Iam going to make my own transformers for valve amplifier and Iam going to use kapton adhesive tape as interwinding insulation.
Electric grade kapton tapes are expensive, but there might be cheaper alternative: kapton tape for 3D printing. It looks like same material except printing tapes dont have certificates for electric use.
What do you think?
Has anyone ever tried to use printing kapton tape for electric insulation?
Electric grade kapton tapes are expensive, but there might be cheaper alternative: kapton tape for 3D printing. It looks like same material except printing tapes dont have certificates for electric use.
What do you think?
Has anyone ever tried to use printing kapton tape for electric insulation?
Define “expensive” and “cheap”. I’ve been using 3M 1350F-2Y - the yellow Mylar stuff specifically for securing SMPS trafo windings. Perfect for regular EI laminated trafos as well. It’s about $30 a roll but goes a long way. I don’t consider it too bad for the money. Kapton may hold up to higher temps, but how hot are you going to be running? And the yellow stuff seems to be a bit less fragile.
Now of you’re trying to get off really cheap, I’ve used regular 99 cent black electrical tape. It makes a gooey mess, and only use it if your trafo has a full plastic bobbin and not just a paper former. And you’ll need to keep it under 40C temp rise. Many times that’s good enough, but I wouldn’t use it for HT windings. Low voltage stuff like a custom heater voltage, hell yeah.
Now of you’re trying to get off really cheap, I’ve used regular 99 cent black electrical tape. It makes a gooey mess, and only use it if your trafo has a full plastic bobbin and not just a paper former. And you’ll need to keep it under 40C temp rise. Many times that’s good enough, but I wouldn’t use it for HT windings. Low voltage stuff like a custom heater voltage, hell yeah.
I can get 15mm width x 33 meter long printing kapton tape for some $7 a piece.
Iam going to use kapton beacuse its superior insulating properties.
Iam going to use kapton beacuse its superior insulating properties.
Ebay Kapton tape = £5 - 12mm x 50m, same but 3M @ Farnell = £60 ish. I've been using Ebay Kapton tape for a few years, it passes a 10kv insulation test, however the last lot I bought was flammable, so, cheap Kapton tape may do, but if your tfmr overheats and catches fire,your screwed,especially if you've sold the amp with said tfmr and it causes a house fire.
See this stuff - Adhesive Yellow Therma Tape - 12mm x 0.060mm x 66m - Brocott UK that's what I'm getting next time. The last lot of Ebay Kapton tape had a label "Kopton" tape,not very reasuring.
Andy.
See this stuff - Adhesive Yellow Therma Tape - 12mm x 0.060mm x 66m - Brocott UK that's what I'm getting next time. The last lot of Ebay Kapton tape had a label "Kopton" tape,not very reasuring.
Andy.
If you’re selling amplifiers you don’t need to be building your own transformers - unless you’ve got the cash to spend on the safety certs and can mark your prices into the stratosphere to compensate. The only thing I ever *sell* are speakers. If someone plugs them in the wall any lawyer will say the result is on them, not me.
For home brew amplifiers for your own use you simply assume all the risk. Nothing bad usually happens if it’s competently built, but you should always be in a position to pull the plug when it isn’t behaving normally.
For home brew amplifiers for your own use you simply assume all the risk. Nothing bad usually happens if it’s competently built, but you should always be in a position to pull the plug when it isn’t behaving normally.
Sorry, what do you want with a just 12 mm wide isolating tape? That allows you only to build very tiny transformers without overlapping, which should be avoided to fill the winding window with as much copper as possible. The isolating layer shoud be as wide as the bobbin, or even more with serrated edges.
Best regards!
Best regards!
Why not use kapton? Its thin and has good insulation properties. However the cheap stuff likely isn’t kapton. Like everything else it’s cheap for a reason.
In Germany since decades most transformer manufacturers use Mylar (here under the moniker Hostaphan) foil as layer or inter winding isolation.
Best regards!
Best regards!
Why not indeed?!
(Kapton tape is quite commonly used for insulating windings on large AC machines, for a Class H temperature rise, if my recall is spot on)
I have used Kapton film as IC insulators, and kapton Sellotape for securing primary layers in a transformer I wound. It's perfectly good for 2kV for a single layer. The biggest issue is it degrades over time, and eventually turns brittle and dusty.
On a similar note, regular rubber amalgamation tape can also be used for isolation of primary and secondary, though its thickness when stretched is still more than most tapes.
Again it's ok if you are not going to get heat above about 60C.
Mylar, or Melinex (TM) is indeed used for winding electrical machines, but I dont think it is used for inter-STRAND insulation, nor coil insulation, but mainly for coil slot liners, and rotors. But mostly nomex is used, in my experience. (Wet transformers where Nomex is incorporated with VPI resins, and VPI insulated Stators. Dry types may differ, as may "Resin Rich" systems, which use tapes with resin pre-impregnated insulation, glass and polyester woven tapes.)
ABB may do it a little different, though that would be the only thing, as everything else is cookie cutter design.
Last edited:
Well, exactly? Why not mylar, mylar-nomex composite or even mylar-pressed paper composite? Kapton alone is quite expensive. Is its high temperature resisting properties justified?
From my point of view, for transformers, the Mylar film I am used to seeing is too heavy gauge, maybe 0.25mm at a guess. (And if an unwanted roll of 50mm wide 0.1mm adhesive free Kapton tape is available, plus winders varnish, then it would be foolish to not use it!)
Nomex/mylar/Nomex is also too cardboard like, and difficult to form for small windings.
I'm sure thinner sheet/tape would be more easily wound around a small formed coil.
For a quick LV transformer, I have been known to use decorators Masking tape, which is just waxed paper.
Glass tape is my favourite though it makes resin potting/varnish painting and a wet lay up, almost a necessity.
Nomex/mylar/Nomex is also too cardboard like, and difficult to form for small windings.
I'm sure thinner sheet/tape would be more easily wound around a small formed coil.
For a quick LV transformer, I have been known to use decorators Masking tape, which is just waxed paper.
Glass tape is my favourite though it makes resin potting/varnish painting and a wet lay up, almost a necessity.
Last edited:
Well, I've seen 0.05mm Mylar and I'm aware even thinner exists. Nomex-Mylar-Nomex composite is 0.15 minimum and is quite easily bending, and single Nomex-Mylar 0.1mm.
Kapton has a fairly high dielectric constant. Use of Kapton between each layer can raise the parasitic capacitances in a transformer considerably over paper based tape. My guess is that Kapton may be the cause for failure in several DIY SMPS transformer experiments that I wound.
As a kid in the late 60's I took apart the power transformers from TV sets. I ripped off all the windings except for the primary and wound my own secondaries using ordinary household masking tape for the interlayer insulation.
Power transformers for solid state amps were not easily found, and beyond my price range, so I made my own. Oddly none of those transformers ever died, but some were rather brown inside after a few years of use.
As a kid in the late 60's I took apart the power transformers from TV sets. I ripped off all the windings except for the primary and wound my own secondaries using ordinary household masking tape for the interlayer insulation.
Power transformers for solid state amps were not easily found, and beyond my price range, so I made my own. Oddly none of those transformers ever died, but some were rather brown inside after a few years of use.
^ This I can agree, is probably the worst side effect of a "thin" primary-secondary isolation tapes.
High Dielectric constant, plus very thin tape, gives worst parasitic capacitance - compared with using something much thicker, or much more "leaky".
I'd agree that masking tape works pretty well, though I havent risked trying it for a valve OPT
High Dielectric constant, plus very thin tape, gives worst parasitic capacitance - compared with using something much thicker, or much more "leaky".
I'd agree that masking tape works pretty well, though I havent risked trying it for a valve OPT
Thank you all for comments!
Especially Andy's warning about flammability of some cheap kapton tapes.
Tubelab note about high dielectric constant i also interesting, but paper based insulation need some varnish impregnation for long term stability and that will also raise the capacity.
Nomex-mylar-nomex is excellent, but in order to get good price you must buy whole factory role and thats too large for just few transformers. If you buy on meter basis its almost expensive as kapton and hard to find. And not self adhesive, so harder to work with.
Especially Andy's warning about flammability of some cheap kapton tapes.
Tubelab note about high dielectric constant i also interesting, but paper based insulation need some varnish impregnation for long term stability and that will also raise the capacity.
Nomex-mylar-nomex is excellent, but in order to get good price you must buy whole factory role and thats too large for just few transformers. If you buy on meter basis its almost expensive as kapton and hard to find. And not self adhesive, so harder to work with.
Totally, Nomex is not cheap in small quantities.
I am lucky that I have several almost used rolls of Kapton, with perhaps only 2metres left on the roll.
Good old reconstituted mica tape, would also be suitable, but may need a couple of layers.
This will need varnish to fix, though also some mica tapes are resin impregnated, to cure requires some time in an oven.
I am lucky that I have several almost used rolls of Kapton, with perhaps only 2metres left on the roll.
Good old reconstituted mica tape, would also be suitable, but may need a couple of layers.
This will need varnish to fix, though also some mica tapes are resin impregnated, to cure requires some time in an oven.
There are local distributors here in Bulgaria, providing small quantities (1kg) of Nomex-Mylar-Nomex. I could give you contact info if you wish to ask them, if they could ship to other countries.
About Tubelab's point. If it will be a mains transformer (50-60Hz), then high capacitance shouldn't be a safety issue. The reactance will be quite high at such low frequency. But for SMPS, the dielectric can get quite loaded due to capacitance and overheat.
About Tubelab's point. If it will be a mains transformer (50-60Hz), then high capacitance shouldn't be a safety issue. The reactance will be quite high at such low frequency. But for SMPS, the dielectric can get quite loaded due to capacitance and overheat.
Here I see Nomex laminate in rolls of various widths, and I have not seen any in tape like widths (though it may exist, unknown to me)
Glass tape 25mm wide is similar in cost to Kapton, maybe £30/30m roll. The glass fabric wont be troubled by temperature, though I'm unsure about the silicone based adhesive.
I'd probably choose to resin dip my own OPTs, which may be worse for stray C than leaving an air void, but will help avoid turn to turn shorts, or inter winding shorts. But I am no expert, it just feels a far better way to make something that will last.
Glass tape 25mm wide is similar in cost to Kapton, maybe £30/30m roll. The glass fabric wont be troubled by temperature, though I'm unsure about the silicone based adhesive.
I'd probably choose to resin dip my own OPTs, which may be worse for stray C than leaving an air void, but will help avoid turn to turn shorts, or inter winding shorts. But I am no expert, it just feels a far better way to make something that will last.
If it's for an OPT, the choice of kapton will be even worse.
Usually you don't need to worry about turn to turn short, only if your winding is sloppy and you're crossing over turns. Voltage gradients in turns inside an OPT is usually extremely small.
The places to worry are usually the high capacitance ones, layers with huge voltage swing closest to the plate of the tube relative to the secondary. There you need distance + low epsilon + high dielectric strength. These layers are mostly located at the start of the coil (bottom), where the Mean turn length is less.
Primary to primary layers usually have low capacitance too. The whole clumped capacitance of a classically wound primary coil can be calculated by the formula:
Cp = 4/3 * Cst * NL - 1 / NL^2
Where:
Cst is the static capacitance, calculated as a plate capacitor
NL is the number of layers in the coil
For a primary coil with 20 layers and an average Cst of 1nF, the self capacitance becomes 63pF. It becomes even less when you start interleaving, because of the series connection of primary sections, where you start dividing the capacitance by the number of sections.
So in a classic step-down OPT, we concentrate our thoughts on primary to secondary capacitances. There's a trick though. In some situations we can deliberately "shift" some of the capacitance into the primary layers.
Usually you don't need to worry about turn to turn short, only if your winding is sloppy and you're crossing over turns. Voltage gradients in turns inside an OPT is usually extremely small.
The places to worry are usually the high capacitance ones, layers with huge voltage swing closest to the plate of the tube relative to the secondary. There you need distance + low epsilon + high dielectric strength. These layers are mostly located at the start of the coil (bottom), where the Mean turn length is less.
Primary to primary layers usually have low capacitance too. The whole clumped capacitance of a classically wound primary coil can be calculated by the formula:
Cp = 4/3 * Cst * NL - 1 / NL^2
Where:
Cst is the static capacitance, calculated as a plate capacitor
NL is the number of layers in the coil
For a primary coil with 20 layers and an average Cst of 1nF, the self capacitance becomes 63pF. It becomes even less when you start interleaving, because of the series connection of primary sections, where you start dividing the capacitance by the number of sections.
So in a classic step-down OPT, we concentrate our thoughts on primary to secondary capacitances. There's a trick though. In some situations we can deliberately "shift" some of the capacitance into the primary layers.
Last edited: