokes are built for efficiency, price, size, low weight, and not specifically for sound quality - Brooke's type chokes tend to be larger diameter, much heavier and more expensive.
There's quite a few companies in England that still specialize in this area.
If you look at the basic Brooke's design, you have the ratios of a : b : c and the equations to build to a specific inductance.
As you see there, using the middle dimensions in fig 2, a=3c/2 and b=c. but a later work (N Theile, 1960's) found that if you make c=2b/3, (c = 0.67b) you get a 'better' sounding inductor and it still produces the same inductance when you use the design equation, depending how well you stack the windings.
Most of the commercially available inductors/chokes are built for efficiency, price, size, low weight, and not specifically for sound quality - Brooke's type chokes tend to be larger diameter, much heavier and more expensive.
There's quite a few companies in England that still specialize in this area.
AES E-Library >> Air-Cored Inductors for Audio
If I remember correctly, the optimal dimensions are 1:1:1 (inner radius : height : width). That is the cross section is square, and the outer diameter is 4x the height. Thiele's formula gave me very close practical vs. calculated inductance.
If I remember correctly, the optimal dimensions are 1:1:1 (inner radius : height : width). That is the cross section is square, and the outer diameter is 4x the height. Thiele's formula gave me very close practical vs. calculated inductance.
Yes, no problems - this is aimed at the max L/R ration and is indeed 1 : 1 : 1 just like most commercial designs today - best "bang for buck", you would say. And even these 'standard' good quality chokes aren't at all cheap.
There are quite a few other ways of building the same value chokes beside these dimensional ratios particularly when it gets into bi & trifilar, tapered stack/slope, etc - is usually a lot more trouble for subtle improvements - it's quite interesting tho ...
There are quite a few other ways of building the same value chokes beside these dimensional ratios particularly when it gets into bi & trifilar, tapered stack/slope, etc - is usually a lot more trouble for subtle improvements - it's quite interesting tho ...
Less than 10% of a 5-page paper which is very worth reading in full:
(The paper has inches and pounds but an end-note gives Metric math.)
Attachments
The biggest problem when winding your own chokes is the physical effort required for even layers and tension if you haven't access to a winder or lathe.
I remember doing a 10mH some with approx C= 40mm but the Din = 60mm and the Dout ended up something like 180mm! This was with dual #16 wires (bifilar, to get the Rs down) and each layer coated in slow cure epoxie.
Quite the challenge that I never did again but 'superb' quality chokes - got the transformer people to do them after that, and this was back when enamelled copper was a reasonable price too.
I eventually 'saw the light' and used electronic Xovers and sometimes combination of both (partic for mids/tops) - looking at dsps these days (can be quite 'tricky' things too) but still analogue Xovers are familiar and do the job quite well, IMO.
I'd suggest for the original poster, rngDeadeye, to just use the dimensions that Paul, Lazlo, etc included above - can't really go wrong with that.
If you want to keep the windings tight, suggest 'slop' some sealer over each layer as you complete it so avoid any possible vibrations under heavy loads - a bit 'messy' but it works quite well.
All the best ...
I remember doing a 10mH some with approx C= 40mm but the Din = 60mm and the Dout ended up something like 180mm! This was with dual #16 wires (bifilar, to get the Rs down) and each layer coated in slow cure epoxie.
Quite the challenge that I never did again but 'superb' quality chokes - got the transformer people to do them after that, and this was back when enamelled copper was a reasonable price too.
I eventually 'saw the light' and used electronic Xovers and sometimes combination of both (partic for mids/tops) - looking at dsps these days (can be quite 'tricky' things too) but still analogue Xovers are familiar and do the job quite well, IMO.
I'd suggest for the original poster, rngDeadeye, to just use the dimensions that Paul, Lazlo, etc included above - can't really go wrong with that.
If you want to keep the windings tight, suggest 'slop' some sealer over each layer as you complete it so avoid any possible vibrations under heavy loads - a bit 'messy' but it works quite well.
All the best ...
Thanks guys.
The 4Kg of enamelled 1.25mm2 copper wire was £50 delivered in the uk.
According to my crossover design, I need a 12mH inductor. Not only do very few places sell inductors that size, I'm sure 1 would be £75 alone!
I don't have a winder but I have intent and a garage.
One last question, can I assume there's no issue using the same gauge wire for high,mid and low?
The 4Kg of enamelled 1.25mm2 copper wire was £50 delivered in the uk.
According to my crossover design, I need a 12mH inductor. Not only do very few places sell inductors that size, I'm sure 1 would be £75 alone!
I don't have a winder but I have intent and a garage.
One last question, can I assume there's no issue using the same gauge wire for high,mid and low?
Cordless drills/screwdrivers can be pressed into service as make-do winders, although cheap hand-winding machines are available that bolt to the bench with a turns counter. By hand round a suitable piece of pipe is of course the simplest approach.
Wire gauge is something you should choose depending on the desired ESR of the inductor, and the power handling (you don't want the inductor to heat up much).
For high current handling and low frequencies you may need to physically support the coils to prevent mechanical movement - ie wind onto a plastic former and epoxy in place.
Wire gauge is something you should choose depending on the desired ESR of the inductor, and the power handling (you don't want the inductor to heat up much).
For high current handling and low frequencies you may need to physically support the coils to prevent mechanical movement - ie wind onto a plastic former and epoxy in place.
I was considering soaking them in yacht varnish or something similar.
I've built the holder for the real and a bobbin holder which will accept 4 cable ties. I'm going to put a cam lobe on the bobbin and screw a computer mouse near it to press the mouse button for each turn 🙂
Little VB app to count turns.
I've built the holder for the real and a bobbin holder which will accept 4 cable ties. I'm going to put a cam lobe on the bobbin and screw a computer mouse near it to press the mouse button for each turn 🙂
Little VB app to count turns.
Last edited:
Yes, varnish will help - transformers are heated in vacuum oven for deep penetration but just a soak will help.
Most of the inductance comes from the outside layers, obviously, so if you want to add a couple of taps for say, 11.5, 11.0 and maybe 10.5mH, you can just scrape and solder 'flying leads' and tape them down while you finish the winding
With the #16, it's fairly stiff so suggest using gardening gloves and a wooden block to 'tamp down' the layers as you go to keep them flat, esp at the sides - it's pretty slow and unfortunately, hard to keep the layers even if the tension is steady but they don't need to be perfect like you see the commercial ones.
All the best with the challenge.
Most of the inductance comes from the outside layers, obviously, so if you want to add a couple of taps for say, 11.5, 11.0 and maybe 10.5mH, you can just scrape and solder 'flying leads' and tape them down while you finish the winding
With the #16, it's fairly stiff so suggest using gardening gloves and a wooden block to 'tamp down' the layers as you go to keep them flat, esp at the sides - it's pretty slow and unfortunately, hard to keep the layers even if the tension is steady but they don't need to be perfect like you see the commercial ones.
All the best with the challenge.
Well, a moderate success.
I wound it until it was 12.5 but I stupidly forgot about the 10mm threaded steal rod. So it's ended up being bang on 11mH which if I plug into Vituix is still ok as its a filler in an elliptical crossover.
I wound it until it was 12.5 but I stupidly forgot about the 10mm threaded steal rod. So it's ended up being bang on 11mH which if I plug into Vituix is still ok as its a filler in an elliptical crossover.
Congratulations, very nicely done. Rather neat winding jig, just change to stainless bolt/thread next time. Did you apply the varnish?
No I haven't yet. I will probably redo this one to get it up to the original 12mH.
Interestingly, if I squeeze the sides and re-measure, I get an extra 0.5mHs.
I was using a drill to wind the coil but I might make a handle as I think I can make it neater. I had to clamp the wire between two pieces of wood to keep it tight while pulling the trigger on the drill.
Interestingly, if I squeeze the sides and re-measure, I get an extra 0.5mHs.
I was using a drill to wind the coil but I might make a handle as I think I can make it neater. I had to clamp the wire between two pieces of wood to keep it tight while pulling the trigger on the drill.
You seem to be able to make quite neat former sides and centres- why not just screw them together with some stainless screws or bolts and leave it assembled, wooden sides and all so any later over-winding is simple to maintain the tight wind
Yes, IMO, using a hand crank (non magnetic) gives better control and add maybe a friction pad too and if you get fancy, add a tensioner in the centre of the jig.
Not sure if you need the really accurate inductance but the units for both channel need to be quite close to each other - if you add a locking pawl, you can scrape off a bit of insulation and measure, and if near the optional values, solder a lead out wire here and continue the wind.
Yes, IMO, using a hand crank (non magnetic) gives better control and add maybe a friction pad too and if you get fancy, add a tensioner in the centre of the jig.
Not sure if you need the really accurate inductance but the units for both channel need to be quite close to each other - if you add a locking pawl, you can scrape off a bit of insulation and measure, and if near the optional values, solder a lead out wire here and continue the wind.
Interesting, I didn't know stainless wasn't magnetic.
Also interesting about soldering a lead half way through, does this mean it's ok to extend a win by soldering on to the end?
The inductance for this coil doesn't have to be accurate but I of course will make sure they are both the same. I was able to scrape off some enamel and measure.
The bobbin took ages to make and I need 7 inductors per channel...
Also interesting about soldering a lead half way through, does this mean it's ok to extend a win by soldering on to the end?
The inductance for this coil doesn't have to be accurate but I of course will make sure they are both the same. I was able to scrape off some enamel and measure.
The bobbin took ages to make and I need 7 inductors per channel...
Oh dear...
It's just dawned on me that I've used 6 steel threaded rods in my speakers (which are laminated 18mm MDF).
It's just dawned on me that I've used 6 steel threaded rods in my speakers (which are laminated 18mm MDF).
There's a few options that aren't magnetic but stainless screws are easy to find.
Just turn/cut the main circles for the sides of the formers, use a holesaw for the centre between them (couple layers, glue and screw) - make them a bigger diameter than design.
About the 'taps' - if you want to add more value to the chokes and later options, suggest picking some possible unductance values lower than the current 12mH (10, 11, etc) then just solder a couple of lead-out wires at those points - they'll probably all be on the outer layer(s) anyway so not that difficult and just tape them out of the way when finishing the wind.
This way, with the same formers, you'll be pretty same mH value for same number of turns - any scrapped off area that's unused, just paint over with something like a pva glue and/or tape - not really high voltages here.
If Xover chokes aren't next to steel rods, very liitle chace of effecting value but try a test with one of the existing chokes you've wound, and position is near to where you want to assemble Xover and see if any change.
Also, multiple chokes near each other in an assembled passive Xover will effect each other so need to be oriented at right angles to each other (there are 3 axis so not a problem and can use Ali bracket and same stainless screws to stand up chokes on edge - this also will show up with measure after assembly.
Just turn/cut the main circles for the sides of the formers, use a holesaw for the centre between them (couple layers, glue and screw) - make them a bigger diameter than design.
About the 'taps' - if you want to add more value to the chokes and later options, suggest picking some possible unductance values lower than the current 12mH (10, 11, etc) then just solder a couple of lead-out wires at those points - they'll probably all be on the outer layer(s) anyway so not that difficult and just tape them out of the way when finishing the wind.
This way, with the same formers, you'll be pretty same mH value for same number of turns - any scrapped off area that's unused, just paint over with something like a pva glue and/or tape - not really high voltages here.
If Xover chokes aren't next to steel rods, very liitle chace of effecting value but try a test with one of the existing chokes you've wound, and position is near to where you want to assemble Xover and see if any change.
Also, multiple chokes near each other in an assembled passive Xover will effect each other so need to be oriented at right angles to each other (there are 3 axis so not a problem and can use Ali bracket and same stainless screws to stand up chokes on edge - this also will show up with measure after assembly.
As it happens I was considering splitting the x-over into 3 separate parts and locating each part in different places in the speaker.
