Filament wiring gauge

[Graydigger]

Hello -

I’m a new member to the forum although I’ve sought wisdom from it’s members for years.

There’s a saying: better to let people think you are an idiot than open your mouth and prove it.  Well, I’m an idiot and I’m proving it with my first post. 

I was trying to figure out why so many folks prefer to use 18 ga. solid core cloth wire for “vintage” filament wiring?  Is there a reason a smaller gauge wire wouldn’t be just fine in this application? (It sure is easier for me to twist/untwist, and make neat runs). I’ve asked a couple of friends that build/mod amps professionally and the answers ranged from:, “it stands up to heat better”,  “that’s what we have always used” to “it’s stiffer (e.g. when they “fly” the filament wiring)”.

Seems to me that since the filament current is low, the voltage is low, heat at the terminals is not extreme, the runs are short between the filaments, the total wire runs aren’t super long, and the resistance difference between a couple feet of 18 vs. 22 ga wire would would seem to be much less than the resistance of the filaments themselves in series. At 60 Hz for non-signal carrying wire runs, is 20-30 mOhms difference between 22 ga and 18 ga in this application that significant? I can’t envision that inductance or reactance of the difference in gauge would make a difference for this application.  Is the reason because most PT’s might have 18 gauge wire?

What am I missing?

Thank you for your help!

[sluckey]

I don't use 18AWG. 20AWG is fine for all the stuff I've built.

[acheld]

You're not missing anything.

22AWG is sufficient for heater wiring.    Ampacity of 22ga (@ 140 F) is 3A, for 20ga 5A.   As you noted, there is a minimal difference in resistance (which accounts for the current heating the wire), but in our guitar amp world, it's not relevant.   If you push 7A through 22ga wire, it will heat not to exceed ~ 170 F, assuming the ambient temp is 85F, and that it is not bundled or in conduit. 

I have been guilty of using 18ga wire -- but grew weary of twisting it down to the tube socket, and have since stopped, moving on to 20ga solid core.    I can make that look neat enough. 

Recently discovered Weico stranded 22ga wire -- that stuff is lovely to work with and for general hookup use, the insulation is actually rated to 600V.  May give that a try as heater wiring in my next build.

[thetragichero]

i used 16 or 18awg solid wire for heaters on a couple of builds and never again. besides being difficult to work with, solid wire ALWAYS ends up breaking on me right before it was in and soldered. stranded wire all day

[Leevi]

https://www.calculator.net/voltage-drop-calculator.html


/Leevi

[trobbins]

One subtlety is that you may notice the insulation on a lighter gauge feed wire shrivelling up and smelling/smoking if there is a fault at a valve base - better that than the power transformer winding becoming collateral damage.

[Mike_J]

I use 18AWG Teflon twisted to the power tubes and 20AWG Teflon twisted to the preamp tubes. No reason other than that is how I have done it. Used to use solid core wire but it can break where you can't see it after you twist it and is hard to locate without more effort than I want to have to expend again.

[scstill]

I just connected heaters in my new build (5c3 style in buick radio chassis) with 22g stranded silcone (previously used 20g)
Tube compliment here is about 3amp. I really like the new smaller gauge.
https://www.amazon.com/gp/product/B01K4RPAH6/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1

My question is why do we twist the wire?
Is it better to have a shorter non-twisted run? or to twist and have a longer run?
I have opted for a short non-twisted run between the power tubes (6v6) and PI (6sc7), do you think this could give me issues?

A followup question is what are some popular color choices for heater wire?
Any reason to use different colors in the twist?

BTW - the two vintage sockets in my build do not have a use so I left them.
Do you like that look? or should I drill them out as well?

[shooter]

My question is why do we twist the wire?

was a sudo "standard"
the only consideration on fil wires, they are "dirty power", keep them away from sensitive signal.  IF they must be close, cross at 90 degrees

[PRR]

> why do we twist the wire?

It tends to cancel electromagnetic field.

[scstill]

My question is why do we twist the wire?

was a sudo "standard"
the only consideration on fil wires, they are "dirty power", keep them away from sensitive signal.  IF they must be close, cross at 90 degrees

I was inspired to clean the heater route up (I think)
The 6v6 socket pair (right most at the bottom) could have been inverted to get heater connections closer to chassis, but I had already modified mounting plates to fit so will take into consideration next time.
At least now the signal lines can be routed somewhat away from heater.

Should I keep 6v6 plate to OT wire away from heater as well?
What about other plate HV? Is it preferred to be on the heater side?

[shooter]

cross at 90 degrees

[trobbins]

scstill, are you really fitting in a pair of 6V6 like that?  How does the top side of that sub-chassis vent?  Are you running them in class C ?

The OT wires (primary side especially), have high signal voltage levels, so separation distance from other wiring/parts is the main issue, and the further the better as the signal level of any part or wire gets smaller (or circuit impedance gets higher) - so 6V heater wiring is already a relatively high signal voltage level.  The more cramped a layout is, the more likely you will get capacitive coupling from heater wiring to sensitive circuits like grids, so a humdinger pot is likely your first line of defence (followed by heater phase swapping on the input tube).

[scstill]

scstill, are you really fitting in a pair of 6V6 like that?  How does the top side of that sub-chassis vent?  Are you running them in class C ?

The OT wires (primary side especially), have high signal voltage levels, so separation distance from other wiring/parts is the main issue, and the further the better as the signal level of any part or wire gets smaller (or circuit impedance gets higher) - so 6V heater wiring is already a relatively high signal voltage level.  The more cramped a layout is, the more likely you will get capacitive coupling from heater wiring to sensitive circuits like grids, so a humdinger pot is likely your first line of defence (followed by heater phase swapping on the input tube).

This is how the original radio was designed, 6v6 in push pull. See the original picture and Schematic.
The rear metal panel is covered with vent holes so I'm hoping the design will suffice as guitar amp.
I'm using the original 6V6 tubes and OT. The orig design used a 16ohm 6x9.

On the subject of OT, I measure the orig one to have 50:1 turns ratio (applied small AC voltage to primary)
I am using a beefier Jensen speak but retaining the 16ohm impedence as the radio used
By my calculation this would be 40kohm to the 6v6, which would be the original design of the radio.
But by spec, doesn't 6v6 like to see 8kohm?? What am I missing?

[Mike_J]

A followup question is what are some popular color choices for heater wire?
Any reason to use different colors in the twist?

I use green and white. The reason is I believe I read somewhere that you need to make sure you keep the same wires going to pins 2 and 7 of the power amp tubes. Never sure it wasn't the preamp tubes that need to have the same wire going to pins 4/5 and 9, although that doesn't seem to make much sense. Anyway I wire my amps so the green wire goes to the same pins be they preamp or power tubes and white wires go to the other side. Using different colors makes doing so much easier.

[sluckey]

Back before vintage became a word, many well known electronics component manufacturers used brown for filament wiring. Can you trace the filanent strings for this example?

     http://sluckeyamps.com/hammond_2/big_guts.jpg
 

[PRR]

> are you really fitting in a pair of 6V6 like that?

That's how they did it. My brother has a 1941 Plymouth with a Dodge radio, just that jam-packed and hot. Note though that it runs nominal 250V, lower than we do on stage. These radios have terrible reliability. Cutting-out the RF and IF helps just a little.

Works as a radio does not ensure it will be a happy guitar amp. The radio has only one stage of audio gain before the final; a g-amp needs two. The radio has a tube-before but it runs IF not AF, a great help in system stability.

[trobbins]

Perhaps re-do the OT turns ratio measurement.  You may not have enough secondary voltage to make an accurate measurement - depends on your meter.  Also somewhat better to apply a voltage (eg. 12Vac) across the half-primary (CT to anode) and measure the voltage on the other half-primary and the secondary.

[scstill]

> are you really fitting in a pair of 6V6 like that?

That's how they did it. My brother has a 1941 Plymouth with a Dodge radio, just that jam-packed and hot. Note though that it runs nominal 250V, lower than we do on stage. These radios have terrible reliability. Cutting-out the RF and IF helps just a little.

Works as a radio does not ensure it will be a happy guitar amp. The radio has only one stage of audio gain before the final; a g-amp needs two. The radio has a tube-before but it runs IF not AF, a great help in system stability.

Do you think the guitar amp would be happier if I separated the tubes more?
I have the two spots with original sockets that I could put the preamps into
Then move the PI into one of the middle spots
Finally separate the 6V6's by one tube space.

[PRR]

I would consider moving the first preamp further away from the 6V6es and OT.

[scstill]

Perhaps re-do the OT turns ratio measurement.  You may not have enough secondary voltage to make an accurate measurement - depends on your meter.  Also somewhat better to apply a voltage (eg. 12Vac) across the half-primary (CT to anode) and measure the voltage on the other half-primary and the secondary.

applied 12.14vac Pri1 to CT; meas 12.04vac Pri2 to CT & .471vac sec; comes out 42k:16ohms right?
odd for a 6v6 design unless there is something wrong with OT or something unique with radio design
I have a identical set still as radio that I will measure tomorrow,

Wondering if if it makes sense to stay with this OT or replace it with one for 6V6 guitar amp at 8000:16?

[sluckey]

I have the two spots with original sockets that I could put the preamps into

I highly recommend replacing those two wafer sockets if you intend to use them. They are terrible quality.

[trobbins]

Yup 42k to 16 ohm - thanks for confirming.  Are you able to confirm that OT belongs to that chassis?  Over decades, it's not unusual for parts to be swapped 'if they work', so just checking if that particular OT is confirmed to be for that chassis.  Given you effectively measured the same voltage on the other half-primary winding, then it seems that the OT is ok.  So sounds like you should look out for a replacement OT to suit the available space - perhaps even 10k:16 if you are at 250V B+ (and losing some volts for cathode bias).

[scstill]

OT looks original is soldered to chassis (will need to grind this off, I rarely have luck heating a chassis solder).
How does this 42k:16 OT work with the push pull 6V6 pair with a plate voltage at 250v?

[scstill]

A followup question is what are some popular color choices for heater wire?
Any reason to use different colors in the twist?

I use green and white. The reason is I believe I read somewhere that you need to make sure you keep the same wires going to pins 2 and 7 of the power amp tubes. Never sure it wasn't the preamp tubes that need to have the same wire going to pins 4/5 and 9, although that doesn't seem to make much sense. Anyway I wire my amps so the green wire goes to the same pins be they preamp or power tubes and white wires go to the other side. Using different colors makes doing so much easier.

Interesting....
What is the reason to keep same filament pin numbers connected if its AC voltage?
some in-phase thing? Seems odd for something that is heating for electron flow.

What do you do if some octals are 2-7 and others are 7-8 (as is the case in my build)
Do you just wire left to left and right to right side from the tube sheet?

As follow-up, answered my own questionl Now I wonder it I should rewire my other builds ....
https://300guitars.com/300s-tips/300-tip-wiring-output-tube-filaments-to-reduce-hum/
https://robrobinette.com/How_To_Build_a_Tube_Amp.htm#Heater_Wiring_Phase
Think I will change my name to always learning :-)

[PRR]

I would really expect a 4 Ohm speaker here. Yes, it is very odd that someone could find an exact-fit for an in-box car radio. But full primary to 3.01 Ohm secondary gives 8,000, as we'd expect; 4 Ohm makes 10k which is a common 2*6V6 loading.

[scstill]

I would really expect a 4 Ohm speaker here. Yes, it is very odd that someone could find an exact-fit for an in-box car radio. But full primary to 3.01 Ohm secondary gives 8,000, as we'd expect; 4 Ohm makes 10k which is a common 2*6V6 loading.

The Orig Speaker has three terminals as shown
The speaker green goes to the OT and the yellow goes to chassis ground.
The OT sec goes to speaker green and the other side to chassis ground

Measure 9.4ohm between the speaker green (center) and yellow (gnd on the left which goes to coil on speaker)
Measure 3.5ohm from the speaker green center to the speaker right terminal.

Wouldn't the OT be seeing the 16ohm since it is the green and yellow that are connected to OT?
Or does it somehow see the 4ohm??

[PRR]

The schematic above in #13 shows a PM speaker.

Your photographs show a field-coil speaker. B+ and Gnd got to it somehow.

Your radio is a year older than that schematic?

We expect the field coil to be "around" 10 Watts (like the amp output).

Measure Yellow to Ground. That's the field coil.

Voice coil is Green to Ground. You say this is 3.5 Ohms, a very likely value.

Implication is that the Field Coil is about 6 Ohms, which at 6V is 6Watts which is a very reasonable value.

[scstill]

So I re-located a couple of the sockets, what do you think?
and stuffed in a new OT which the 6v6's should like a little better (8500/16ohm @15w), thoughts?
I will use color coded filament wire to keep them wired to the same pins....

[acheld]

Wow!  That is a packed chassis . . . 

[trobbins]

That seems like the best outcome you have available for part layout, given the input valve is as far from everything else as possible.

If you get noticeable hum/noise, or some accidental feedback that causes some resonant peaks, then I guess be prepared for some methodical 'fault-finding' to isolate main causes (ie. don't 'bury' parts or terminals if practical, as future access could be needed).  That type of fault-finding typically needs a spectrum analyser software using a soundcard and probe, and an AC meter with good low mV resolution connected to a speaker output resistor load.

[wdelaney72]

All of my builds have 22awg 600v stranded for filament.....no issue at all. same wire for circuit board. Solid wire is nice in that it the twisted wire holds in place better. Twist or not twist....Marshall got real lazy in the 80's here and you can see some JCM amps with no twisting. I always twist just to be on the safe side, but I highly doubt 18awg is necessary.


The twisted wire also looks clean

[shooter]

does the OT have room to rotate 90 degrees?  or the PT?

[wdelaney72]

I kept this one retro....ala 12xxx series 1968. A little vintage hum never hurt anyone 

[scstill]

does the OT have room to rotate 90 degrees?  or the PT?

Yes. Lucky for me the PT is relatively square.
Hopefully the hum will be non existant with the revised layout.
Thanks

[scstill]

Is the routing of High Voltage B+ important
Does it matter if B+ to OT feeding 6v6 PushPull plates is close to PT (like picture)?

Does it matter if other B+ is close to filaments?
Twist with other B+ or leave loose? Short as Possible?

[shooter]

as long as it's not close to preamp stuff you should be ok, I would look at "in front" of PT, maybe through a cable clamp attached to the PT bolt 

[scstill]

as long as it's not close to preamp stuff you should be ok, I would look at "in front" of PT, maybe through a cable clamp attached to the PT bolt 

I decided to run it underneath the PT. Separated by the chassis metal.
Taking care to keep power and filament away from signal.
Also planning shielded wire for signal line.