Wiring question... Bundles/twist or separate?

[Bob Simpson]

I can't find the answer with the search...

I've noticed looking at the pictures that some builders seem to bundle with wire ties, or twists, and some just seem to throw them in the general vicinity of where they're connected...

Is there a general preference?

Heater wiring twisted, but some in the air, and some round the back of the chassis...

OT secondarys to the switch/jacks...

Bob

[Phil_S]

Hi Bob,
I'll take a stab at this one. In general, there are several generally accepted principles in play, but implementation can take a number of variations.

First, I'd say that a/c and d/c wiring should be separated. That is why you see the filament wires separated from the high voltage DC and signal wires. Some of us separate them by putting them up in the air and others by routing them on the opposite side of the tube socket from the other wires. Both work. A/C supplies are twisted as a hum canceling measure. This includes filament wires and any other wires coming from the power transformer.

HT (high tension) or HV (high voltage), same thing, or B+ is the DC power supply. These are not signal wires and these should probably be kept as clear of the signal wires as possible. Neat freaks like to bundle wires; others just lay them out loose. Again, both give acceptable results. If signal wires must cross, cross them at right angles.

The wires to the output transformer and to the speaker are also carrying a/c, so it is a good idea to twist those, as well.

I think I kept this pretty simple, so it doesn't cover everything, but it should cover you for most of what you see in an amp.

[Structo]

Yes, it all falls under the heading of proper lead dressing.

You will notice after looking at a lot of pictures of tube amps built by people that know what they are doing that there is a similar theme.

As Phil pointed out, if wires have to be close or cross each other, it's best done at a 90 degree angle.
That way the lines of force surrounding the wire does not affect the other wire as much.

Interestingly enough, Mr. Dumble liked to run his plate and cathode wires close together on the way to the board and some have suggested that it adds to the tone of the amp.

Fender always ran their heater wires overhead and a few other modern companies do this as well.

Marshall always ran their heater wires down on the floor in the back corner of the chassis.
Both work and achieve the same end, which is to keep them away from low voltage signal wires.

The power section of an amp can sometimes be a challenge to dress properly.
A lot of wires and a small space to contain them.

I found on one of the first amps I built that I had some hum.
I double checked my lead dress around the tubes and everything looked like it should.

However when I bundled the power transformer wires I had bundled the Standby wire with them which is a DC power wire.
It was picking up the 60 Hz noise from the AC wires and injecting it into the B+ supply.
Once I separated that wire and ran it by itself the amp quieted down.

[David Root]

I've seen that note about Dumble running plate & cathodes wires together before. Did he twist them together or just run them side by side, and what would be the mechanism of the tone modification that might make?

From a layout perspective the plate wire DC current is running from the plate to the plate resistor and the cathode wire DC current is running from the cathode resistor to the cathode, which is the opposite direction when you put them side by side. (Assuming plate and cathode resistors are board mounted). Similarly the AC signal current is flowing in opposite directions too.

Could there be some kind of phase modification involving the cathode bypass cap and/or the plate coupling cap?

Maybe I'm dreaming in Technicolor too. (I also do that.)

[Bob-I]

I've seen that note about Dumble running plate & cathodes wires together before. Did he twist them together or just run them side by side, and what would be the mechanism of the tone modification that might make?

I'm not sure why, but he ran them side by side. It seems to smooth out the amp somewhat. I'll bet theres a small amount of capacitance in this lead dress taking some of the extreme highs out.

As far as general lead dress, I've found it best to examine as many pics as you can find on the type of amp you're building, and follow. Why not learn from others.

[martin manning]

I think that's it, Bob. The small capacitance formed between the parallel cathode and plate leads acts like the small-value snubber cap often seen soldered from plate to cathode. Since the cathode resistor is bypassed to ground with another cap, this trick shorts HF on the plate signal to ground.

MPM

[Bob-I]

I think that's it, Bob. The small capacitance formed between the parallel cathode and plate leads acts like the small-value snubber cap often seen soldered from plate to cathode. Since the cathode resistor is bypassed to ground with another cap, this trick shorts HF on the plate signal to ground.

MPM

I tried to measure the capacitance on the parallel leads. I found that just moving the leads to the meter changed the measurement from 3-5pF.

Yes, capacitance is a factor, but I'll get there something else as well. Bottom line, any lead movement makes a difference, sometimes very subtle but combining all the little differences can make a big difference.

[tubeswell]

Having high potential DC wires next to low-potential ones is theoretically not a problem (as long as insulation is adequate to stop arcing), but in practice most DC supply wires have a small amount of residual AC in them after filtering, which is not good next to sensitive signal wires.

Any AC pairs should be twisted including the HT secondaries and rectifier and heater windings, and even the OT primaries (and secondaries).

2CW

[DonMoose]

I tried to measure the capacitance on the parallel leads. I found that just moving the leads to the meter changed the measurement from 3-5pF.

At vacuum tube impedances, 3-5pF matters. Only a little, but it matters.

[Structo]

I think that's it, Bob. The small capacitance formed between the parallel cathode and plate leads acts like the small-value snubber cap often seen soldered from plate to cathode. Since the cathode resistor is bypassed to ground with another cap, this trick shorts HF on the plate signal to ground.

MPM

I tried to measure the capacitance on the parallel leads. I found that just moving the leads to the meter changed the measurement from 3-5pF.

Yes, capacitance is a factor, but I'll get there something else as well. Bottom line, any lead movement makes a difference, sometimes very subtle but combining all the little differences can make a big difference.

I wonder if it is something more than just the capacitance going on there.
I was thinking it was more like induction between the two wires, so the lines of force around the wires are interacting with each other.
I would think since the plate has much more voltage that it would induce some effect into the cathodes.
Heck, I don't know.

[DonMoose]

Magnetic coupling is about current.

Capacitive coupling is about voltage.

Hope this helps!

[Andy Le Blanc]

Here's the inside of an old galvanometer, mil.spec. pieces are even cleaner.
One surplus projector set I saw had the wire bundles run thru braided wire
tube that was bolted to inside of the chassis.
The models used for a piece of wire are a resistor in series with an inductor
and a capacitor to ground, and they all add up differently with length and gauge.
I like the old fashion bundle and lace, with lower gain amps the coupling and
wire proximity issue isn't so bad, but you still have to watch it. Chopsticking
something like a BF or SF fender, opposed to a lead Marshal layout is a great demo.
Insulation type is also an issue, I've seen a outfit on the web catering to the
ultra high end Hi-Fi, offering 99.9 pure gold, silver and nickel alloy conductor
with a braided cotton tube insulator saying that they found it to have the least
capacitive interaction in a circuit. Makes sense in phone cartridge and arm
before a pre.

[Bob Simpson]

I appreciate all the info.

I'm working on building a Blue Monkey reverb 88, ( slowly ), and I'd like to do it "right".

I've looked at all the "Dumble" pictures I can find, d'lites, Ceriatone, et. al., and I'm getting confused.

Thanks.

Bob

[Structo]

Magnetic coupling is about current.

Capacitive coupling is about voltage.

Hope this helps!

How would you define cross talk between two adjacent wires?