Preamp Heater Wires

[fiftynine]

......dropping down from the top of the tube Fender style or laid flat to the chassis and going across the back of the tube (not round it) with the preamp circuit wires pressed all over them like spaghetti?

[shooter]

Most, I suspect have done both with great success, and failure.  The key is keeping signal wires and DC wires at right angles from heaters(AC), shielded where needed, good air-gap tween 'em.  C.T. on the heaters.

[Willabe]

There's another thing to consider that HBP has posted here.

If you look at the Fender tweeds the heater wires were laid/dressed down into the corner lip of the chassis. The brown and black face amps had the heater wires run from above.

HBP can explain this much better than I can, but it's was because of the proximity of the eyelet boards parts, gride R's to the heater wires. 

Do a search for pictures of Fender chassis gut shots and you'll see it.     

[HotBluePlates]

......dropping down from the top of the tube Fender style or laid flat to the chassis ...

Well, it depends... How good are you at thinking in 3-D?

Fender did it at least 2 ways. Original tweed amps lay the heater wires against the chassis, but dressed away from the sockets and tucked into the lip formed at the chassis edge, as shown below.



Next image is an actual vintage Fender tweed amp, but someone re-did the heater wires (probably because the amp was old enough to ground 1 heater pin at all 6.3v tubes to use the chassis in place of one of the heater wires). Notice this puts the heater wiring closer to the board components, and also makes the working area more cluttered. Fender didn't do it that way, in these amps.



So the "Fender rule" is if the sockets are at 90-degrees to the board/components, dress the heaters against the chassis to keep them away from the board.

The photo below shows a 60's-era Deluxe Reverb. Now the sockets are on the same plane as the board, and the heaters are dressed up in the air and drop down to the sockets. That keeps them at max distance away from everything.



Not everyone followed these "rules" or even considered twisted heater wiring mandatory. McIntosh made many hi-fi items and typically rotated tube sockets so heater pins of each were aligned, then passed a bare solid wire straight through similar-pins of each socket. So that places 1 side of the heater wiring on either side of the socket. See below.



Soldano did the same, as well as making other common connections between output tube sockets in this buss-fashion.



I did a variation of some of these in an amp I built by passing my heaters flat along the chassis, but running wires from socket-to-socket directly on top of the socket instead of going in the air or around. When I generated my layout, I knew I'd have the components mounted between parallel strips with the tube sockets themselves below the components (the way Tektronix built their tube scopes). Running the heaters right-to-left against the chassis kep that wiring at 90-degrees to the socket pins themselves (on one plane) while also keeping it at 90-degrees to the components (on another plane). It seemed to me the neatest way to wire them while minimizing the chance of hum would be this unconventional arrangement. I've had zero hum issues.



So it just depends on your circumstances and what your experience tells you should work best (or well-enough). If you don't have the experience to draw from, copy what worked for others, as long as your situation also truly matches what their situation was.

[Ed_Chambley]

According to the layout I decide how I am going to run them.  Usually, I run them next to the chassis and across the tube sockets and then bring my other leads over and down.

Most amps I build are tube up designs.  Currently, I am using regulated DC heaters on all preamps.  I find if building separate preamps that plug into power amp DC seems to be best.

[Mike_J]

To me it all depends on the individual tube and how it needs to be wired.  In some, probably most, cases it is best to run the wires twisted tightly above the sockets for most of the builds I have made which are tubes on the same plane as the circuit board.  However, sometimes it makes sense to tuck the wires in the back for a specific tube in order to stay away from wires for the reasons mentioned above.


Thanks
Mike

[jjasilli]

I've seen it theorized that running heater wires close to the chassis increase parasitic capacitance to good effect in this case -- bleeding-off hum & noise from the heater wires.

[Fresh_Start]

HBP thank you so much for your examples and explanations. Adapting heater placement to your chassis and circuit board layout is such a great basic concept.

It wasn't until my 4th build that I really grasped use of all 3 dimensions to separate high current wires from sensitive, low voltage wires. I also solved a problem with the tremolo pulse getting into my PI by running the bias supply lead up above the circuit board and then dropping straight down to the insertion point.

JJ mentions running leads flat on the chassis. I've done that with plate wires to good effect (heaters up in the air).  If you Google "Hiwatt gut shots", you'll see another approach with the heaters flat on the deck - allowed by the big real estate inside a Hiwatt chassis.
I love this place!Chip

[HotBluePlates]

I've seen it theorized that running heater wires close to the chassis increase parasitic capacitance to good effect in this case -- bleeding-off hum & noise from the heater wires.

I think this is unlikely to be the case.

How much capacitance is there? I tried measuring capacitance from a heater pin to ground in one of my amps but only got an "OL" reading. I got the same reading pressing my meter probes together. So I don't know if that represents a very high capacitance or a very low one. It could also be the 100Ω resistors from heater to chassis swamping the reading.

How much capacitance to the chassis might thee be? Typically, wiring stray capacitance to ground/chassis is estimate around 100pF. Maybe we could argue that should be higher in this case. Should we agree to call heater wiring 400-500pF to the chassis?

The heaters are 60Hz and do not have as much added harmonics as we'd find in the high voltage B+ supply. So what's the impedance to ground to that 60Hz for even 500pF of heater-to-chassis capacitance?

X_C = 1/(2*Π*F*C) = 1/(2*Π*60Hz*0.0000000005 Farads) = ~5.3MΩ

I'd argue the heater-to-chassis capacitance would have to be at least as large as your amp's filter caps (maybe bigger) to have a useful shunting effect for noise.

And if it doesn't, you already have a heater center-tap or 100Ω resistors to ground to shunt noise more effectively than the capacitance. And by experience/experiment, you should already know that heaters are noisy without the CT or resistors.

[Paul1453]

The concept to consider is the magnetic field inducing voltage and current in nearby wires.

Exactly how this magnetic field is affected by nearby metal that is tied to earth ground, is a question I would like to know more about.   

I would not expect the nearby grounded metal to completely suck up this magnetic field.

I'm not even sure that the grounded metal would completely stop this magnetic field from penetrating to wires on the other side of the metal.   

I know that magnetism and electricity are interrelated, but not how all that works.   

Any Electro-Magnetic scientists care to try to enlighten us?

[shooter]

but not how all that works

I spent 24yrs fixing giant magnets, MRI's, was paid well to fix them, not engineer them.  But for fun, find a 3T magnet, cram your head in and out of the bore real fast
start here;
https://en.wikipedia.org/wiki/Electromagnetic_field
once you get Maxwell equations wired; check back