I found this statement online regarding center taps vs. "virtual" center taps for tube heaters.
I read somewhere that the filament windings might not be balanced via the center tap, and you could achieve better balance by using matched 100R resistors with a virtual center tap.
When should we use an actual center tap vs. a "virtual" center tap? If I am having a power transformer wound for a small (EL84) PP power amp, and I need to specify whether it will have center taps on the heater windings or not, is there a demonstrable benefit either way?
I read somewhere that the filament windings might not be balanced via the center tap, and you could achieve better balance by using matched 100R resistors with a virtual center tap.
When should we use an actual center tap vs. a "virtual" center tap? If I am having a power transformer wound for a small (EL84) PP power amp, and I need to specify whether it will have center taps on the heater windings or not, is there a demonstrable benefit either way?
Last edited:
The center tap adds almost nothing to the cost and is usually a better way to ground the filaments. Unless you are building a high-gain preamp circuit and need a hum-balance pot, I'd add the center tap. You can always NOT use it, of course. ;-)
If your amplifier tubes are sensitive to a slightly un-balanced filament center tap, versus a pair of precision resistors used as a pseudo center tap
Then . . .
You need better quality tubes
You need a better circuit
You need DC power for your filaments
One, two, or three of the above.
I never had a problem with either a center tapped filament winding; and never had a problem with a two resistor pseudo center tap.
A typical problem with hum from an EL84 push pull amplifier, is if it uses a Concertina phase splitter, and
Either / And / Or
there is a leakage from the filament to the cathode (replace the tube)
the filament winding center tap/pseudo center tap - is not elevated to a DC voltage that is close to the concertina cathode voltage.
I once had a non elevated cathode in a concertina, there was hum.
Un-bypassed self bias resistors on the input stage can also be a problem if there is leakage from the filament to the cathode (replace the tube).
Just my opinions and experience
Your Mileage May Vary
Take care of that, and then all you have to worry about that can cause hum is . . .
Input ground loop
B+ ground loop, the rectifier, B+ center tap, and first filter cap must be a short path and self contained ground loop. Only then do you pass ground to the second filter cap.
output transformer angular orientation versus power transformer and choke angular orientation; and spacing to the output transformer
magnetic steel chassis
Most of my amplifiers have less than 100uV hum into 8 Ohms (still not good enough for headphones).
Have fun designing and building
Then . . .
You need better quality tubes
You need a better circuit
You need DC power for your filaments
One, two, or three of the above.
I never had a problem with either a center tapped filament winding; and never had a problem with a two resistor pseudo center tap.
A typical problem with hum from an EL84 push pull amplifier, is if it uses a Concertina phase splitter, and
Either / And / Or
there is a leakage from the filament to the cathode (replace the tube)
the filament winding center tap/pseudo center tap - is not elevated to a DC voltage that is close to the concertina cathode voltage.
I once had a non elevated cathode in a concertina, there was hum.
Un-bypassed self bias resistors on the input stage can also be a problem if there is leakage from the filament to the cathode (replace the tube).
Just my opinions and experience
Your Mileage May Vary
Take care of that, and then all you have to worry about that can cause hum is . . .
Input ground loop
B+ ground loop, the rectifier, B+ center tap, and first filter cap must be a short path and self contained ground loop. Only then do you pass ground to the second filter cap.
output transformer angular orientation versus power transformer and choke angular orientation; and spacing to the output transformer
magnetic steel chassis
Most of my amplifiers have less than 100uV hum into 8 Ohms (still not good enough for headphones).
Have fun designing and building
Last edited:
The need to 'balance' the heaters is a noval way to use poor grade valves that have a minor heater cathode leak.
Some people tie the heater centre tap to a possitive voltage other use a 220R pot to 'balance' them.
I use quality valves and the centre tap of the winding; 3.15 0 3.15v or a pot.
Some people tie the heater centre tap to a possitive voltage other use a 220R pot to 'balance' them.
I use quality valves and the centre tap of the winding; 3.15 0 3.15v or a pot.
I have seen a trimmer used to provide a path to ground.
The trimmer can be adjusted for least hum.
The trimmer can be adjusted for least hum.
Quality Valves?
. . . Yes.
How many of your preamp and line amplifier tubes have Spiral Wound Filaments in the cathode?
Example: Some early data sheets for the 12AU7 specify the Spiral filament.
The JJ small tubes I use have Spiral filaments, most of them do. Some JJ low to medium power output tubes also have spiral filaments.
If your tube has multiple up, down, up, down, up, down filament wires in the cathode, then try using some JJ tubes.
JJ's quality has been constantly increasing over the last 25 to 30 years.
I always get my latest production JJ tubes from Eurotubes.com. They very extensively re-test them, including for noise and hum.
They ship to many countries worldwide.
Well, It is a 25 minute trip to Eurotubes from my house. I drive over and pick up what I need (lucky me).
If you use DHT tubes, there is not much choice . . . up, down, up, down, up, down filament wires for all the original ones (new modified designs do not count, the are not the same as the original tubes).
. . . Yes.
How many of your preamp and line amplifier tubes have Spiral Wound Filaments in the cathode?
Example: Some early data sheets for the 12AU7 specify the Spiral filament.
The JJ small tubes I use have Spiral filaments, most of them do. Some JJ low to medium power output tubes also have spiral filaments.
If your tube has multiple up, down, up, down, up, down filament wires in the cathode, then try using some JJ tubes.
JJ's quality has been constantly increasing over the last 25 to 30 years.
I always get my latest production JJ tubes from Eurotubes.com. They very extensively re-test them, including for noise and hum.
They ship to many countries worldwide.
Well, It is a 25 minute trip to Eurotubes from my house. I drive over and pick up what I need (lucky me).
If you use DHT tubes, there is not much choice . . . up, down, up, down, up, down filament wires for all the original ones (new modified designs do not count, the are not the same as the original tubes).
It seems that the virtual center tap is a cheap and easy way to provide some benefit in a world where many power tubes are "less than ideal." Both EL84 PP amplifier kits that I built this year use a virtual center tap on the power tube heaters.
So this guitar tech I follow over on IG (AmpTech74) has an interesting view of this.
Octal-based tubes have the pin with the highest voltage, #3, right next to one of the pins of the heater circuit. In the event of a tube failure, there is a chance that the HV going through the heater circuit to find ground. If you are using a traditional center tap, that HV is going thru the 6.3v windings of your power transformer on its way to ground, and that might kill the whole PT. But if you are using a virtual center tap, the HV goes through the pair of 100R resistors, costing you a few pennies.
Of course, EL84 HV pins are on the other side, so it's less of a thing. I've taken to putting the 100R resistors tight in the socket of the first power tube, wired to the cathode, so the heater circuit is elevated that 10 volts or so.
Octal-based tubes have the pin with the highest voltage, #3, right next to one of the pins of the heater circuit. In the event of a tube failure, there is a chance that the HV going through the heater circuit to find ground. If you are using a traditional center tap, that HV is going thru the 6.3v windings of your power transformer on its way to ground, and that might kill the whole PT. But if you are using a virtual center tap, the HV goes through the pair of 100R resistors, costing you a few pennies.
Of course, EL84 HV pins are on the other side, so it's less of a thing. I've taken to putting the 100R resistors tight in the socket of the first power tube, wired to the cathode, so the heater circuit is elevated that 10 volts or so.
If your amplifier output stage is arcing . . .
Make sure you have a proper load on the output transformer secondary
Make sure the amplifier voltages do not exceed the tubes maximum voltage ratings
Clean the Rosin off of the tube sockets, and PCB traces.
Control the humidity of your listening room.
Use Quality output tubes; and tubes that were not abused before.
Make sure you have a proper load on the output transformer secondary
Make sure the amplifier voltages do not exceed the tubes maximum voltage ratings
Clean the Rosin off of the tube sockets, and PCB traces.
Control the humidity of your listening room.
Use Quality output tubes; and tubes that were not abused before.
Ohhhhhhh ... the enemy of all things electronic where I live. Humidity combined with a dust bunny will kill lots of things.
DC heating that does not also elevate the filament can sometimes be a problem.
The 7199 triode concertina that does not elevate the heater can develop filament to cathode leakage.
I measured one 7199 without filament elevation, it degraded over time. It had 100k Ohms leakage from the filament to the cathode.
Think of a concertina that has 33k Rp, and 33k Rk. Then consider 100k filament to cathode leakage that is in parallel with Rk.
With DC heating hum is not a problem, but the large un-balanced loads of the concertina cathode and plate is a problem.
There is a reason why tube manufacturers rate the maximum cathode to filament voltages.
The 7199 triode concertina that does not elevate the heater can develop filament to cathode leakage.
I measured one 7199 without filament elevation, it degraded over time. It had 100k Ohms leakage from the filament to the cathode.
Think of a concertina that has 33k Rp, and 33k Rk. Then consider 100k filament to cathode leakage that is in parallel with Rk.
With DC heating hum is not a problem, but the large un-balanced loads of the concertina cathode and plate is a problem.
There is a reason why tube manufacturers rate the maximum cathode to filament voltages.