I have been reading for a while and finally had a question that I could not find an answer to.
I understand how twisting wires can help reject unwanted common signals.
Wikipedia has a good short version of the benefits however they leave out a lot of the math.
I also understand that the heater voltage should be higher than the cathode so that it minimizes the influence of the plate/cathode current flow.
Is there a sound reference that explains the reason for twisting the heater wires?
I understand that high impedance, small signal parts of the circuit would benefit. Alexander Graham Bell figured this out.
I don't understand the need in a low impedance, non- amplified situation like heaters.
I connected a pair of un-twisted wires to my scope and saw a lot of rf in the +\- 5mV range. But as soon as I terminate the wires with a hundred ohm resistor the rf goes to zero. This sort of simulates the heater resistance.
I understand how twisting wires can help reject unwanted common signals.
Wikipedia has a good short version of the benefits however they leave out a lot of the math.
I also understand that the heater voltage should be higher than the cathode so that it minimizes the influence of the plate/cathode current flow.
Is there a sound reference that explains the reason for twisting the heater wires?
I understand that high impedance, small signal parts of the circuit would benefit. Alexander Graham Bell figured this out.
I don't understand the need in a low impedance, non- amplified situation like heaters.
I connected a pair of un-twisted wires to my scope and saw a lot of rf in the +\- 5mV range. But as soon as I terminate the wires with a hundred ohm resistor the rf goes to zero. This sort of simulates the heater resistance.
Just as twisting the wires reduces noise pick-up it also helps cancel noise emission. The heaters are not necessarily quieter, but all of the sensitive stuff (signal) around the wires don't have to listen to it.
It also pays to route the twisted heater wires away from the signal path in the corner of the chassis if possible to minimize pick-up.
It also pays to route the twisted heater wires away from the signal path in the corner of the chassis if possible to minimize pick-up.
It's Magnetic
The best reason to twist tube filament wiring is to reduce the emission of magnetic flux radiation type hum. The heater circuit of a few tubes in parallel can carry several amps of AC current and can radiate hum into audio grounds if filament wires aren't twisted to cancel hum flux radiation. That's my best guess.
The best reason to twist tube filament wiring is to reduce the emission of magnetic flux radiation type hum. The heater circuit of a few tubes in parallel can carry several amps of AC current and can radiate hum into audio grounds if filament wires aren't twisted to cancel hum flux radiation. That's my best guess.
Yes, but the hum pickup is taking place in the valve grid circuit, which is both high impedance and amplified.
Twisted pair provides immunity from induced (differential) noise (on the pair) because the wire picking up radiation differentially (the one nearest the source in the case of untwisted lines) is constantly being switched for the one furthest away. A similar reduction in radiated interference (hum radiation) is achieved in twisted pair by cancellation of the radiated fields. This can be thought of as analogous to the reciprocity theorem in antennas without resort to a mathematical treatment.
Not my best use of English, but you probably get the idea...
w
Last edited:
I concur with above comments, and can elaborate that distance is your friend between EMI emitter (heater wiring carrying 50Hz current) and antenna (high impedance grid circuits being amplified), given that radiation level drops with distance (ie. keep the two sets of wiring separated by as much distance as possible).
Radiation level also drops with heater current level (ie. either use discrete heater feeds to preamp sections, or make those feeds the 'end of the daisy-chain').
There are also many subtle techniques to improve hum immunity (such as DC elevation of heater AC voltage, and wiring configuration at the valve socket, and using a socket with centre metal tube.
Ciao, Tim
Radiation level also drops with heater current level (ie. either use discrete heater feeds to preamp sections, or make those feeds the 'end of the daisy-chain').
There are also many subtle techniques to improve hum immunity (such as DC elevation of heater AC voltage, and wiring configuration at the valve socket, and using a socket with centre metal tube.
Ciao, Tim
Two of the factors that affects a circuit's ability to radiate a magnetic field are the area enclosed by the circuit and the current through it. Heater wiring in its simplest case (the transformer with its two wires connected to a load) is essentially a single-turn coil with the heater current flowing through it. If the area of this coil is made very small by placing the conductors close to one another (most conveniently done by twisting them together) the radiated field will be minimized.
Conversely, the worst thing you can do is to wire the heaters in such a way as to run each of the conductors in opposite directions from the transformer so as to make a single loop enclosing the entire chassis.
I'm just looking at the radiated magnetic field here- none of this has anything to do with grounding, etc.
Nor does it have anything to do with the reason we use spiral heaters within the certain tubes.
Nor does it have anything to do with a balanced (differential) circuit's ability to reject a common-mode signal (such as a hum field)
This has only to do with a means to reduce the radiated magnetic field from the heater wiring -attacking the problem at its source.
Conversely, the worst thing you can do is to wire the heaters in such a way as to run each of the conductors in opposite directions from the transformer so as to make a single loop enclosing the entire chassis.
I'm just looking at the radiated magnetic field here- none of this has anything to do with grounding, etc.
Nor does it have anything to do with the reason we use spiral heaters within the certain tubes.
Nor does it have anything to do with a balanced (differential) circuit's ability to reject a common-mode signal (such as a hum field)
This has only to do with a means to reduce the radiated magnetic field from the heater wiring -attacking the problem at its source.
Last edited:
Hi
Forgive my ignorance but as more of an SS guy, why would you not use a regulated DC source for the low voltage heater current with local decoupling cap at each tube? Unless of course if your goal is to have a valve only circuit🙄 might complicate it a bit.😱
Small diodes and regulator IC's are very cheap
.
Forgive my ignorance but as more of an SS guy, why would you not use a regulated DC source for the low voltage heater current with local decoupling cap at each tube? Unless of course if your goal is to have a valve only circuit🙄 might complicate it a bit.😱
Small diodes and regulator IC's are very cheap
.
hiya cbs - sounds great, but often you want an elevated or floating supply - tough to do when regulating and often unnecessary if good lead dress is followed.
CBS, it's a pain having to rectify and filter an AC winding to get all that heater power. Most PT's have 6V3 windings, and require schottky's to reduce losses otherwise you go under voltage on the heater. Largish electrolytics are needed to cope with the ripple current. Rectifying the heater can introduce rectifier noise. And one often gets the same hum reduction with a much simpler elevated DC applied to the heater.
Wrt kruesi's comments, the tpi and turn uniformity, and thickness of insulation of the wires used for heater cabling can be an influence, but starts to get in to the extremes.
Tim
Wrt kruesi's comments, the tpi and turn uniformity, and thickness of insulation of the wires used for heater cabling can be an influence, but starts to get in to the extremes.
Tim
Don't bother with DC on the heaters, or with shielding.
Tests I ran years ago showed twisted and properly run AC lines gave results just as good as far as hum and noise was concerned. However proper routing is critical. Don't try to put them on a PCB. Larger bases (eg Octal valves 6V6.6SN7 etc )are easier to achieve good results than those with smaller valve bases. I also prefer to use lower gain triodes in the early stages. High gain pentode pre-amps can be difficult if the layout is poor.
Tests I ran years ago showed twisted and properly run AC lines gave results just as good as far as hum and noise was concerned. However proper routing is critical. Don't try to put them on a PCB. Larger bases (eg Octal valves 6V6.6SN7 etc )are easier to achieve good results than those with smaller valve bases. I also prefer to use lower gain triodes in the early stages. High gain pentode pre-amps can be difficult if the layout is poor.
PCB's are fine as long as you use twisted pair heater supplies as arched add ons, with space between the board. Last 2 amps with PCB's are quiet as a church mouse....
I even used twisted pair for the inputs.....
I even used twisted pair for the inputs.....
Just for interest,
Siver plated PTFE cables have a greater current carrying capacity at a smaller size. So it's easier to get a neat finish and works well with B9A bases. 🙂
Just an example:
RS | Cables | Equipment Wire/Instrumentation | PTFE Cable | PTFE (** 3G 210)
These are full rolls I guess others have a source that is cheaper.
Regards
M. Gregg
Siver plated PTFE cables have a greater current carrying capacity at a smaller size. So it's easier to get a neat finish and works well with B9A bases. 🙂
Just an example:
RS | Cables | Equipment Wire/Instrumentation | PTFE Cable | PTFE (** 3G 210)
These are full rolls I guess others have a source that is cheaper.
Regards
M. Gregg
Last edited:
Another hint is that the (turns per inch) is critical in a twisted pair and related to the frequency, as in this case 50/60 Hz...
Chris
Chris
So what is the correct TPI for 60 Hz? I can't relate cycles/second to length of wire. I've googled 6, but without explanation.
I'd suggest just twisting as best you can with what you've got - I'm sure their are photos of people twisting heater cables with a drill.
If you want to go to the level of ensuring a certain TPI, then you'll also be getting the protractor out to use the best crossover angle (which will then depend on clockwise rotation of twist) and physically locating equidistant with all other cables. And then you may as well use the smallest cable size and thinnest insulation to get wire spacing as minimal as possible. Not to mention star distribution. And I'm sure the list is longer...
Ciao, Tim
If you want to go to the level of ensuring a certain TPI, then you'll also be getting the protractor out to use the best crossover angle (which will then depend on clockwise rotation of twist) and physically locating equidistant with all other cables. And then you may as well use the smallest cable size and thinnest insulation to get wire spacing as minimal as possible. Not to mention star distribution. And I'm sure the list is longer...
Ciao, Tim
Roline, You mentioned using TP for inputs. Do you find them just as effective as shielded cable? Do you leave the ground side disconnected at the input grid and just ground at the input jack?
I'm not sure that most of the balanced mic cables I've seen are twisted tight enough.
I'm one of the "hand drill" users to twist my heater cables. I don't go as far as measuring the TPI, but I go for "very tight". 😉
I've built a few amps with DC heaters, but found that the disadvantages have to be balanced against the advantages, such as having to use a separate heater supply transformer and mounting the regulator away from sensitive components.
I agree entirely. I quess mic cable is never twisted so tightly is that the tighter the twist , the shorter the length.
I have recently completed a mic amp with 70dB of gain, that has AC heaters and there's no hum.
- Status
- Not open for further replies.