I have a design I'm currently designing the build out for. It's a cascode with B+ of -320V to +320V thus the design has separate heater supplies to stay within the heater-cathode limits.
Is there any benefit from twisting/platting/tying the different heater pairs together?
The only issue I could see is ensuring the conductor to conductor insulation is high enough (ie from the lowest heater reference to the highest taking into account any AC peaks). Given they're running off separate isolated secondaries the only other issue is a small amount of capacitance.
It would look neater..
Is there any benefit from twisting/platting/tying the different heater pairs together?
The only issue I could see is ensuring the conductor to conductor insulation is high enough (ie from the lowest heater reference to the highest taking into account any AC peaks). Given they're running off separate isolated secondaries the only other issue is a small amount of capacitance.
It would look neater..
I've written this before, but it might be good to put it down again. There is a proper way to twist wires, and almost everyone I've met don't know it!
When done, the twisted pair will have almost no inclination to unravel.
Even when heating during soldering.
Moreover, there is no latent or induced residual strain in the conductors inside.
Just though y'all'd like to know.
⋅-=≡ GoatGuy ✓ ≡=-⋅
PS: for practice ... and good bench-keeping, I usually make about 10 to 20 of these, all at the same time, with my wife handling the drill. The uncut loop-end hangs over a headless nail on the wall quite conveniently. Especially if I make them in a variety of colors. Reach up and grab one! Its amazing what you can do in a couple of minutes...
Take 1 insulated, stranded wire, about 2 meter (6 ft) long, and 'fold in half'.
Put the loop into a standard-chuck drill.
Tighten. Do not twist yet!
With a pair of cloth (not leather or rubber) gloves, hold the pair sticking out of chuck
… between the thumb and first 2 fingers, keeping the wires separate.
With other hand, engage the drill to very slowly twist the wires.
Move the gloved down to match the point-of-twisting.
It might get tangled up the first time you try, but it definitely is worth the trouble. Put the loop into a standard-chuck drill.
Tighten. Do not twist yet!
With a pair of cloth (not leather or rubber) gloves, hold the pair sticking out of chuck
… between the thumb and first 2 fingers, keeping the wires separate.
With other hand, engage the drill to very slowly twist the wires.
Move the gloved down to match the point-of-twisting.
When done, the twisted pair will have almost no inclination to unravel.
Even when heating during soldering.
Moreover, there is no latent or induced residual strain in the conductors inside.
Just though y'all'd like to know.
⋅-=≡ GoatGuy ✓ ≡=-⋅
PS: for practice ... and good bench-keeping, I usually make about 10 to 20 of these, all at the same time, with my wife handling the drill. The uncut loop-end hangs over a headless nail on the wall quite conveniently. Especially if I make them in a variety of colors. Reach up and grab one! Its amazing what you can do in a couple of minutes...
Last edited:
The old school and proper way of heater lead-dress is twisted and layed flat against the chassis for mimimal hum issues.
I don't see any details on twist angle or turns per inch or insulation thickness? They all impact the external field level that is the 'raison d'etre' of twisting.
NickKUK, you could measure the capacitance between the heater windings on the power transformer - if that is measurable (due to the windings being on adjacent layers) then that capacitance may be more than the external capacitance added by bundling the heater wiring. The elevated DC nodes need to be suitably decoupled to gnd, to avoid heater-cathode leakage signal from being an influence.
I read somewhere that it's 11 turns per foot, can't find the source for this, which means most valve amp htr wiring is over twisted. Whether this matters I've no idea.
As to the angle, surely T/ft dictates the angle,the two being inter-dependent.
Andy.
I might be wrong, but I don't think the original question has been answered:
"Is there any benefit from twisting/platting/tying the different heater pairs together?"
Here's my attempt:
Twisting a (single) pair is the normal approach. The magnetic field depends on the distances to a current. As the currents are of the same magnitude but have an opposite direction, the sum of the two magnetic fields can be minimized by keeping the wires close together. Then the distance to them will be nearly equal and the two fields will nearly cancel each other.
Twisting the two wires will keep them close and thus the fields will cancel nicely. Any resulting magnetic field will be the result of 1 wire being closer (on average) than the other. Twisting results in both wires taking turns which one is closest, thereby averaging the two out of phase residual magnetic fields resulting in even better field cancelation. All this depends on equal and out of phase currents.
(At the moment...) I don't see how twisting both pairs also would help in field cancelation.
"Is there any benefit from twisting/platting/tying the different heater pairs together?"
Here's my attempt:
Twisting a (single) pair is the normal approach. The magnetic field depends on the distances to a current. As the currents are of the same magnitude but have an opposite direction, the sum of the two magnetic fields can be minimized by keeping the wires close together. Then the distance to them will be nearly equal and the two fields will nearly cancel each other.
Twisting the two wires will keep them close and thus the fields will cancel nicely. Any resulting magnetic field will be the result of 1 wire being closer (on average) than the other. Twisting results in both wires taking turns which one is closest, thereby averaging the two out of phase residual magnetic fields resulting in even better field cancelation. All this depends on equal and out of phase currents.
(At the moment...) I don't see how twisting both pairs also would help in field cancelation.
Distance between twists should be less than distance to other circuit of interest.
Color the wires white and black. Put your eye at the other circuit. View through ground glass. If the cable appears uniform 50% gray, equal parts black and white, the field cancels. If it is more-white or more-black, it isn't cancelling.
This is fairly easy to do with overhead power and telephone cables. Actually because the overall length is typically long, a transposition every 100 or 1,000 feet (30m-300m) was often sufficient.
It is downright impossible to twist all the way down to a 6V 12AX7 socket. You always end up with one wire bothering triode 1 and the other wire bothering triode 2. A possible trick is to drill the chassis and bring in only 1/8" of heater wire to each side, but this has other problems. The
There have been a smattering of scientific articles on the topic of twisted pairs and external field level since Moser and Spencer, 1968. One of the easier to read is linked below.
There are two sides to the coin where twisting may be important. If the aim is to reduce coupling into a wire with a sensitive signal, then twist the sensitive signal wire with its ground path (or at least locate them close and parallel to each other, so as to minimise the 'pickup loop' side of the coin for magnetic field transfer.
If the aim is to reduce the external field level from wires running heater current, then use coax, or twisted wire where the wires are as close together as possible, and twisted as uniformly as possible. A better outcome is when using thin insulation walled wire - even to the point of using new motor winding wire with thick quality enamelled insulation - or fancy high temp thin walled insulation. So no, don't twist two separate heater wire runs together if at all possible.
But for the bulk of us it is all about routing heater wiring away from sensitive leads (due mainly to capacitive coupling rather than magnetic), and that can mean a 3D assessment of what is happening at each valve socket, and is often counter to what may look aesthetically pleasing to the eye. Separation distance is often the best medicine. But then we roll back to square one and ask whether it makes a detectable difference anyway - as so many have proffered that their amps are silent and they don't twist heater wiring.
(PDF) Analysis of Twisting of the Wires as a Method Reducing the Influence of Stray Field on Accuracy of Magnetic Measurements | T. Meydan and Stan Zurek - Academia.edu
There are two sides to the coin where twisting may be important. If the aim is to reduce coupling into a wire with a sensitive signal, then twist the sensitive signal wire with its ground path (or at least locate them close and parallel to each other, so as to minimise the 'pickup loop' side of the coin for magnetic field transfer.
If the aim is to reduce the external field level from wires running heater current, then use coax, or twisted wire where the wires are as close together as possible, and twisted as uniformly as possible. A better outcome is when using thin insulation walled wire - even to the point of using new motor winding wire with thick quality enamelled insulation - or fancy high temp thin walled insulation. So no, don't twist two separate heater wire runs together if at all possible.
But for the bulk of us it is all about routing heater wiring away from sensitive leads (due mainly to capacitive coupling rather than magnetic), and that can mean a 3D assessment of what is happening at each valve socket, and is often counter to what may look aesthetically pleasing to the eye. Separation distance is often the best medicine. But then we roll back to square one and ask whether it makes a detectable difference anyway - as so many have proffered that their amps are silent and they don't twist heater wiring.
(PDF) Analysis of Twisting of the Wires as a Method Reducing the Influence of Stray Field on Accuracy of Magnetic Measurements | T. Meydan and Stan Zurek - Academia.edu
Last edited:
"aesthetically pleasing to the eye"........ hmm?
This thread is tipping the scales at sillyness.
There is a proper and sensible way to dress wiring, that has been used for ages, and results in proper operation.
But worrying about how the wiring "looks", how many "twists per inch" and such just borders on OCD and rediculousness.
This thread is tipping the scales at sillyness.
There is a proper and sensible way to dress wiring, that has been used for ages, and results in proper operation.
But worrying about how the wiring "looks", how many "twists per inch" and such just borders on OCD and rediculousness.
Wise, it is for that reason that I just gave a recipe for how to competently make hanks of twisted pair from otherwise ordinary hookup singles. With a drill. Ideally, with ah helper. Such stuff comes in surprisingly handy and can be made far in advance of actual need. Cloth gloves are the magic. And a helper. 🙂
Neat idea GG.
However, being the Old School buzzard that I am, I hand-twist heater wiring, my projects are small, and hey, it kills some time, and I got plenty.
If we really want to get into whackoland, consider the CAT-5 and -6, etc. cables, that have 4 twisted pairs run together. Each pair is twisted at a different pitch, to minimize interaction. Riddle me that one, BatFriends!
All good fortune,
Chris
All good fortune,
Chris
Those cat cables transmit digital information at tremendous speeds, naturally.
So they have to minimise corruption of that information.
These are heater wires...my molding machine had them in a braided shield.
Copper covered with mica, over that Teflon/fiberglass braid, over that braided metal wire shielding.
2.5 square mm, single phase, that means line and neutral inside the same braided shield.
After many years, I replaced them with 4 square mm (mica covered) woven/braided fiberglass covered copper, as the old wires were stiff and chafing.
I used regular single wires in pairs, no shielding.
Try those shielded wires if you feel like, for heater supplies in tube radios.
Ground the shielding. No interference possible.
Copper covered with mica, over that Teflon/fiberglass braid, over that braided metal wire shielding.
2.5 square mm, single phase, that means line and neutral inside the same braided shield.
After many years, I replaced them with 4 square mm (mica covered) woven/braided fiberglass covered copper, as the old wires were stiff and chafing.
I used regular single wires in pairs, no shielding.
Try those shielded wires if you feel like, for heater supplies in tube radios.
Ground the shielding. No interference possible.
Last edited:
Hello, Mr. Goat,
I put two different colors of wire in that drill motor and twist them up. Two colors to keep track of polarity.
Thanks DT
Next I am going to try the star-quad style with 4 twisted wires.
For low level signal four wires have better hum rejection than 2 wires.
google "star quad"
Thanks DT
DualTriode, yes. Yes! But not for filament wires.
Once, not so many years back, I had a bunch more money than good sense, and in the charitable-giving spirit, funded the reworking of our mid-sized Church's choir-and-instrumental mixer, amplifiers, speakers, microphones (all of them), and so on. However, knowing my way around a soldering iron, and not at all amused with how crâhppy the commercial mic cables held up (no more than a couple of years, and they'd create all sorts of hard-to-track-down phantom problems, especially in live performance grrrrrrrr) …
I had read about Mogami brand wire, 4 wires, all co-twisted together in a sheath. A separate very flexible braid (not foil) as a secondary shield. Seemed really chic. That and the Neutrix brand of full german-silver (nickel-and-tin-and-copper) shileded XLR connectors at $20 a pair.
So, I bought 1,000 feet of the stuff, and with 'the boys', sat in our kitchen one long afternoon, stripping wire and making cables of all different lengths. Used up every inch of the cable, too. Could have bought another 500 feet, actually.
For about $1,000ℓ in outlay, we had over 35 cables, excellent, 4 wire, carefully potted, and well anchored cables. Several bottles o wine, too, went into the effort.
Thing is, some 10 years later, reports are that not a single XLR mic cable has failed yet. Even when people trip over then, and others have rolled 500+ pound wheels of grand pianos, of the Hammond 'ho-ho-ho-its-portable' organ, and everything else. Some bunch of yahoos even rolled a new d•••med ALTAR over the cables one evening while we were practicing. That kind of thing ruined the previous ones.
So. 4 wire, twisted. Mogami cable. Neutrix connectors. Do NOT waste time with anything less, if you want to really have cables that work. It was also remarkably good to have had the presence of mind to lay down little 'standard colors' color tape ((4 mm color over 10 mm other base color, 'readable' in any direction). Putting these on before soldering the last end allowed a sleeve of absolutely transparent heat-shrink to cover the tape; 10 years later, all the tape markers are still there.
To be able to glance across the stage, eyeball a cable's color code, have a glance at the mixer-board inputs, and know with REALLY high probability that you've got the right socket … has saved our bacon so many times.
Again, for tube filaments, all the reasons for twisting make sense, but none of the reasons for 4 wire twisted. No pianos rolling over the wires. THat's a first point. LOL.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
Once, not so many years back, I had a bunch more money than good sense, and in the charitable-giving spirit, funded the reworking of our mid-sized Church's choir-and-instrumental mixer, amplifiers, speakers, microphones (all of them), and so on. However, knowing my way around a soldering iron, and not at all amused with how crâhppy the commercial mic cables held up (no more than a couple of years, and they'd create all sorts of hard-to-track-down phantom problems, especially in live performance grrrrrrrr) …
I had read about Mogami brand wire, 4 wires, all co-twisted together in a sheath. A separate very flexible braid (not foil) as a secondary shield. Seemed really chic. That and the Neutrix brand of full german-silver (nickel-and-tin-and-copper) shileded XLR connectors at $20 a pair.
So, I bought 1,000 feet of the stuff, and with 'the boys', sat in our kitchen one long afternoon, stripping wire and making cables of all different lengths. Used up every inch of the cable, too. Could have bought another 500 feet, actually.
For about $1,000ℓ in outlay, we had over 35 cables, excellent, 4 wire, carefully potted, and well anchored cables. Several bottles o wine, too, went into the effort.
Thing is, some 10 years later, reports are that not a single XLR mic cable has failed yet. Even when people trip over then, and others have rolled 500+ pound wheels of grand pianos, of the Hammond 'ho-ho-ho-its-portable' organ, and everything else. Some bunch of yahoos even rolled a new d•••med ALTAR over the cables one evening while we were practicing. That kind of thing ruined the previous ones.
So. 4 wire, twisted. Mogami cable. Neutrix connectors. Do NOT waste time with anything less, if you want to really have cables that work. It was also remarkably good to have had the presence of mind to lay down little 'standard colors' color tape ((4 mm color over 10 mm other base color, 'readable' in any direction). Putting these on before soldering the last end allowed a sleeve of absolutely transparent heat-shrink to cover the tape; 10 years later, all the tape markers are still there.
To be able to glance across the stage, eyeball a cable's color code, have a glance at the mixer-board inputs, and know with REALLY high probability that you've got the right socket … has saved our bacon so many times.
Again, for tube filaments, all the reasons for twisting make sense, but none of the reasons for 4 wire twisted. No pianos rolling over the wires. THat's a first point. LOL.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅