Virtual Center Tap Resistor Value

[mbratch]

I'm familiar with the "virtual center tap" for the heater wiring in an amp using 100 ohm resistors. My question is: why is it always 100 ohms? Can it be a different value not too far from that? Say, 150 ohms?

[Chuck H]

Dunno for sure. I've seen lower, like 47ohm resistors used. I think the low value is intended to keep the impedance of the circuit low. Higher shouldn't be a problem other than that. HUMMMMMMMMMMMM!!!!!!!!!!!!!!!!!!!!!!

[mooreamps]

It probably could be a different value. But I think the intention here is ; if using resistors ; match them up as a pair within a small tolerance.

Ex : R1 = 99.5 ohms
R2 = 99.6 ohms

That sort of thing....

-g

[Chuck H]

Ahh yes. This assuming all the tubes that share a filament are matched each side for gain and the power tubes are matched. There is also the option of using a pot as a "hum balance" control. One of the few good silver face additions. This allows you to intentionally imbalance the position of the CT to find the least hum. I wouldn't use just a pot though. Probably a 100 ohm pot and a pair of 47 ohm resistors (one on each end) and the wiper at 0V or your DC elevation point. Actually you probably wouldn't need the adjustment if your DC elevating the CT. I usually just get a pair of 100 ohm 1 watt resistors (and yes Gary, I do get 1% for these ) and CT them at a 30V to 50V elevated DC source.

Point is... I can't imagine an advantage to using higher value resistors other than "it's what I have on hand". You should be able to walk into Fry's or Radio Shack and get a pair of 100 ohm 1 watt resistors.

Oh... The reason not to get too low a value would probably be to not interfere with the lower impedance of the winding itself. Go too low and you will start to dump current to ground.

[redelephant]

The center tap deal on heater windings is a left-over from the directly heated tubes days...Have anyone here actually found any difference when simply connecting one side of the heater winding to GND? I dont think so!
Per old tube manuals, and basic understanding of how electrons may leak into the cathode, the best way to eliminate heater hum is to elevate the potential of the heater to a voltage above the potential of the cathodes of sensitive tubes (preamp). In a cathode biased amp, the easiest way is to connect one side to the positive side of the cathode resistor. Or u can connect it to any voltage divider that provides positive 15-80volts.

[Enzo]

Actually Chuck, low resistance wouldn;t shunt current to ground, there is no circuit for current to ground. As suggested in the last post, you can connect one point of the circuit directly to ground. However the two resistors in series are across the winding, so no need to heat them up with a low value. The value doesn;t matter a lot.

And yes, redelephant, I have many times seen it matter. Not all hum balance pots are best nulled at center. SOme like at the end better, and some like off-center somewhere. Yes, one could add in the wiring to elevate the heaters DC, but two resistors seems to work just fine, and those only cost Leo Fender maybe a penny each.


Hum balance pots were generally wire wound instead of carbon. A 100 ohm pot would have to dissipate close to a half watt.

[mooreamps]

Hum balance pots were generally wire wound instead of carbon. A 100 ohm pot would have to dissipate close to a half watt.

0.3969 watts @ 6.3 volts... A 1 watt rated pot should suffice for this purpose....



-g

[Steve Conner]

Have anyone here actually found any difference when simply connecting one side of the heater winding to GND? I dont think so!

I have! That really does increase the hum level in even a moderately high-gain amp.

I also found that if I apply a positive DC voltage to the heater string, the hum balance is much less critical, so that's my preferred method now.

If you have a bias tap, you can add another diode and cap to get +60V out of it. In my last build, I couldn't do that because it was cathode biased, so I connected the whole heater string to the cathode of the power tube to put about +9V on it.

[mbratch]

Point is... I can't imagine an advantage to using higher value resistors other than "it's what I have on hand". You should be able to walk into Fry's or Radio Shack and get a pair of 100 ohm 1 watt resistors

That's the scenario. I happen to have some 150's handy. We don't have Fry's here in the NE, and Radio Shack part stock is dwindling, in favor of cell phones, overpriced PC accessories, and electronic toys, and other sundry home-use electronics.

But... I still haven't seen any explanation as to why it's almost exactly 100 ohms other than "that's what everybody uses".

[Pedro Vecino]

I use 250 Ohm pots, especially in Engl (its adjustable resistors fail often) and work fine.

[pdf64]

Engineering is about using proven solutions to get the required outcomes, minimising risk and development cost - experimentation is for scientists and should be the last resort. Therefore as a value of 100 ohms for the balancing resistors is know to work, there's no benefit in risking the use of 150.
Whereas you don't have those engineering constraints, and can experiment to your heart's content, so give your 150 ohm resistors a go in this application.

[Steve Conner]

Bah pdf64, if everyone thought like you, we would all still be driving around in horseless carriages based on 150-year-old gasoline burning technology. Oh, wait.

I don't see any reason in the world why 150 wouldn't work. But if you're worried, why not use two 150s in parallel for each resistor, that's 75 ohms.

[Chuck H]

*palm smacks forehead*

That'll do. 75 ohms per side. My soldering skills are adequate at best so I try not to do things like parallel resistors in the interest of tidiness I guess. So I didn't even think it.

But, the primary issue is still open. Why is 47 to 125 ohms per side the "range" that has been seen in common applications of this circuit??? If shunting isn't an issue and "the value hardly matters" why do we never see 10 ohm resistors used. or 1M?

[Merlinb]

The balance resistors could be almost any value.

However, if they are very large then the heater supply becomes more prone to picking up radio noise.
If they are very small then they consume lots of current, and dissipate lots of power, which is wasteful. 100ohms is a small, VERY common value that doesn't draw much current (so you can safely use 1/4W resistors / humdinger).
Bu you could use anything from 100 ohms to 1k ohms, no problem, and even bigger if you don't encounter radio hash.

They do not need to be specially matched since the method of hum coupling into the audio circuit is never perfectly balanced in nature.

[mooreamps]

But, the primary issue is still open. Why is 47 to 125 ohms per side the "range" that has been seen in common applications of this circuit??? If shunting isn't an issue and "the value hardly matters" why do we never see 10 ohm resistors used. or 1M?

Perhaps the intent would be to put a "light" load on the secondary coil of the filament winding... Well, what would that mean ?? If we say we have a filament winding rated for 4 amperes, then 6.3/4 is a winding impedance of 1.575 ohms.. A 100 ohm load would present a "light" load on the secondary winding. A 1 megohm load would be no load at all ; practically... 10 ohms would be more a moderate load and could start to sag on the output voltage. Also, 10 ohms at 6.3 volts would draw ; of course ; 630 milliamperes dissipating approx 4 watts through a 10 ohm resistor...