Question for experienced electronics folk, Re: Toroids

I’m doing a prototyping project at home and I have to wrap some very small toroids with a lot of windings. I’ve wound plenty before and am familiar with the conventional methods but I’m wondering if you might have any wizzy tips that will make these ones easier.

I’m hoping to get 350 windings of #44 on to a B64290P37X38. Actually it will be a pair of 175 each windings so I can either wind a pair of wires x175 or a single wire of 350 windings and tap the middle.

Outside diameter: 6.3mm
Height: 2.5mm
Inside diameter: 3.8mm


And just for fun, I'm trying to use all the same components for prototyping as I will for the final design so I had to solder some leads onto some small parts like this schottky diode so I could breadboard with them. Crappy job I know, but I'm not wasting my time making breadboarding components look pretty.

Thanks in advance for any advice.
Cheers!
CB

a reply to: CraftBuilder

This is not intended to be a serious reply but calculate the totlal length of wire required, cut it and start wrapping. Don't fret, you Will get it done, eventually, HA!

a reply to: CraftBuilder

Are you winding a transformer or an inductor?

If a transformer, you could pair the primary/primaries and secondary/secondaries wires and pull them through at the same time.

You could also use this same method for faster winding, i.e: pull more than one wire through at once. The downside of this method is that the neatness, and therefore the tightness of the pack of the windings is likely to be harder to achieve than if you only wound a single wire (or perhaps a pair, in the case of a transformer). If you wind, say, four wires, there are likely to be multiple crossovers of wires and it will not lay as 'flat' as you might like.

You could also use the 'multiple wire method' to achieve a bifilar wind, too.

An articulated pin vice would also be very handy, too! Cyanoacrilate (superglue), used sparingly, will help hold your partially completed windings in place without too much bulk.

With the small diameter torroid and the shottky, I'd hazard a guess that you are wiring for HF?

Be careful soldering wires on the smd's with a standard iron and solder. Smd's do not sink heat well, and are easily heat damaged. Low temp iron and solder.

Does it have to be insulated wire? .. just curious

a reply to: chr0naut
Thanks for the advice. It’s a transformer, actually an inductor with a sense winding so basically a transformer. Wrapping the windings together as a pair is what I meant in the OP. The math looks good to over 500 turns so I think pairing them up will be fine for now. The biggest problem with hand winding a single wire is that by the time you do a few hundred turns with #44 it's ready to fall apart from all of the flexing.

The application is for a DC-DC converter. A very small one. The input has up to 0.3 volts at 650 micro amps available sporadically. It has to boost it to 3.3 volts and store it until there is enough power to run a microcontroller which performs a function and then sleeps or powers down completely until there is enough charge to do it again. The entire device can’t go over 3.5 grams with peripherals, housing and transport system so I’ve got the DC-DC converter section down to three components. Fortunately, low V and I so very small components. The toroid is the size of the entire rest of the prototype and I'm working to get that down quite a bit. Eventually everything will end up getting die bonded onto a single substrate.

a reply to: smirkley
All components have soldering specifications in the datasheet and I’m always well below them when I hand solder. I unfortunately, had a soldering pencil thrust into my hand before I was given a soother and have soldered thousands of devices since SMD came along. I don’t remember ever overdoing a component. As long as you are following proper soldering practices there is a decent margin. The work I mainly do now is on systems so small that we can’t use PCBs, so all of our prototypes and most production assemblies are done with hand soldering or manual jigs, even BGA. We design with mechanical CAD in 3D instead of layout software and have 3D libraries for all of the components and traces we use. It’s like the soldering version of Ultimate Tetris.

a reply to: nOraKat
Yep.

originally posted by: CraftBuilder
a reply to: chr0naut
Thanks for the advice. It’s a transformer, actually an inductor with a sense winding so basically a transformer. Wrapping the windings together as a pair is what I meant in the OP. The math looks good to over 500 turns so I think pairing them up will be fine for now. The biggest problem with hand winding a single wire is that by the time you do a few hundred turns with #44 it's ready to fall apart from all of the flexing.

The application is for a DC-DC converter. A very small one. The input has up to 0.3 volts at 650 micro amps available sporadically. It has to boost it to 3.3 volts and store it until there is enough power to run a microcontroller which performs a function and then sleeps or powers down completely until there is enough charge to do it again. The entire device can’t go over 3.5 grams with peripherals, housing and transport system so I’ve got the DC-DC converter section down to three components. Fortunately, low V and I so very small components. The toroid is the size of the entire rest of the prototype and I'm working to get that down quite a bit. Eventually everything will end up getting die bonded onto a single substrate.

I have just been reviewing surface mount transformers and the majority of really small ones use a rod or bobbin (Iron or Ferrite) fitted into a soft iron 'cap', rather than a toroid. I understand the weight and efficiency advantages of a ferrite toroid but wonder if an off-the-shelf small SMD xformer might fit the spec?

a reply to: chr0naut
I’ve been through a plethora of smd, bobbin style transformers and even rewound a few of them but I’m really pushing the envelope with this power supply and they aren’t cutting it. Its tight specs when even the scope probes pull down your power supply. I think manufactures like the bobbins wherever they can get away with it because they can be machine wound. There is a point where you can’t get a winding shuttle into a small toroid core and manual winding is very expensive. I’ve gone as far as to grind the sides off of two channel balun cores (images below) so that I could get more windings on and that basically turns them into a bobbin.

For anyone interested -

You can build something that can mimic the technology used in the Japanese Mag-Lev (magnetic levitation) train.

It employs induction.