Filament Wiring - Shielded vs. Twisted?

[k9swc]

Anyone have any thoughts or preferences about whether to use shielded wire rather than tightly twisted for filaments in a preamp?

I've built several using the twisted method but there always seems to be a slight hum remaining at higher volumes, and I'm considering duplicating the circuit using some silver plated, shielded, teflon insulated cable instead, grounding the shield at each tube (3) and source. Will the extra work be worth it?

Thanks,
Bud

[David Davenport]

Hi Bud,

Twisting greatly reduces the magnetic field that is radiated. That, together with careful routing of the filament wiring should be all you need.

Shielding reduces the electric field, but that is really not a problem with hum.

Dave

[zigzagflux]

FWIW, the proper manner would be to elevate the heaters as personal preference dictates, connecting the center tap of your transformer to this DC reference voltage. The shield, at the source end, would be tied not to ground, but to this DC reference. A common misconception. Furthermore, I would suggest leaving the shield unterminated at the tube end. If using IDHT, small ceramic caps from each heater pin to earth is a good idea.

I have investigated multiple instrumentation/shielding/grounding books, and done a bit of research at Analog Devices site, and they always recommend the same 'optimal' shielding method as above. Granted, most of these applications were regarding instrumentation amplifiers sensing mV or uV signals, but what works in the 'signal' domain should apply just as well in the 'power' domain. Consistent recommendations have convinced me, and in practical use, it works darn well. My entire system is AC heated, with two of those stages DHT. Hum is less than 1mV rms with the volume control at max.

That is not to say you NEED shielded leads, as Dave has said. But if you do want to use shields, that is the correct way to terminate them.

my $0.02

[David Davenport]

Hi zigzagflux,

Thanks for your informative post - that is good to know.

Dave

[Brett]

Excellent, concise synopsis zigzag.

[Funker]

Hi folks,
I will underwrite Zigzagflux comment.

I had my own experience with my hombrew mc preamp. Its got AC heated valves . One of the main thing is that the heater circuit is well balanced against ground. The heater windings center tap straped to 0V . The 0V rail is only groundet near the input. I used twisted and shielded wires.The shield is only attached on one side by the trafo end. These are telecom shielded patch wires which normal used on the MDF to establish subscriber circuits for PCM / ISDN or DSL services. The hum level is 70db below +6db line level. (50Hz)

regards from Hamburg

Wolfgang

[mashaffer]

Wolfgang, a few questions if I might. What is the gauge of the twisted pairs in that cable? Can you point to an example of it? Is this a European spec item or is there a US equiv. for which I can look? Do you daisy chain the heater wiring as normal or is there an individual run to each tube directly from the PT? If daisy chained how do you connect the shields?

Thanks for your patience.

[Funker]

Hi mashaffer,

the wire is DIN/ IEC specified. Its designation is YVO(st)2x05/09. I do not know the US equivalent , but I´m sure Belden or ITT make a similar kind of wire.

Here the pics of the wire:




The wire size is 0,5mm in diamter, which is 0,2mm² . the DC loop resistance is 90 mOhms . This can handle 3 Amps by a loss of 0.3 Volts at 1m lenght (3.3 feet).

I arrange an individual wire to each tube to the mains trafo. the shields are grounded at trafo side.

some pics of ma hombrew preamp:

Power supply board:



main wiring:



top view of the preamp:



mains unit and amp fitted :




you can see the grey wires in the pics. I use the same wire for the level potmeters.

The preamp has 3 heater windings , each one for the both input tube with an individual hum balance potmeter. And the third winding for the line driver tubes.

The preamp is mainly made out of srap material from our lacal srapyard.

regards from Hamburg
Wolfgang

[Gabdx1]

Zigzag is totally right. Shielded wire, ground the insulation at the source.

I would like to add that going DC can help a lot too, some tubes catch hum of the heater filaments. Again regulation is simple : 4 diodes, resistance (+wire) , big cap. bingo.

[Johan Potgieter]

After all said, just that I have used 6VAC on heaters of RIAA stages. That is, to put matters in perspective, 6 300mV not influencing a 0,3mV signal entry point, equalising included. That was with careful routing but unscreened, so it is possible. I used a pot across the 6V to adjust the earth-point; any residual hum was better balanced out that way.

That said, it is relatively easy these days to connect all low current heaters in series and feed with 36 - 60Vdc - whatever is required. It could even be regulated for extra smoothing; regulators are <1 US$. Only, to warn about simply 6VAC through a bridge+cap, if such a source is not well-filtered, the charging rectification current peaks could easily introduce more interference than 'smooth' 6VAC would have done.

I have also found elevation up to say 30 - 40V advantageous, should there be heater-cathode leakage somewhere.